=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 41a40305d136800f944133e53e6b8ab400f7c868 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, SPGLIB 1.16.2, LIBGRPP 20231215 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 25.01.2024 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-hdf5 --with-libvdwxc --with-pexsi --with-plumed --with-sirius=no; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBGRPP := 20231215 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 #USE_DEEPMD := 2.2.7 #USE_SIRIUS := 7.5.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS #LIBVDWXC_VER := 0.4.0 #SPFFT_VER := 1.0.6 #SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_DEEPMD),) USE_DEEPMD := $(strip $(USE_DEEPMD)) DEEPMD_INC := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/include DEEPMD_LIB := $(INSTALL_PATH)/libdeepmd_c-$(USE_DEEPMD)/lib CFLAGS += -I$(DEEPMD_INC) DFLAGS += -D__DEEPMD LIBS += $(DEEPMD_LIB)/libdeepmd.so LIBS += $(DEEPMD_LIB)/libdeepmd_c.so LIBS += $(DEEPMD_LIB)/libdeepmd_cc.so LIBS += $(DEEPMD_LIB)/libdeepmd_dyn_cudart.so LIBS += $(DEEPMD_LIB)/libdeepmd_op.so LIBS += $(DEEPMD_LIB)/libdeepmd_op_cuda.so LIBS += $(DEEPMD_LIB)/libtensorflow_cc.so.2 LIBS += $(DEEPMD_LIB)/libtensorflow_framework.so.2 endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBGRPP),) USE_LIBGRPP := $(strip $(USE_LIBGRPP)) LIBGRPP_INC := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/include LIBGRPP_LIB := $(INSTALL_PATH)/libgrpp-main-$(USE_LIBGRPP)/lib CFLAGS += -I$(LIBGRPP_INC) DFLAGS += -D__LIBGRPP LIBS += $(LIBGRPP_LIB)/liblibgrpp.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/01 job id: 51775559 --- Point --- name: 10 plot: h2o_32_ri_rpa_mp2 regex: Total RI-RPA Time= label: RI-RPA (8n/2r/6t) --- Point --- name: 11 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-RPA (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-MP2.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-HF-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-MP2-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/02 job id: 51775561 --- Point --- name: 20 plot: h2o_32_ri_rpa_mp2 regex: Total MP2 Time= label: RI-MP2 (8n/6r/2t) --- Point --- name: 21 plot: h2o_32_ri_rpa_mp2_mem regex: Estimated peak process memory label: RI-MP2 (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/03 job id: 51775562 --- Point --- name: 100 plot: h2o_64_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 101 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/04 job id: 51775563 --- Point --- name: 102 plot: h2o_64_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 103 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/05 job id: 51775564 --- Point --- name: 104 plot: h2o_64_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 105 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/06 job id: 51775565 --- Point --- name: 106 plot: h2o_64_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 107 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/07 job id: 51775566 --- Point --- name: 108 plot: h2o_64_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 109 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-64 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-64.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/08 job id: 51775567 --- Point --- name: 110 plot: h2o_64_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 111 plot: h2o_64_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/09 job id: 51775568 --- Point --- name: 200 plot: h2o_128_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 201 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/10 job id: 51775569 --- Point --- name: 202 plot: h2o_128_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 203 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/11 job id: 51775570 --- Point --- name: 204 plot: h2o_128_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 205 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/12 job id: 51775571 --- Point --- name: 206 plot: h2o_128_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 207 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/13 job id: 51775572 --- Point --- name: 208 plot: h2o_128_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 209 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-128 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-128.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/14 job id: 51775573 --- Point --- name: 210 plot: h2o_128_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 211 plot: h2o_128_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/15 job id: 51775577 --- Point --- name: 400 plot: h2o_256_md regex: CP2K label: (8n/12r/1t) --- Point --- name: 401 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/16 job id: 51775579 --- Point --- name: 402 plot: h2o_256_md regex: CP2K label: (8n/6r/2t) --- Point --- name: 403 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/17 job id: 51775583 --- Point --- name: 404 plot: h2o_256_md regex: CP2K label: (8n/4r/3t) --- Point --- name: 405 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/18 job id: 51775586 --- Point --- name: 406 plot: h2o_256_md regex: CP2K label: (8n/3r/4t) --- Point --- name: 407 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/19 job id: 51775587 --- Point --- name: 408 plot: h2o_256_md regex: CP2K label: (8n/2r/6t) --- Point --- name: 409 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-256 test - DBCSR dominated (MPI/OMP) input file: benchmarks/QS/H2O-256.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 30 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/20 job id: 51775589 --- Point --- name: 410 plot: h2o_256_md regex: CP2K label: (8n/1r/12t) --- Point --- name: 411 plot: h2o_256_md_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/21 job id: 51775590 --- Point --- name: 500 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/12r/1t) --- Point --- name: 501 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 6 # threads/rank = 2 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/22 job id: 51775592 --- Point --- name: 502 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/6r/2t) --- Point --- name: 503 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/6r/2t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 4 # threads/rank = 3 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/23 job id: 51775594 --- Point --- name: 504 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/4r/3t) --- Point --- name: 505 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/4r/3t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 3 # threads/rank = 4 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/24 job id: 51775595 --- Point --- name: 506 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/3r/4t) --- Point --- name: 507 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/3r/4t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/25 job id: 51775596 --- Point --- name: 508 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/2r/6t) --- Point --- name: 509 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/2r/6t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 (NREP 3) linear scaling test (864 H2O) input file: benchmarks/QS_DM_LS/H2O-dft-ls.inp required files: [] output file: result.log # nodes = 8 # ranks/node = 1 # threads/rank = 12 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/26 job id: 51775597 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 20 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/27 job id: 51775599 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 172669. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 139.180 139.181 farming_run 1 2.0 138.321 138.339 139.151 139.154 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 4194304 0.0% 0.0% 100.0% flops 14 x 32 x 32 154140672 0.0% 0.0% 100.0% flops 29 x 32 x 32 159645696 0.0% 0.0% 100.0% flops 14 x 14 x 32 208732160 0.0% 0.0% 100.0% flops 29 x 14 x 32 212860928 0.0% 0.0% 100.0% flops 14 x 29 x 32 212860928 0.0% 0.0% 100.0% flops 29 x 29 x 32 227352576 0.0% 0.0% 100.0% flops 14 x 32 x 14 896801644032 0.0% 0.0% 100.0% flops 29 x 32 x 14 928925089792 0.0% 0.0% 100.0% flops 14 x 32 x 29 928925089792 0.0% 0.0% 100.0% flops 29 x 32 x 29 962100985856 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693169221632 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753639550976 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.164741E+12 0.0% 0.0% 100.0% flops max/rank 447.801317E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249492158 0.0% 0.0% 100.0% number of processed stacks 164328 0.0% 0.0% 100.0% average stack size 0.0 0.0 1518.3 marketing flops 7.165779E+12 ------------------------------------------------------------------------------- # multiplications 1160 max memory usage/rank 1.538925E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.005 0.023 119.581 119.582 qs_energies 1 2.0 0.000 0.000 119.347 119.350 mp2_main 1 3.0 0.000 0.000 117.231 117.233 mp2_gpw_main 1 4.0 0.031 0.044 116.202 116.205 mp2_ri_gpw_compute_in 1 5.0 0.179 0.196 97.127 97.459 mp2_ri_gpw_compute_in_loop 1 6.0 0.003 0.004 57.312 57.643 mp2_eri_3c_integrate_gpw 272 7.0 0.150 0.164 43.760 48.838 get_2c_integrals 1 6.0 0.008 0.009 39.032 39.640 integrate_v_rspace 273 8.0 0.439 0.459 25.519 30.104 fft_wrap_pw1pw2 5465 10.4 0.066 0.069 28.113 28.720 grid_integrate_task_list 273 9.0 20.763 26.292 20.763 26.292 fft_wrap_pw1pw2_100 2178 11.4 1.312 1.447 25.559 26.193 compute_2c_integrals 1 7.0 0.002 0.002 21.454 21.458 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 20.838 21.165 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.426 20.836 21.162 rpa_ri_compute_en 1 5.0 0.019 0.022 18.952 19.057 cp_fm_cholesky_decompose 12 8.2 17.527 18.101 17.527 18.101 fft3d_s 5443 12.4 16.872 17.303 16.893 17.322 cholesky_decomp 1 7.0 0.000 0.000 16.416 16.989 ao_to_mo_and_store_B_mult_1 272 7.0 10.766 15.314 10.766 15.314 calculate_wavefunction 272 8.0 5.394 5.535 12.902 13.800 calc_potential_gpw 544 9.5 0.004 0.004 10.823 11.661 rpa_num_int 1 6.0 0.001 0.018 10.684 10.694 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.632 10.663 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 9.793 10.553 potential_pw2rs 545 10.0 0.106 0.108 8.523 10.351 calc_mat_Q 8 8.0 0.000 0.000 9.474 9.562 contract_S_to_Q 8 9.0 0.000 0.000 8.899 8.985 collocate_single_gaussian 272 10.0 0.038 0.040 7.848 8.644 parallel_gemm_fm 14 9.1 0.000 0.000 8.500 8.588 parallel_gemm_fm_cosma 14 10.1 8.500 8.588 8.500 8.588 create_integ_mat 1 6.0 0.014 0.028 7.844 7.854 array2fm 1 7.0 0.000 0.000 6.741 7.301 pw_gather_s 2722 12.2 4.005 4.903 4.005 4.903 pw_poisson_solve 545 10.5 0.010 0.011 4.664 4.887 pw_scatter_s 2720 12.7 4.439 4.750 4.439 4.750 pw_poisson_set 548 11.5 0.018 0.018 3.339 3.560 array2fm_buffer_send 1 8.0 2.987 3.124 2.987 3.124 pw_copy 4911 11.6 2.618 2.763 2.618 2.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=116.201891, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2787.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 200775. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.304 536.348 536.351 farming_run 1 2.0 533.091 533.252 536.064 536.134 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.274089E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 9782991. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 325.462 325.462 qs_energies 1 2.0 0.003 0.026 325.132 325.132 mp2_main 1 3.0 0.012 0.098 210.001 210.005 mp2_gpw_main 1 4.0 0.107 0.346 209.051 209.063 mp2_ri_gpw_compute_en 1 5.0 0.037 0.041 128.371 130.455 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.739 1.906 126.733 126.739 scf_env_do_scf 1 3.0 0.000 0.000 114.842 114.842 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.956 113.962 rebuild_ks_matrix 4 6.0 0.000 0.000 113.955 113.961 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.060 113.955 113.961 hfx_ks_matrix 4 8.0 0.001 0.001 113.604 113.607 integrate_four_center 4 9.0 0.153 0.472 113.603 113.606 mp2_ri_gpw_compute_en_expansio 172 7.0 0.464 0.497 111.022 112.033 local_gemm 172 8.0 110.558 111.564 110.558 111.564 integrate_four_center_main 4 10.0 0.096 0.538 102.035 105.118 integrate_four_center_bin 268 11.0 101.940 104.948 101.940 104.948 init_scf_loop 1 4.0 0.000 0.000 96.678 96.678 mp2_ri_gpw_compute_in 1 5.0 0.066 0.077 80.143 81.171 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 56.758 57.763 mp2_eri_3c_integrate_gpw 91 7.0 0.147 0.160 44.531 49.233 integrate_v_rspace 95 8.0 0.397 0.568 28.165 32.659 fft_wrap_pw1pw2 1868 10.4 0.028 0.034 30.553 31.155 fft_wrap_pw1pw2_100 730 11.4 0.645 0.719 28.255 28.888 grid_integrate_task_list 95 9.0 23.456 28.211 23.456 28.211 ao_to_mo_and_store_B_mult_1 91 7.0 10.559 26.982 10.559 26.982 get_2c_integrals 1 6.0 0.000 0.000 23.295 23.341 compute_2c_integrals 1 7.0 0.002 0.003 22.275 22.303 compute_2c_integrals_loop_lm 1 8.0 0.003 0.018 21.820 22.163 mp2_eri_2c_integrate_gpw 1 9.0 1.732 1.894 21.817 22.145 fft3d_s 1823 12.4 19.242 19.879 19.256 19.892 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.162 18.162 calc_potential_gpw 182 9.5 0.002 0.002 12.365 13.284 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 10.753 11.379 calculate_wavefunction 91 8.0 2.020 2.059 10.044 10.784 potential_pw2rs 186 10.0 0.033 0.035 9.094 10.515 collocate_single_gaussian 91 10.0 0.025 0.155 8.347 9.204 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.528 7.867 9.056 mp_sendrecv_dm3 2068 8.0 5.895 7.053 5.895 7.053 integrate_four_center_load 4 10.0 0.000 0.000 6.750 6.754 hfx_load_balance 1 11.0 0.000 0.000 6.750 6.754 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=209.052950, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1527.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 445.521920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.060 62.783 62.785 qs_mol_dyn_low 1 2.0 0.023 0.063 62.152 62.204 qs_forces 11 3.9 0.002 0.003 56.622 56.644 qs_energies 11 4.9 0.003 0.005 55.125 55.157 scf_env_do_scf 11 5.9 0.000 0.001 48.478 48.478 scf_env_do_scf_inner_loop 108 6.5 0.015 0.044 46.058 46.059 qs_scf_new_mos 108 7.5 0.000 0.001 35.345 35.636 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.344 35.636 dbcsr_multiply_generic 2286 12.5 0.095 0.099 34.846 35.368 ot_scf_mini 108 9.5 0.002 0.002 33.638 33.820 velocity_verlet 10 3.0 0.047 0.137 32.761 32.824 multiply_cannon 2286 13.5 0.188 0.195 26.699 28.132 multiply_cannon_loop 2286 14.5 1.801 1.905 25.858 27.310 ot_mini 108 10.5 0.001 0.001 20.496 20.751 qs_ot_get_derivative 108 11.5 0.001 0.001 17.451 17.639 mp_waitall_1 245248 16.5 9.080 15.302 9.080 15.302 multiply_cannon_metrocomm3 54864 15.5 0.072 0.078 6.222 13.158 multiply_cannon_multrec 54864 15.5 3.646 5.733 7.656 11.049 qs_ot_get_p 119 10.4 0.001 0.002 8.391 8.692 rebuild_ks_matrix 119 8.3 0.000 0.000 8.366 8.510 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.100 8.366 8.510 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.406 7.537 mp_sum_l 7287 12.8 5.808 7.508 5.808 7.508 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.982 6.425 multiply_cannon_sync_h2d 54864 15.5 5.140 6.080 5.140 6.080 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.882 5.998 dbcsr_mm_accdrv_process 76910 16.1 1.814 2.861 3.925 5.669 mp_sum_dm 438 4.9 5.416 5.549 5.416 5.549 update_particle_set 20 4.0 0.000 0.000 5.365 5.466 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.351 5.411 md_output 10 3.0 0.000 0.000 0.062 5.400 md_write_output 11 3.9 0.009 0.168 0.070 5.364 write_trajectory 44 4.9 0.002 0.009 0.056 5.195 write_particle_coordinates 11 5.9 0.054 5.189 0.054 5.189 init_scf_run 11 5.9 0.000 0.001 5.162 5.163 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.162 5.162 sum_up_and_integrate 119 10.3 0.008 0.053 4.817 4.828 integrate_v_rspace 119 11.3 0.002 0.002 4.799 4.819 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.505 4.655 calculate_rho_elec 119 8.7 0.012 0.026 4.504 4.655 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.519 4.519 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.250 4.251 cp_fm_redistribute_end 50 14.0 2.168 4.220 2.175 4.223 cp_fm_diag_elpa_base 50 14.0 2.040 4.108 2.044 4.115 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.924 3.136 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.940 3.135 apply_single 119 13.6 0.000 0.000 2.940 3.135 calculate_dm_sparse 119 9.5 0.000 0.000 2.942 3.106 calculate_first_density_matrix 1 7.0 0.001 0.009 2.815 2.820 ot_diis_step 108 11.5 0.006 0.006 2.768 2.768 jit_kernel_multiply 13 15.8 2.048 2.737 2.048 2.737 acc_transpose_blocks 54864 15.5 0.215 0.234 2.200 2.735 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.503 2.588 density_rs2pw 119 9.7 0.003 0.004 2.451 2.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.487 2.490 init_scf_loop 11 6.9 0.000 0.000 2.394 2.394 wfi_extrapolate 11 7.9 0.002 0.013 2.249 2.249 potential_pw2rs 119 12.3 0.004 0.004 2.162 2.204 make_m2s 4572 13.5 0.053 0.055 2.060 2.137 grid_integrate_task_list 119 12.3 2.011 2.115 2.011 2.115 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.006 2.057 make_images 4572 14.5 0.132 0.139 1.978 2.054 fft_wrap_pw1pw2 1201 11.6 0.011 0.013 1.904 2.005 mp_sum_d 4139 12.0 1.368 1.922 1.368 1.922 fft3d_ps 1201 13.6 0.366 0.472 1.681 1.778 mp_alltoall_d11v 2130 13.8 1.522 1.741 1.522 1.741 transfer_rs2pw 487 10.6 0.006 0.006 1.621 1.733 transfer_pw2rs 487 13.2 0.005 0.006 1.538 1.562 fft_wrap_pw1pw2_140 487 12.2 0.047 0.052 1.457 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.406 1.439 grid_collocate_task_list 119 9.7 1.354 1.424 1.354 1.424 mp_waitany 12084 13.8 1.294 1.423 1.294 1.423 acc_transpose_blocks_sync 164592 16.5 1.185 1.421 1.185 1.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=62.785000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=424.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 484.274176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 934262. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.052 42.221 42.221 qs_mol_dyn_low 1 2.0 0.003 0.003 41.793 41.858 qs_forces 11 3.9 0.003 0.004 41.706 41.706 qs_energies 11 4.9 0.002 0.003 39.973 39.977 scf_env_do_scf 11 5.9 0.001 0.001 33.725 33.725 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.960 30.960 dbcsr_multiply_generic 2286 12.5 0.119 0.123 23.179 23.548 qs_scf_new_mos 108 7.5 0.001 0.001 21.398 21.620 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.397 21.619 ot_scf_mini 108 9.5 0.002 0.003 20.447 20.608 velocity_verlet 10 3.0 0.011 0.029 19.699 19.700 multiply_cannon 2286 13.5 0.210 0.220 17.539 19.107 multiply_cannon_loop 2286 14.5 1.183 1.252 16.290 17.766 ot_mini 108 10.5 0.001 0.001 12.578 12.806 mp_waitall_1 200699 16.5 6.280 11.642 6.280 11.642 qs_ot_get_derivative 108 11.5 0.001 0.001 10.195 10.358 multiply_cannon_metrocomm3 27432 15.5 0.071 0.073 4.221 9.908 multiply_cannon_multrec 27432 15.5 1.838 4.137 6.479 9.491 rebuild_ks_matrix 119 8.3 0.000 0.000 7.349 7.499 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 7.349 7.499 dbcsr_mm_accdrv_process 47894 16.0 3.647 6.345 4.560 6.814 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.479 6.615 qs_ot_get_p 119 10.4 0.001 0.001 4.862 5.085 init_scf_run 11 5.9 0.000 0.001 4.952 4.952 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.952 4.952 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.893 4.739 sum_up_and_integrate 119 10.3 0.001 0.001 4.187 4.192 integrate_v_rspace 119 11.3 0.002 0.003 4.173 4.178 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.136 4.174 calculate_rho_elec 119 8.7 0.021 0.024 4.135 4.173 mp_sum_l 7287 12.8 2.080 4.162 2.080 4.162 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.021 4.097 apply_single 119 13.6 0.000 0.000 3.021 4.097 make_m2s 4572 13.5 0.052 0.053 3.102 3.406 make_images 4572 14.5 0.206 0.247 3.012 3.318 calculate_first_density_matrix 1 7.0 0.000 0.000 3.270 3.273 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.079 3.097 init_scf_loop 11 6.9 0.000 0.000 2.737 2.737 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.661 2.662 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.618 2.621 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.459 2.542 jit_kernel_multiply 11 16.2 0.854 2.537 0.854 2.537 density_rs2pw 119 9.7 0.003 0.004 2.428 2.516 calculate_dm_sparse 119 9.5 0.000 0.000 2.290 2.373 ot_diis_step 108 11.5 0.011 0.011 2.333 2.333 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.275 2.276 multiply_cannon_sync_h2d 27432 15.5 1.701 2.270 1.701 2.270 cp_fm_redistribute_end 50 14.0 1.153 2.246 1.156 2.249 cp_fm_diag_elpa_base 50 14.0 1.058 2.151 1.087 2.185 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.090 2.131 acc_transpose_blocks 27432 15.5 0.112 0.118 1.702 2.090 potential_pw2rs 119 12.3 0.006 0.006 1.930 1.952 grid_integrate_task_list 119 12.3 1.834 1.949 1.834 1.949 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.822 1.869 prepare_preconditioner 11 7.9 0.000 0.000 1.788 1.814 make_preconditioner 11 8.9 0.000 0.000 1.788 1.814 make_images_data 4572 15.5 0.047 0.053 1.352 1.789 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.679 1.742 transfer_rs2pw 487 10.6 0.005 0.006 1.591 1.672 wfi_extrapolate 11 7.9 0.001 0.001 1.616 1.616 fft3d_ps 1201 13.6 0.513 0.567 1.542 1.584 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.180 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.458 1.468 mp_alltoall_d11v 2130 13.8 1.329 1.448 1.329 1.448 transfer_pw2rs 487 13.2 0.004 0.005 1.383 1.405 fft_wrap_pw1pw2_140 487 12.2 0.047 0.051 1.346 1.397 grid_collocate_task_list 119 9.7 1.290 1.348 1.290 1.348 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.254 1.302 make_images_sizes 4572 15.5 0.005 0.005 0.802 1.121 mp_alltoall_i44 4572 16.5 0.797 1.116 0.797 1.116 mp_allgather_i34 2286 14.5 0.670 1.092 0.670 1.092 mp_sum_d 4139 12.0 0.567 1.042 0.567 1.042 dbcsr_complete_redistribute 329 12.2 0.196 0.380 0.944 1.016 acc_transpose_blocks_kernels 27432 16.5 0.187 0.274 0.755 1.012 qs_energies_init_hamiltonians 11 5.9 0.037 0.061 0.990 0.991 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.927 0.940 acc_transpose_blocks_sync 82296 16.5 0.810 0.938 0.810 0.938 mp_alltoall_z22v 1201 15.6 0.769 0.855 0.769 0.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.221000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.363636, yerr=1.067940 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 515.334144E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.113 0.274 43.078 43.103 qs_mol_dyn_low 1 2.0 0.028 0.116 40.861 40.869 qs_forces 11 3.9 0.004 0.015 39.697 39.716 qs_energies 11 4.9 0.022 0.035 38.052 38.079 scf_env_do_scf 11 5.9 0.000 0.001 30.732 30.732 scf_env_do_scf_inner_loop 108 6.5 0.015 0.069 27.508 27.508 dbcsr_multiply_generic 2286 12.5 0.119 0.123 20.061 20.293 velocity_verlet 10 3.0 0.031 0.107 19.032 19.075 qs_scf_new_mos 108 7.5 0.001 0.001 18.491 18.525 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.491 18.525 ot_scf_mini 108 9.5 0.002 0.002 17.633 17.655 multiply_cannon 2286 13.5 0.194 0.197 15.477 16.554 multiply_cannon_loop 2286 14.5 0.857 0.894 14.489 15.663 ot_mini 108 10.5 0.001 0.001 10.913 10.952 qs_ot_get_derivative 108 11.5 0.001 0.001 9.286 9.310 multiply_cannon_multrec 18288 15.5 1.854 2.835 7.288 7.720 rebuild_ks_matrix 119 8.3 0.000 0.000 6.614 6.654 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.613 6.654 dbcsr_mm_accdrv_process 38222 16.0 5.287 6.202 5.344 6.266 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.829 5.865 mp_waitall_1 158411 16.6 4.033 5.536 4.033 5.536 init_scf_run 11 5.9 0.000 0.001 4.993 4.993 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.993 4.993 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.732 4.501 qs_ot_get_p 119 10.4 0.001 0.002 4.243 4.280 sum_up_and_integrate 119 10.3 0.001 0.001 4.094 4.099 integrate_v_rspace 119 11.3 0.003 0.003 4.081 4.088 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.838 3.854 calculate_rho_elec 119 8.7 0.031 0.032 3.838 3.854 calculate_first_density_matrix 1 7.0 0.000 0.000 3.560 3.561 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 2.124 3.416 init_scf_loop 11 6.9 0.000 0.000 3.196 3.196 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.779 2.787 calculate_dm_sparse 119 9.5 0.000 0.001 2.612 2.632 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.218 2.626 apply_single 119 13.6 0.000 0.000 2.218 2.626 mp_sum_l 7287 12.8 1.937 2.518 1.937 2.518 make_m2s 4572 13.5 0.044 0.045 2.370 2.511 density_rs2pw 119 9.7 0.003 0.003 2.372 2.485 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.446 2.447 make_images 4572 14.5 0.193 0.205 2.284 2.426 acc_transpose_blocks 18288 15.5 0.077 0.080 2.259 2.370 prepare_preconditioner 11 7.9 0.000 0.000 2.362 2.367 make_preconditioner 11 8.9 0.000 0.000 2.362 2.367 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.209 2.295 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.086 2.097 cp_fm_diag_elpa_base 50 14.0 2.059 2.072 2.084 2.094 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.065 2.073 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.057 2.063 grid_integrate_task_list 119 12.3 1.793 1.893 1.793 1.893 potential_pw2rs 119 12.3 0.007 0.007 1.870 1.881 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.864 1.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.852 1.867 qs_energies_init_hamiltonians 11 5.9 0.030 0.052 1.832 1.854 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.808 1.832 transfer_rs2pw 487 10.6 0.005 0.005 1.638 1.782 ot_diis_step 108 11.5 0.012 0.012 1.564 1.564 acc_transpose_blocks_kernels 18288 16.5 0.215 0.222 1.434 1.540 fft3d_ps 1201 13.6 0.533 0.553 1.493 1.520 create_qs_kind_set 1 2.0 0.000 0.000 1.133 1.386 read_qs_kind 2 3.0 0.162 0.484 1.133 1.386 parser_read_line 2821 4.0 0.001 0.001 0.971 1.378 parser_read_line_low 5 5.0 0.004 0.111 0.970 1.377 broadcast_input_information 5 6.0 0.002 0.003 0.966 1.377 wfi_extrapolate 11 7.9 0.001 0.001 1.374 1.374 jit_kernel_transpose 5 15.6 1.219 1.328 1.219 1.328 transfer_pw2rs 487 13.2 0.004 0.004 1.315 1.328 grid_collocate_task_list 119 9.7 1.234 1.318 1.234 1.318 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.231 1.303 fft_wrap_pw1pw2_140 487 12.2 0.061 0.063 1.278 1.302 mp_bcast_i_src 20 7.0 0.249 1.289 0.249 1.289 make_images_data 4572 15.5 0.047 0.051 0.982 1.205 mp_bcast_am_src 5 7.0 0.714 1.198 0.715 1.199 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.144 1.190 multiply_cannon_sync_h2d 18288 15.5 0.960 1.179 0.960 1.179 mp_sum_dm 438 4.9 1.072 1.149 1.072 1.149 cp_fm_cholesky_invert 11 10.9 1.102 1.108 1.102 1.108 update_particle_set 20 4.0 0.000 0.000 1.034 1.106 qs_create_task_list 11 7.9 0.010 0.048 0.933 1.053 mp_alltoall_d11v 2130 13.8 0.875 1.037 0.875 1.037 generate_qs_task_list 11 8.9 0.151 0.322 0.923 1.013 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.841 0.986 parallel_gemm_fm 81 9.0 0.000 0.000 0.976 0.979 parallel_gemm_fm_cosma 81 10.0 0.976 0.979 0.976 0.979 md_write_output 11 3.9 0.016 0.406 0.054 0.975 md_output 10 3.0 0.000 0.000 0.032 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.857 0.863 mp_alltoall_z22v 1201 15.6 0.773 0.862 0.773 0.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=43.103000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=490.363636, yerr=2.100767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 552.898560E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.084 40.989 40.994 qs_mol_dyn_low 1 2.0 0.029 0.141 40.456 40.488 qs_forces 11 3.9 0.002 0.003 39.498 39.530 qs_energies 11 4.9 0.015 0.050 37.675 37.699 scf_env_do_scf 11 5.9 0.011 0.036 31.539 31.544 scf_env_do_scf_inner_loop 108 6.5 0.008 0.024 27.502 27.502 dbcsr_multiply_generic 2286 12.5 0.103 0.106 20.372 20.516 velocity_verlet 10 3.0 0.101 0.246 20.227 20.284 qs_scf_new_mos 108 7.5 0.001 0.001 18.867 18.930 qs_scf_loop_do_ot 108 8.5 0.001 0.003 18.866 18.929 ot_scf_mini 108 9.5 0.004 0.014 17.820 17.885 multiply_cannon 2286 13.5 0.219 0.226 16.151 16.709 multiply_cannon_loop 2286 14.5 1.493 1.613 15.117 15.478 ot_mini 108 10.5 0.001 0.001 10.634 10.715 multiply_cannon_multrec 27432 15.5 2.456 3.123 9.023 9.315 qs_ot_get_derivative 108 11.5 0.001 0.001 8.738 8.805 dbcsr_mm_accdrv_process 47916 15.9 6.118 7.793 6.465 7.855 rebuild_ks_matrix 119 8.3 0.000 0.000 6.846 6.899 qs_ks_build_kohn_sham_matrix 119 9.3 0.035 0.067 6.846 6.898 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.124 6.176 qs_ot_get_p 119 10.4 0.002 0.002 4.268 4.340 init_scf_run 11 5.9 0.001 0.002 4.327 4.328 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.327 4.328 init_scf_loop 11 6.9 0.035 0.125 4.004 4.018 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.882 3.924 calculate_rho_elec 119 8.7 0.040 0.047 3.882 3.924 sum_up_and_integrate 119 10.3 0.001 0.001 3.840 3.849 integrate_v_rspace 119 11.3 0.002 0.003 3.825 3.838 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.182 3.679 prepare_preconditioner 11 7.9 0.000 0.000 2.800 2.824 make_preconditioner 11 8.9 0.001 0.007 2.800 2.823 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.739 2.749 make_m2s 4572 13.5 0.054 0.056 2.620 2.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.400 2.706 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.211 2.679 apply_single 119 13.6 0.000 0.000 2.211 2.679 mp_waitall_1 137007 16.6 2.204 2.679 2.204 2.679 make_images 4572 14.5 0.274 0.335 2.511 2.634 calculate_first_density_matrix 1 7.0 0.009 0.036 2.562 2.584 acc_transpose_blocks 27432 15.5 0.113 0.116 2.398 2.545 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.430 2.458 density_rs2pw 119 9.7 0.009 0.026 2.272 2.351 cp_dbcsr_syevd 50 12.0 0.007 0.015 2.331 2.332 calculate_dm_sparse 119 9.5 0.000 0.000 2.208 2.258 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.033 2.045 cp_fm_diag_elpa_base 50 14.0 1.988 2.003 2.023 2.031 fft_wrap_pw1pw2 1201 11.6 0.012 0.014 1.950 1.990 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.951 1.957 grid_integrate_task_list 119 12.3 1.829 1.910 1.829 1.910 ot_diis_step 108 11.5 0.012 0.012 1.855 1.855 mp_sum_l 7287 12.8 1.223 1.782 1.223 1.782 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.739 1.752 transfer_rs2pw 487 10.6 0.004 0.005 1.544 1.711 fft3d_ps 1201 13.6 0.577 0.636 1.583 1.650 potential_pw2rs 119 12.3 0.013 0.022 1.630 1.645 fft_wrap_pw1pw2_140 487 12.2 0.074 0.081 1.556 1.596 acc_transpose_blocks_sync 82296 16.5 1.383 1.541 1.383 1.541 qs_energies_init_hamiltonians 11 5.9 0.006 0.007 1.444 1.476 wfi_extrapolate 11 7.9 0.001 0.001 1.472 1.472 grid_collocate_task_list 119 9.7 1.286 1.411 1.286 1.411 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.309 1.327 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.296 1.307 dbcsr_complete_redistribute 329 12.2 0.119 0.162 0.881 1.184 make_images_data 4572 15.5 0.047 0.052 1.006 1.182 cp_fm_upper_to_full 72 14.2 0.831 1.180 0.831 1.180 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.712 1.150 transfer_pw2rs 487 13.2 0.004 0.004 1.077 1.093 jit_kernel_multiply 6 15.8 0.282 1.085 0.282 1.085 hybrid_alltoall_any 4725 16.4 0.066 0.155 0.843 1.063 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.918 0.996 mp_waitany 7280 13.7 0.784 0.969 0.784 0.969 mp_alltoall_z22v 1201 15.6 0.890 0.959 0.890 0.959 mp_alltoall_d11v 2130 13.8 0.820 0.955 0.820 0.955 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.643 0.939 transfer_rs2pw_140 130 11.5 0.108 0.120 0.758 0.935 mp_sum_dm 438 4.9 0.763 0.906 0.763 0.906 acc_transpose_blocks_kernels 27432 16.5 0.273 0.297 0.873 0.905 build_core_hamiltonian_matrix_ 11 4.9 0.014 0.092 0.815 0.888 cp_fm_cholesky_invert 11 10.9 0.871 0.874 0.871 0.874 update_particle_set 20 4.0 0.000 0.000 0.733 0.874 mp_alltoall_i22 627 13.8 0.545 0.854 0.545 0.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.846 0.850 mp_sum_d 4137 12.0 0.636 0.832 0.636 0.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=40.994000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.818182, yerr=2.854603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 628.064256E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.068 32.800 32.800 qs_mol_dyn_low 1 2.0 0.003 0.003 32.397 32.404 qs_forces 11 3.9 0.004 0.005 32.070 32.071 qs_energies 11 4.9 0.002 0.002 30.316 30.321 scf_env_do_scf 11 5.9 0.000 0.001 24.981 24.981 scf_env_do_scf_inner_loop 108 6.5 0.013 0.018 22.076 22.076 velocity_verlet 10 3.0 0.047 0.058 16.543 16.563 dbcsr_multiply_generic 2286 12.5 0.101 0.104 14.569 14.692 qs_scf_new_mos 108 7.5 0.001 0.001 13.771 13.795 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.771 13.795 ot_scf_mini 108 9.5 0.002 0.002 12.982 13.008 multiply_cannon 2286 13.5 0.226 0.236 11.297 11.873 multiply_cannon_loop 2286 14.5 0.643 0.662 10.203 10.482 ot_mini 108 10.5 0.001 0.001 7.466 7.504 multiply_cannon_multrec 9144 15.5 1.807 1.987 6.273 6.464 rebuild_ks_matrix 119 8.3 0.000 0.000 6.163 6.194 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.163 6.193 qs_ot_get_derivative 108 11.5 0.001 0.001 6.092 6.118 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.496 5.524 dbcsr_mm_accdrv_process 12550 15.8 3.627 4.380 4.357 4.462 sum_up_and_integrate 119 10.3 0.001 0.001 3.742 3.746 integrate_v_rspace 119 11.3 0.003 0.003 3.731 3.736 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.692 3.707 calculate_rho_elec 119 8.7 0.060 0.061 3.692 3.706 init_scf_run 11 5.9 0.000 0.001 3.668 3.668 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.667 3.668 qs_ot_get_p 119 10.4 0.001 0.002 3.466 3.502 mp_waitall_1 115863 16.7 2.164 2.959 2.164 2.959 init_scf_loop 11 6.9 0.000 0.000 2.878 2.879 make_m2s 4572 13.5 0.034 0.035 2.266 2.457 make_images 4572 14.5 0.269 0.300 2.176 2.367 calculate_first_density_matrix 1 7.0 0.000 0.000 2.336 2.337 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.259 2.262 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.151 2.169 density_rs2pw 119 9.7 0.003 0.003 2.036 2.116 prepare_preconditioner 11 7.9 0.000 0.000 2.095 2.101 make_preconditioner 11 8.9 0.000 0.000 2.095 2.101 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.961 2.000 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.984 1.985 fft_wrap_pw1pw2 1201 11.6 0.012 0.013 1.943 1.956 grid_integrate_task_list 119 12.3 1.881 1.956 1.881 1.956 calculate_dm_sparse 119 9.5 0.000 0.000 1.880 1.904 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.691 1.693 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.622 1.630 cp_fm_diag_elpa_base 50 14.0 1.592 1.612 1.620 1.628 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.589 1.604 fft3d_ps 1201 13.6 0.642 0.656 1.556 1.573 acc_transpose_blocks 9144 15.5 0.040 0.042 1.531 1.572 jit_kernel_multiply 8 15.5 0.691 1.564 0.691 1.564 fft_wrap_pw1pw2_140 487 12.2 0.094 0.095 1.530 1.542 potential_pw2rs 119 12.3 0.010 0.010 1.506 1.511 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.433 1.442 qs_energies_init_hamiltonians 11 5.9 0.068 0.106 1.417 1.420 grid_collocate_task_list 119 9.7 1.308 1.370 1.308 1.370 ot_diis_step 108 11.5 0.013 0.013 1.346 1.346 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.279 1.316 apply_single 119 13.6 0.000 0.000 1.278 1.316 make_images_data 4572 15.5 0.042 0.046 1.010 1.291 hybrid_alltoall_any 4725 16.4 0.065 0.176 0.955 1.285 wfi_extrapolate 11 7.9 0.001 0.001 1.273 1.273 transfer_rs2pw 487 10.6 0.004 0.005 1.125 1.227 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.198 1.206 cp_fm_cholesky_invert 11 10.9 1.124 1.128 1.124 1.128 mp_alltoall_d11v 2130 13.8 0.951 1.039 0.951 1.039 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.935 0.949 transfer_pw2rs 487 13.2 0.003 0.003 0.937 0.942 mp_allgather_i34 2286 14.5 0.406 0.942 0.406 0.942 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.876 0.928 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.394 0.841 mp_alltoall_z22v 1201 15.6 0.789 0.833 0.789 0.833 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.742 0.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.791 0.793 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.758 0.771 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.513 0.760 acc_transpose_blocks_sync 27432 16.5 0.717 0.748 0.717 0.748 mp_sum_l 7287 12.8 0.588 0.742 0.588 0.742 dbcsr_complete_redistribute 329 12.2 0.168 0.229 0.697 0.733 make_images_sizes 4572 15.5 0.005 0.005 0.474 0.676 mp_alltoall_i44 4572 16.5 0.469 0.671 0.469 0.671 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.800000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=593.454545, yerr=7.831912 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 782.290944E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.041 46.610 46.610 qs_mol_dyn_low 1 2.0 0.003 0.003 45.991 45.999 qs_forces 11 3.9 0.003 0.004 45.892 45.893 qs_energies 11 4.9 0.005 0.012 43.832 43.835 scf_env_do_scf 11 5.9 0.001 0.001 37.024 37.024 scf_env_do_scf_inner_loop 108 6.5 0.005 0.009 28.673 28.674 velocity_verlet 10 3.0 0.028 0.029 25.357 25.365 dbcsr_multiply_generic 2286 12.5 0.116 0.117 20.801 20.886 qs_scf_new_mos 108 7.5 0.001 0.001 18.795 18.867 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.794 18.867 ot_scf_mini 108 9.5 0.002 0.002 17.590 17.672 multiply_cannon 2286 13.5 0.296 0.303 16.260 17.073 multiply_cannon_loop 2286 14.5 0.880 0.914 14.823 15.528 ot_mini 108 10.5 0.001 0.001 10.737 10.826 multiply_cannon_multrec 9144 15.5 3.417 4.758 9.243 9.535 qs_ot_get_derivative 108 11.5 0.001 0.001 8.667 8.747 init_scf_loop 11 6.9 0.000 0.000 8.322 8.323 rebuild_ks_matrix 119 8.3 0.000 0.000 7.460 7.575 qs_ks_build_kohn_sham_matrix 119 9.3 0.020 0.034 7.460 7.575 prepare_preconditioner 11 7.9 0.000 0.000 7.300 7.317 make_preconditioner 11 8.9 0.000 0.000 7.300 7.317 dbcsr_mm_accdrv_process 12550 15.8 4.864 7.164 5.689 7.197 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.802 7.190 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.714 6.821 cp_fm_upper_to_full 72 14.2 3.248 4.687 3.248 4.687 init_scf_run 11 5.9 0.000 0.001 4.646 4.646 scf_env_initial_rho_setup 11 6.9 0.002 0.002 4.646 4.646 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.269 4.293 calculate_rho_elec 119 8.7 0.118 0.121 4.269 4.292 mp_waitall_1 94719 16.7 3.114 4.174 3.114 4.174 sum_up_and_integrate 119 10.3 0.001 0.001 4.012 4.017 integrate_v_rspace 119 11.3 0.003 0.003 4.002 4.007 qs_ot_get_p 119 10.4 0.002 0.003 3.760 3.870 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.420 3.836 make_m2s 4572 13.5 0.037 0.038 2.929 3.224 make_images 4572 14.5 0.352 0.383 2.808 3.103 calculate_first_density_matrix 1 7.0 0.000 0.000 3.081 3.087 dbcsr_complete_redistribute 329 12.2 0.297 0.304 2.108 2.963 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.770 2.628 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.260 2.604 apply_single 119 13.6 0.000 0.000 2.260 2.604 fft_wrap_pw1pw2 1201 11.6 0.014 0.014 2.560 2.573 calculate_dm_sparse 119 9.5 0.000 0.000 2.528 2.547 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.636 2.510 mp_alltoall_i22 627 13.8 1.564 2.470 1.564 2.470 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.493 2.345 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.247 2.288 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.264 2.269 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.264 2.264 density_rs2pw 119 9.7 0.003 0.003 2.222 2.241 grid_integrate_task_list 119 12.3 2.143 2.173 2.143 2.173 acc_transpose_blocks 9144 15.5 0.044 0.044 2.088 2.147 fft_wrap_pw1pw2_140 487 12.2 0.179 0.180 2.130 2.146 ot_diis_step 108 11.5 0.014 0.014 2.046 2.046 fft3d_ps 1201 13.6 0.860 0.880 2.012 2.022 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.904 1.958 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.881 1.881 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.368 1.858 make_images_data 4572 15.5 0.045 0.048 1.367 1.829 qs_energies_init_hamiltonians 11 5.9 0.022 0.023 1.827 1.828 mp_sum_l 7287 12.8 1.061 1.673 1.061 1.673 cp_fm_cholesky_invert 11 10.9 1.660 1.664 1.660 1.664 grid_collocate_task_list 119 9.7 1.565 1.583 1.565 1.583 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.551 1.551 cp_fm_diag_elpa_base 50 14.0 1.404 1.461 1.550 1.550 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.491 1.501 wfi_extrapolate 11 7.9 0.001 0.001 1.494 1.494 potential_pw2rs 119 12.3 0.013 0.014 1.431 1.435 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.392 1.410 mp_alltoall_d11v 2130 13.8 1.283 1.331 1.283 1.331 acc_transpose_blocks_sync 27432 16.5 1.131 1.187 1.131 1.187 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.131 1.144 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.003 1.061 jit_kernel_multiply 5 15.4 0.796 1.057 0.796 1.057 mp_alltoall_z22v 1201 15.6 1.017 1.042 1.017 1.042 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.955 0.964 transfer_rs2pw 487 10.6 0.005 0.005 0.911 0.959 mp_allgather_i34 2286 14.5 0.393 0.949 0.393 0.949 qs_create_task_list 11 7.9 0.000 0.000 0.934 0.944 generate_qs_task_list 11 8.9 0.367 0.386 0.934 0.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=46.610000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=733.545455, yerr=15.263525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 496.922624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.094 87.347 87.352 qs_mol_dyn_low 1 2.0 0.011 0.040 86.242 86.252 qs_forces 11 3.9 0.003 0.004 86.000 86.004 qs_energies 11 4.9 0.004 0.006 83.058 83.075 scf_env_do_scf 11 5.9 0.000 0.001 73.014 73.016 scf_env_do_scf_inner_loop 99 6.5 0.006 0.021 66.838 66.839 dbcsr_multiply_generic 2055 12.4 0.106 0.108 52.121 52.389 qs_scf_new_mos 99 7.5 0.000 0.001 49.162 49.286 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.161 49.285 ot_scf_mini 99 9.5 0.002 0.002 46.768 46.868 velocity_verlet 10 3.0 0.023 0.075 44.917 44.919 multiply_cannon 2055 13.4 0.186 0.190 43.140 44.093 multiply_cannon_loop 2055 14.4 1.770 1.815 42.032 42.923 ot_mini 99 10.5 0.001 0.001 27.213 27.331 qs_ot_get_derivative 99 11.5 0.001 0.001 20.421 20.553 multiply_cannon_multrec 49320 15.4 11.312 12.222 17.378 18.447 rebuild_ks_matrix 110 8.3 0.000 0.000 14.920 15.066 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.068 14.920 15.066 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.087 13.219 mp_waitall_1 220248 16.4 11.591 12.403 11.591 12.403 qs_ot_get_p 110 10.4 0.001 0.001 11.037 11.211 multiply_cannon_sync_h2d 49320 15.4 9.535 10.137 9.535 10.137 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.958 8.503 multiply_cannon_metrocomm3 49320 15.4 0.082 0.086 6.699 7.906 init_scf_run 11 5.9 0.000 0.001 7.733 7.734 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.733 7.734 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.263 7.722 apply_single 110 13.6 0.000 0.000 7.263 7.722 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.517 7.542 sum_up_and_integrate 110 10.3 0.002 0.003 7.280 7.316 integrate_v_rspace 110 11.3 0.003 0.003 7.255 7.296 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.639 6.772 calculate_rho_elec 110 8.6 0.020 0.024 6.638 6.771 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.635 6.635 ot_diis_step 99 11.5 0.006 0.006 6.593 6.594 dbcsr_mm_accdrv_process 87628 16.1 3.054 3.181 5.936 6.221 init_scf_loop 11 6.9 0.000 0.000 6.144 6.145 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.941 5.962 cp_fm_diag_elpa_base 48 14.0 5.923 5.946 5.939 5.960 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.455 5.541 mp_sum_l 6594 12.7 4.370 5.269 4.370 5.269 make_m2s 4110 13.4 0.061 0.065 4.339 4.442 make_images 4110 14.4 0.178 0.190 4.241 4.349 wfi_extrapolate 11 7.9 0.001 0.001 4.145 4.145 calculate_dm_sparse 110 9.5 0.001 0.001 3.938 4.031 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 4.012 4.016 prepare_preconditioner 11 7.9 0.000 0.000 3.895 3.914 make_preconditioner 11 8.9 0.000 0.000 3.895 3.914 density_rs2pw 110 9.6 0.003 0.004 3.672 3.822 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.630 3.764 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.684 3.728 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.555 3.601 calculate_first_density_matrix 1 7.0 0.000 0.001 3.483 3.488 grid_integrate_task_list 110 12.3 3.258 3.470 3.258 3.470 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.239 3.292 fft_wrap_pw1pw2 1111 11.6 0.014 0.017 2.986 3.097 potential_pw2rs 110 12.3 0.006 0.006 2.905 2.936 jit_kernel_multiply 13 15.9 2.603 2.878 2.603 2.878 acc_transpose_blocks 49320 15.4 0.202 0.212 2.715 2.876 fft3d_ps 1111 13.6 0.731 0.821 2.529 2.625 fft_wrap_pw1pw2_140 451 12.1 0.092 0.101 2.417 2.534 mp_alltoall_d11v 2046 13.8 2.098 2.506 2.098 2.506 transfer_rs2pw 451 10.6 0.006 0.006 2.283 2.443 grid_collocate_task_list 110 9.6 2.160 2.264 2.160 2.264 cp_fm_cholesky_invert 11 10.9 2.125 2.130 2.125 2.130 transfer_pw2rs 451 13.1 0.006 0.006 2.077 2.117 mp_waitany 14300 13.8 1.871 2.078 1.871 2.078 make_images_data 4110 15.4 0.043 0.047 1.879 2.039 mp_sum_d 3893 11.9 1.402 2.011 1.402 2.011 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.973 1.987 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.615 1.864 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.831 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=87.352000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.090909, yerr=2.314168 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.706944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2933536. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.078 74.213 74.214 qs_mol_dyn_low 1 2.0 0.011 0.071 73.682 73.693 qs_forces 11 3.9 0.008 0.024 73.408 73.422 qs_energies 11 4.9 0.002 0.003 70.003 70.025 scf_env_do_scf 11 5.9 0.000 0.001 60.840 60.843 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 52.062 52.063 dbcsr_multiply_generic 2055 12.4 0.119 0.132 39.227 39.543 velocity_verlet 10 3.0 0.028 0.207 38.553 38.561 qs_scf_new_mos 99 7.5 0.001 0.001 35.642 35.779 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.641 35.779 ot_scf_mini 99 9.5 0.003 0.003 33.916 34.050 multiply_cannon 2055 13.4 0.226 0.245 31.863 32.783 multiply_cannon_loop 2055 14.4 1.160 1.184 30.363 31.592 ot_mini 99 10.5 0.001 0.001 19.345 19.485 multiply_cannon_multrec 24660 15.4 6.956 8.664 14.004 15.885 rebuild_ks_matrix 110 8.3 0.000 0.000 13.963 14.098 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.963 14.097 qs_ot_get_derivative 99 11.5 0.001 0.001 13.504 13.635 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.295 12.415 mp_waitall_1 176588 16.5 8.207 11.451 8.207 11.451 init_scf_loop 11 6.9 0.000 0.000 8.731 8.732 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.409 8.327 qs_ot_get_p 110 10.4 0.001 0.002 7.759 7.926 multiply_cannon_sync_h2d 24660 15.4 6.363 7.448 6.363 7.448 dbcsr_mm_accdrv_process 52282 16.1 5.522 6.400 6.881 7.226 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.462 7.063 apply_single 110 13.6 0.000 0.001 6.462 7.063 sum_up_and_integrate 110 10.3 0.001 0.003 6.647 6.658 prepare_preconditioner 11 7.9 0.000 0.000 6.625 6.650 make_preconditioner 11 8.9 0.000 0.000 6.625 6.650 integrate_v_rspace 110 11.3 0.002 0.003 6.620 6.632 init_scf_run 11 5.9 0.000 0.001 6.593 6.593 scf_env_initial_rho_setup 11 6.9 0.001 0.003 6.592 6.593 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.190 6.358 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.130 6.142 calculate_rho_elec 110 8.6 0.039 0.047 6.129 6.141 ot_diis_step 99 11.5 0.010 0.011 5.766 5.766 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.001 5.728 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.552 5.578 make_m2s 4110 13.4 0.057 0.059 4.594 5.120 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.022 5.023 make_images 4110 14.4 0.406 0.463 4.480 5.002 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.291 4.307 cp_fm_diag_elpa_base 48 14.0 4.237 4.252 4.288 4.304 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.654 3.723 wfi_extrapolate 11 7.9 0.001 0.001 3.678 3.678 density_rs2pw 110 9.6 0.004 0.004 3.414 3.596 fft_wrap_pw1pw2 1111 11.6 0.014 0.016 3.322 3.481 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.478 3.481 grid_integrate_task_list 110 12.3 3.131 3.360 3.131 3.360 cp_fm_cholesky_invert 11 10.9 3.254 3.263 3.254 3.263 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.046 3.095 calculate_dm_sparse 110 9.5 0.001 0.001 3.042 3.068 mp_sum_l 6594 12.7 2.177 2.966 2.177 2.966 make_images_data 4110 15.4 0.048 0.052 2.384 2.911 hybrid_alltoall_any 4261 16.3 0.105 0.452 2.066 2.860 fft3d_ps 1111 13.6 1.072 1.281 2.703 2.855 calculate_first_density_matrix 1 7.0 0.000 0.000 2.813 2.816 fft_wrap_pw1pw2_140 451 12.1 0.112 0.120 2.566 2.724 potential_pw2rs 110 12.3 0.008 0.008 2.548 2.572 grid_collocate_task_list 110 9.6 2.166 2.327 2.166 2.327 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.117 2.142 mp_alltoall_d11v 2046 13.8 1.817 2.126 1.817 2.126 transfer_rs2pw 451 10.6 0.007 0.007 1.886 2.107 acc_transpose_blocks 24660 15.4 0.112 0.117 2.053 2.092 qs_energies_init_hamiltonians 11 5.9 0.003 0.016 2.014 2.034 jit_kernel_multiply 10 16.3 0.992 1.907 0.992 1.907 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.833 1.850 mp_allgather_i34 2055 14.4 0.806 1.806 0.806 1.806 cp_fm_cholesky_decompose 22 10.9 1.723 1.731 1.723 1.731 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.766 1.623 dbcsr_complete_redistribute 325 12.2 0.242 0.350 1.327 1.613 transfer_pw2rs 451 13.1 0.005 0.006 1.566 1.592 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.562 1.579 mp_waitany 10164 13.8 1.291 1.536 1.291 1.536 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 1.388 1.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.214000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.363636, yerr=6.168435 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 665.554944E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.128 0.159 68.432 68.435 qs_mol_dyn_low 1 2.0 0.021 0.144 67.566 67.575 qs_forces 11 3.9 0.019 0.026 67.126 67.128 qs_energies 11 4.9 0.028 0.040 63.810 63.829 scf_env_do_scf 11 5.9 0.001 0.002 54.624 54.625 scf_env_do_scf_inner_loop 99 6.5 0.007 0.013 44.693 44.695 velocity_verlet 10 3.0 0.014 0.105 35.464 35.492 dbcsr_multiply_generic 2055 12.4 0.115 0.118 30.911 31.360 qs_scf_new_mos 99 7.5 0.001 0.001 29.673 29.769 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.672 29.768 ot_scf_mini 99 9.5 0.002 0.003 28.349 28.460 multiply_cannon 2055 13.4 0.212 0.222 23.580 24.977 multiply_cannon_loop 2055 14.4 0.813 0.848 22.137 24.011 ot_mini 99 10.5 0.001 0.001 15.166 15.284 rebuild_ks_matrix 110 8.3 0.000 0.000 12.509 12.680 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.508 12.680 mp_waitall_1 139946 16.5 7.741 12.229 7.741 12.229 multiply_cannon_multrec 16440 15.4 3.769 5.346 9.993 11.512 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.020 11.172 qs_ot_get_derivative 99 11.5 0.001 0.001 10.631 10.744 init_scf_loop 11 6.9 0.000 0.001 9.882 9.883 multiply_cannon_metrocomm3 16440 15.4 0.045 0.046 4.530 8.862 qs_ot_get_p 110 10.4 0.001 0.001 8.120 8.269 prepare_preconditioner 11 7.9 0.000 0.000 8.058 8.074 make_preconditioner 11 8.9 0.000 0.001 8.058 8.074 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.330 7.690 sum_up_and_integrate 110 10.3 0.001 0.002 6.575 6.590 integrate_v_rspace 110 11.3 0.003 0.003 6.549 6.564 init_scf_run 11 5.9 0.000 0.001 6.539 6.539 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.539 6.539 dbcsr_mm_accdrv_process 34862 16.1 5.364 5.702 6.072 6.285 qs_ot_p2m_diag 48 11.0 0.042 0.044 6.042 6.063 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.862 5.873 calculate_rho_elec 110 8.6 0.058 0.058 5.862 5.872 cp_dbcsr_syevd 48 12.0 0.014 0.091 5.606 5.607 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.986 5.556 apply_single 110 13.6 0.000 0.000 4.986 5.556 make_m2s 4110 13.4 0.049 0.050 4.587 5.083 make_images 4110 14.4 0.396 0.516 4.469 4.966 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.775 4.788 cp_fm_diag_elpa_base 48 14.0 4.672 4.709 4.747 4.756 ot_diis_step 99 11.5 0.011 0.011 4.468 4.468 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.622 4.270 multiply_cannon_sync_h2d 16440 15.4 3.270 3.920 3.270 3.920 cp_fm_cholesky_invert 11 10.9 3.447 3.457 3.447 3.457 grid_integrate_task_list 110 12.3 3.176 3.432 3.176 3.432 density_rs2pw 110 9.6 0.004 0.007 3.119 3.335 calculate_first_density_matrix 1 7.0 0.001 0.010 3.249 3.252 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.227 3.231 wfi_extrapolate 11 7.9 0.001 0.001 3.184 3.184 fft_wrap_pw1pw2 1111 11.6 0.013 0.017 3.129 3.148 make_images_data 4110 15.4 0.045 0.050 2.436 3.068 mp_sum_l 6594 12.7 2.158 3.023 2.158 3.023 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.955 3.014 hybrid_alltoall_any 4261 16.3 0.108 0.379 2.140 2.935 calculate_dm_sparse 110 9.5 0.001 0.001 2.859 2.896 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.782 2.842 fft_wrap_pw1pw2_140 451 12.1 0.127 0.133 2.527 2.549 fft3d_ps 1111 13.6 1.085 1.115 2.445 2.461 potential_pw2rs 110 12.3 0.010 0.016 2.432 2.454 grid_collocate_task_list 110 9.6 2.216 2.453 2.216 2.453 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.864 2.446 mp_irecv_dv 48980 15.7 0.789 2.309 0.789 2.309 mp_alltoall_d11v 2046 13.8 1.968 2.221 1.968 2.221 dbcsr_complete_redistribute 325 12.2 0.334 0.366 1.717 2.195 qs_energies_init_hamiltonians 11 5.9 0.005 0.021 2.023 2.041 cp_fm_cholesky_decompose 22 10.9 2.018 2.040 2.018 2.040 acc_transpose_blocks 16440 15.4 0.075 0.077 1.837 2.026 transfer_rs2pw 451 10.6 0.005 0.006 1.636 1.823 cp_fm_upper_to_full 70 14.2 1.415 1.792 1.415 1.792 mp_allgather_i34 2055 14.4 0.755 1.777 0.755 1.777 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.734 1.748 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.709 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.159 1.624 transfer_pw2rs 451 13.1 0.005 0.005 1.544 1.563 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.008 1.373 1.515 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.481 1.496 mp_waitany 17072 13.8 1.233 1.437 1.233 1.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=68.435000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=629.727273, yerr=9.026133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 735.592448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.058 72.311 72.312 qs_mol_dyn_low 1 2.0 0.003 0.003 71.698 71.723 qs_forces 11 3.9 0.004 0.005 71.617 71.618 qs_energies 11 4.9 0.001 0.002 68.181 68.186 scf_env_do_scf 11 5.9 0.000 0.001 57.936 57.938 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.187 45.188 velocity_verlet 10 3.0 0.015 0.017 38.833 38.836 dbcsr_multiply_generic 2055 12.4 0.141 0.145 33.084 33.304 qs_scf_new_mos 99 7.5 0.001 0.001 30.347 30.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.347 30.446 ot_scf_mini 99 9.5 0.002 0.003 28.629 28.738 multiply_cannon 2055 13.4 0.236 0.259 25.328 26.666 multiply_cannon_loop 2055 14.4 1.388 1.479 23.745 24.383 ot_mini 99 10.5 0.001 0.001 15.984 16.106 multiply_cannon_multrec 24660 15.4 4.112 6.920 14.113 15.549 init_scf_loop 11 6.9 0.000 0.000 12.703 12.704 rebuild_ks_matrix 110 8.3 0.000 0.000 12.219 12.321 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.218 12.321 qs_ot_get_derivative 99 11.5 0.001 0.001 11.770 11.881 prepare_preconditioner 11 7.9 0.000 0.000 10.912 10.931 make_preconditioner 11 8.9 0.000 0.000 10.912 10.931 dbcsr_mm_accdrv_process 52304 16.0 8.444 10.248 9.845 10.915 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.805 10.894 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.114 10.598 qs_ot_get_p 110 10.4 0.001 0.002 7.396 7.546 init_scf_run 11 5.9 0.000 0.001 7.444 7.445 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.444 7.445 mp_waitall_1 121746 16.5 5.005 6.974 5.005 6.974 sum_up_and_integrate 110 10.3 0.001 0.002 6.373 6.387 integrate_v_rspace 110 11.3 0.003 0.003 6.347 6.360 make_m2s 4110 13.4 0.060 0.061 5.842 6.162 make_images 4110 14.4 0.572 0.694 5.699 6.017 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.855 5.928 calculate_rho_elec 110 8.6 0.077 0.081 5.855 5.928 qs_ot_p2m_diag 48 11.0 0.055 0.064 5.274 5.292 cp_fm_upper_to_full 70 14.2 3.395 4.928 3.395 4.928 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.674 4.675 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.169 4.298 apply_single 110 13.6 0.000 0.000 4.169 4.298 calculate_first_density_matrix 1 7.0 0.000 0.000 4.234 4.241 ot_diis_step 99 11.5 0.011 0.012 4.167 4.167 dbcsr_complete_redistribute 325 12.2 0.437 0.582 2.922 4.070 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.915 4.060 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.950 3.957 cp_fm_diag_elpa_base 48 14.0 3.783 3.849 3.947 3.955 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.751 3.753 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.576 3.636 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.661 3.494 grid_integrate_task_list 110 12.3 3.244 3.488 3.244 3.488 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.334 3.474 cp_fm_cholesky_invert 11 10.9 3.400 3.411 3.400 3.411 calculate_dm_sparse 110 9.5 0.001 0.001 3.355 3.384 fft_wrap_pw1pw2 1111 11.6 0.013 0.016 3.232 3.271 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.229 3.271 make_images_data 4110 15.4 0.048 0.052 2.900 3.259 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.397 3.242 density_rs2pw 110 9.6 0.004 0.004 2.999 3.241 wfi_extrapolate 11 7.9 0.001 0.001 3.108 3.108 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.789 2.909 mp_alltoall_i22 605 13.7 1.696 2.881 1.696 2.881 acc_transpose_blocks 24660 15.4 0.106 0.108 2.578 2.742 fft_wrap_pw1pw2_140 451 12.1 0.146 0.148 2.657 2.701 multiply_cannon_sync_h2d 24660 15.4 2.390 2.553 2.390 2.553 fft3d_ps 1111 13.6 1.123 1.167 2.502 2.531 grid_collocate_task_list 110 9.6 2.262 2.489 2.262 2.489 qs_energies_init_hamiltonians 11 5.9 0.005 0.015 2.271 2.274 potential_pw2rs 110 12.3 0.012 0.013 2.214 2.227 mp_alltoall_d11v 2046 13.8 1.835 2.183 1.835 2.183 jit_kernel_multiply 8 15.7 1.056 2.151 1.056 2.151 cp_fm_cholesky_decompose 22 10.9 1.927 1.966 1.927 1.966 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.881 1.913 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.890 1.911 mp_sum_l 6594 12.7 1.220 1.802 1.220 1.802 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.710 mp_allgather_i34 2055 14.4 0.668 1.705 0.668 1.705 transfer_rs2pw 451 10.6 0.005 0.006 1.433 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.657 1.668 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.850 1.621 mp_irecv_dv 62702 16.1 0.746 1.536 0.746 1.536 acc_transpose_blocks_sync 73980 16.4 1.367 1.506 1.367 1.506 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.312000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.272727, yerr=7.398905 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 870.273024E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.061 57.582 57.583 qs_mol_dyn_low 1 2.0 0.003 0.003 56.860 56.879 qs_forces 11 3.9 0.033 0.059 56.626 56.628 qs_energies 11 4.9 0.045 0.048 52.959 52.976 scf_env_do_scf 11 5.9 0.000 0.001 43.516 43.517 scf_env_do_scf_inner_loop 99 6.5 0.018 0.030 35.646 35.647 velocity_verlet 10 3.0 0.019 0.020 31.350 31.362 dbcsr_multiply_generic 2055 12.4 0.132 0.135 24.452 24.601 qs_scf_new_mos 99 7.5 0.001 0.001 21.313 21.394 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.312 21.393 ot_scf_mini 99 9.5 0.002 0.002 20.045 20.119 multiply_cannon 2055 13.4 0.238 0.246 18.720 19.990 multiply_cannon_loop 2055 14.4 0.603 0.622 17.437 17.707 rebuild_ks_matrix 110 8.3 0.000 0.000 11.592 11.664 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.592 11.664 ot_mini 99 10.5 0.001 0.001 11.133 11.199 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.315 10.381 multiply_cannon_multrec 8220 15.4 3.322 4.660 8.294 9.524 init_scf_loop 11 6.9 0.000 0.000 7.822 7.823 mp_waitall_1 103326 16.6 6.071 7.816 6.071 7.816 qs_ot_get_derivative 99 11.5 0.001 0.001 7.369 7.442 init_scf_run 11 5.9 0.000 0.001 6.236 6.236 scf_env_initial_rho_setup 11 6.9 0.010 0.015 6.235 6.235 prepare_preconditioner 11 7.9 0.000 0.000 6.137 6.147 make_preconditioner 11 8.9 0.000 0.000 6.137 6.147 sum_up_and_integrate 110 10.3 0.001 0.002 6.103 6.116 integrate_v_rspace 110 11.3 0.003 0.003 6.076 6.089 dbcsr_mm_accdrv_process 17442 15.9 3.400 4.515 4.830 5.995 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.706 5.778 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.738 5.751 calculate_rho_elec 110 8.6 0.115 0.116 5.737 5.751 qs_ot_get_p 110 10.4 0.013 0.025 4.936 5.008 make_m2s 4110 13.4 0.037 0.038 4.307 4.556 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.090 4.546 make_images 4110 14.4 0.646 0.707 4.176 4.422 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.791 3.841 apply_single 110 13.6 0.000 0.000 3.791 3.841 ot_diis_step 99 11.5 0.012 0.013 3.741 3.742 grid_integrate_task_list 110 12.3 3.379 3.500 3.379 3.500 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.484 3.488 calculate_first_density_matrix 1 7.0 0.000 0.000 3.434 3.437 fft_wrap_pw1pw2 1111 11.6 0.014 0.015 3.377 3.383 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.135 3.136 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.962 2.965 cp_fm_cholesky_invert 11 10.9 2.943 2.947 2.943 2.947 fft_wrap_pw1pw2_140 451 12.1 0.192 0.195 2.901 2.910 density_rs2pw 110 9.6 0.003 0.004 2.727 2.839 calculate_dm_sparse 110 9.5 0.001 0.001 2.790 2.831 qs_energies_init_hamiltonians 11 5.9 0.003 0.006 2.692 2.707 wfi_extrapolate 11 7.9 0.001 0.001 2.697 2.698 hybrid_alltoall_any 4261 16.3 0.201 0.841 2.304 2.690 make_images_data 4110 15.4 0.041 0.047 2.299 2.626 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.549 2.576 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.545 2.553 cp_fm_diag_elpa_base 48 14.0 2.489 2.515 2.543 2.551 fft3d_ps 1111 13.6 1.274 1.285 2.534 2.542 multiply_cannon_sync_h2d 8220 15.4 2.349 2.504 2.349 2.504 grid_collocate_task_list 110 9.6 2.364 2.443 2.364 2.443 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.117 2.166 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.002 2.033 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.791 1.994 mp_alltoall_d11v 2046 13.8 1.773 1.991 1.773 1.991 potential_pw2rs 110 12.3 0.015 0.015 1.871 1.879 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.727 1.740 cp_fm_cholesky_decompose 22 10.9 1.713 1.728 1.713 1.728 jit_kernel_multiply 7 15.6 1.115 1.645 1.115 1.645 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.511 1.639 acc_transpose_blocks 8220 15.4 0.038 0.040 1.561 1.623 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.606 1.613 dbcsr_complete_redistribute 325 12.2 0.591 0.628 1.525 1.613 mp_allgather_i34 2055 14.4 0.452 1.557 0.452 1.557 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.477 1.392 mp_irecv_dv 24056 15.7 0.450 1.345 0.450 1.345 qs_create_task_list 11 7.9 0.000 0.000 1.219 1.317 generate_qs_task_list 11 8.9 0.376 0.445 1.218 1.316 transfer_rs2pw 451 10.6 0.005 0.005 1.150 1.292 mp_waitany 9240 13.8 1.085 1.247 1.085 1.247 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.217 1.240 multiply_cannon_metrocomm1 8220 15.4 0.022 0.022 0.850 1.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.583000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=820.000000, yerr=15.521246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/14/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 1.612391E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1464624 0.0% 0.0% 100.0% average stack size 0.0 0.0 672.0 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 1.409675E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.039 95.940 95.940 qs_mol_dyn_low 1 2.0 0.003 0.003 95.300 95.311 qs_forces 11 3.9 0.025 0.046 94.991 94.992 qs_energies 11 4.9 0.021 0.023 90.790 90.803 scf_env_do_scf 11 5.9 0.000 0.001 79.297 79.297 velocity_verlet 10 3.0 0.008 0.008 59.364 59.399 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.382 48.385 dbcsr_multiply_generic 2055 12.4 0.124 0.128 32.166 32.234 init_scf_loop 11 6.9 0.000 0.000 30.840 30.845 qs_scf_new_mos 99 7.5 0.001 0.001 30.248 30.266 qs_scf_loop_do_ot 99 8.5 0.001 0.001 30.247 30.265 prepare_preconditioner 11 7.9 0.000 0.000 28.616 28.638 make_preconditioner 11 8.9 0.000 0.000 28.616 28.638 ot_scf_mini 99 9.5 0.002 0.002 28.361 28.377 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.457 28.048 multiply_cannon 2055 13.4 0.330 0.349 24.269 25.374 multiply_cannon_loop 2055 14.4 0.825 0.846 22.294 22.987 cp_fm_upper_to_full 70 14.2 13.086 18.935 13.086 18.935 ot_mini 99 10.5 0.001 0.001 15.666 15.692 rebuild_ks_matrix 110 8.3 0.000 0.000 14.224 14.241 qs_ks_build_kohn_sham_matrix 110 9.3 0.040 0.051 14.224 14.240 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.915 12.934 dbcsr_complete_redistribute 325 12.2 1.030 1.048 7.876 11.338 qs_ot_get_derivative 99 11.5 0.001 0.001 10.897 10.910 multiply_cannon_multrec 8220 15.4 4.251 4.415 10.396 10.599 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.804 10.268 mp_waitall_1 84994 16.7 8.816 9.705 8.816 9.705 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.144 9.566 mp_alltoall_i22 605 13.7 5.770 9.238 5.770 9.238 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.462 7.502 calculate_rho_elec 110 8.6 0.227 0.227 7.461 7.501 qs_ot_get_p 110 10.4 0.004 0.004 7.229 7.250 sum_up_and_integrate 110 10.3 0.002 0.002 7.169 7.184 integrate_v_rspace 110 11.3 0.003 0.003 7.141 7.156 init_scf_run 11 5.9 0.000 0.001 7.038 7.038 scf_env_initial_rho_setup 11 6.9 0.005 0.005 7.038 7.038 cp_fm_cholesky_invert 11 10.9 6.377 6.380 6.377 6.380 dbcsr_mm_accdrv_process 11614 15.7 4.123 4.350 5.995 6.267 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.646 6.241 make_m2s 4110 13.4 0.043 0.044 5.793 6.225 make_images 4110 14.4 0.880 0.934 5.602 6.035 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.913 5.335 apply_single 110 13.6 0.000 0.000 4.913 5.335 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.253 5.257 fft_wrap_pw1pw2 1111 11.6 0.016 0.016 4.974 4.988 ot_diis_step 99 11.5 0.015 0.015 4.723 4.723 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.684 4.685 fft_wrap_pw1pw2_140 451 12.1 0.364 0.365 4.252 4.270 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.959 3.959 cp_fm_diag_elpa_base 48 14.0 3.396 3.605 3.956 3.956 density_rs2pw 110 9.6 0.004 0.004 3.846 3.870 fft3d_ps 1111 13.6 1.841 1.873 3.833 3.849 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.340 3.836 grid_integrate_task_list 110 12.3 3.695 3.776 3.695 3.776 qs_energies_init_hamiltonians 11 5.9 0.022 0.045 3.754 3.766 hybrid_alltoall_any 4261 16.3 0.262 0.561 3.027 3.698 make_images_data 4110 15.4 0.045 0.048 3.004 3.651 calculate_dm_sparse 110 9.5 0.001 0.001 3.541 3.582 wfi_extrapolate 11 7.9 0.001 0.001 3.521 3.522 calculate_first_density_matrix 1 7.0 0.000 0.000 3.391 3.393 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.347 3.366 multiply_cannon_sync_h2d 8220 15.4 3.127 3.165 3.127 3.165 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.056 3.064 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.775 2.785 grid_collocate_task_list 110 9.6 2.682 2.717 2.682 2.717 potential_pw2rs 110 12.3 0.021 0.021 2.589 2.597 cp_fm_cholesky_decompose 22 10.9 2.496 2.521 2.496 2.521 mp_alltoall_d11v 2046 13.8 2.311 2.383 2.311 2.383 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.220 2.283 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.190 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.967 1.985 acc_transpose_blocks 8220 15.4 0.040 0.040 1.917 1.973 qs_create_task_list 11 7.9 0.000 0.000 1.899 1.943 generate_qs_task_list 11 8.9 0.730 0.781 1.899 1.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.919 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=95.940000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1273.909091, yerr=58.232222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.325824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57761. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2488523. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.111 0.247 220.202 220.212 qs_mol_dyn_low 1 2.0 0.057 0.158 218.771 218.783 qs_forces 11 3.9 0.006 0.008 218.579 218.639 qs_energies 11 4.9 0.017 0.104 212.859 212.929 scf_env_do_scf 11 5.9 0.004 0.026 194.993 194.997 scf_env_do_scf_inner_loop 117 6.6 0.032 0.179 173.119 173.123 qs_scf_new_mos 117 7.6 0.001 0.001 132.020 132.320 qs_scf_loop_do_ot 117 8.6 0.002 0.007 132.020 132.319 dbcsr_multiply_generic 2507 12.6 0.189 0.209 128.576 129.641 velocity_verlet 10 3.0 0.022 0.058 128.498 128.510 ot_scf_mini 117 9.6 0.003 0.005 125.289 125.596 multiply_cannon 2507 13.6 0.239 0.247 103.002 104.693 multiply_cannon_loop 2507 14.6 2.402 2.460 100.644 102.325 ot_mini 117 10.6 0.001 0.001 68.603 68.918 qs_ot_get_derivative 117 11.6 0.001 0.001 43.481 43.767 multiply_cannon_multrec 60168 15.6 31.798 33.774 41.838 43.644 rebuild_ks_matrix 128 8.3 0.001 0.001 34.666 35.008 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.666 35.007 qs_ot_get_p 128 10.4 0.001 0.001 34.132 34.471 mp_waitall_1 267128 16.5 30.827 34.118 30.827 34.118 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.229 31.533 multiply_cannon_sync_h2d 60168 15.6 26.284 28.056 26.284 28.056 qs_ot_p2m_diag 83 11.4 0.079 0.091 27.083 27.133 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.522 26.069 apply_single 128 13.6 0.001 0.001 24.521 26.068 ot_diis_step 117 11.6 0.008 0.008 24.903 24.904 cp_dbcsr_syevd 83 12.4 0.005 0.005 24.116 24.118 init_scf_loop 11 6.9 0.024 0.187 21.789 21.791 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 21.236 21.435 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.635 20.668 cp_fm_diag_elpa_base 83 14.4 20.537 20.597 20.629 20.664 multiply_cannon_metrocomm3 60168 15.6 0.117 0.122 16.697 18.985 prepare_preconditioner 11 7.9 0.000 0.000 16.914 16.945 make_preconditioner 11 8.9 0.000 0.002 16.914 16.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.128 16.333 make_m2s 5014 13.6 0.105 0.114 14.526 14.908 make_images 5014 14.6 0.399 0.418 14.341 14.732 sum_up_and_integrate 128 10.3 0.002 0.004 14.636 14.652 integrate_v_rspace 128 11.3 0.003 0.004 14.578 14.596 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.916 14.097 calculate_rho_elec 128 8.7 0.045 0.065 13.915 14.096 init_scf_run 11 5.9 0.000 0.002 13.279 13.283 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.279 13.283 mp_sum_l 7950 12.9 9.957 11.137 9.957 11.137 dbcsr_mm_accdrv_process 124484 16.2 4.772 4.924 9.603 10.199 cp_fm_cholesky_invert 11 10.9 9.697 9.706 9.697 9.706 wfi_extrapolate 11 7.9 0.001 0.001 9.345 9.345 calculate_dm_sparse 128 9.5 0.001 0.001 8.700 8.775 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.404 8.572 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.279 8.354 multiply_cannon_metrocomm1 60168 15.6 0.090 0.096 6.705 8.268 density_rs2pw 128 9.7 0.005 0.006 7.357 8.077 make_images_data 5014 15.6 0.065 0.072 7.055 8.012 grid_integrate_task_list 128 12.3 7.045 7.610 7.045 7.610 hybrid_alltoall_any 5200 16.5 0.294 2.270 6.152 7.494 fft_wrap_pw1pw2 1291 11.7 0.018 0.026 7.068 7.399 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.989 7.003 fft3d_ps 1291 13.7 2.094 2.597 5.894 6.167 fft_wrap_pw1pw2_140 523 12.2 0.234 0.251 5.815 6.078 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.900 6.000 mp_alltoall_d11v 2415 14.1 4.699 5.883 4.699 5.883 grid_collocate_task_list 128 9.7 4.836 5.164 4.836 5.164 potential_pw2rs 128 12.3 0.009 0.010 5.087 5.159 cp_fm_cholesky_decompose 22 10.9 5.005 5.017 5.005 5.017 transfer_rs2pw 523 10.6 0.008 0.009 3.939 4.723 mp_sum_d 4474 12.1 3.789 4.662 3.789 4.662 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=220.212000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.545455, yerr=7.062882 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 842.002432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132588E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702167E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828901424 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57636. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 5154197. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.143 0.213 200.116 200.147 qs_mol_dyn_low 1 2.0 0.051 0.145 199.098 199.114 qs_forces 11 3.9 0.005 0.006 198.658 198.711 qs_energies 11 4.9 0.022 0.128 191.861 191.917 scf_env_do_scf 11 5.9 0.021 0.163 173.724 173.742 scf_env_do_scf_inner_loop 118 6.6 0.047 0.300 138.155 138.171 velocity_verlet 10 3.0 0.103 0.312 122.890 122.904 dbcsr_multiply_generic 2529 12.6 0.203 0.209 99.332 100.712 qs_scf_new_mos 118 7.6 0.001 0.001 98.589 99.298 qs_scf_loop_do_ot 118 8.6 0.021 0.158 98.588 99.297 ot_scf_mini 118 9.6 0.004 0.005 93.699 94.388 multiply_cannon 2529 13.6 0.511 0.566 78.273 83.196 multiply_cannon_loop 2529 14.6 1.581 1.650 74.694 77.527 ot_mini 118 10.6 0.001 0.001 51.926 52.728 mp_waitall_1 216598 16.6 24.746 39.205 24.746 39.205 multiply_cannon_multrec 30348 15.6 21.127 25.576 31.853 36.809 init_scf_loop 11 6.9 0.017 0.135 35.444 35.459 rebuild_ks_matrix 129 8.3 0.001 0.001 34.167 34.807 qs_ks_build_kohn_sham_matrix 129 9.3 0.119 0.542 34.166 34.806 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.941 31.549 qs_ot_get_derivative 118 11.6 0.001 0.002 29.696 30.408 prepare_preconditioner 11 7.9 0.000 0.000 29.762 29.879 make_preconditioner 11 8.9 0.019 0.148 29.762 29.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.395 29.019 multiply_cannon_metrocomm3 30348 15.6 0.096 0.101 15.896 28.692 qs_ot_get_p 129 10.4 0.007 0.014 23.663 24.272 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.322 23.620 apply_single 129 13.6 0.001 0.001 22.322 23.620 ot_diis_step 118 11.6 0.015 0.015 22.052 22.054 multiply_cannon_sync_h2d 30348 15.6 18.131 20.372 18.131 20.372 qs_ot_p2m_diag 84 11.4 0.191 0.219 18.542 18.580 cp_dbcsr_syevd 84 12.4 0.011 0.046 17.266 17.268 cp_fm_cholesky_invert 11 10.9 17.215 17.228 17.215 17.228 make_m2s 5058 13.6 0.087 0.093 14.556 16.315 make_images 5058 14.6 1.173 1.375 14.337 16.099 sum_up_and_integrate 129 10.3 0.025 0.183 14.969 15.035 integrate_v_rspace 129 11.3 0.003 0.004 14.861 14.979 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.856 13.926 calculate_rho_elec 129 8.7 0.092 0.137 13.855 13.925 cp_fm_diag_elpa 84 13.4 0.000 0.000 13.710 13.744 cp_fm_diag_elpa_base 84 14.4 13.428 13.542 13.700 13.729 init_scf_run 11 5.9 0.000 0.001 12.723 12.724 scf_env_initial_rho_setup 11 6.9 0.002 0.006 12.723 12.724 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.682 12.171 multiply_cannon_metrocomm4 27819 15.6 0.107 0.122 3.893 11.035 dbcsr_mm_accdrv_process 62758 16.2 5.512 6.427 10.170 10.856 mp_irecv_dv 70084 16.3 3.681 10.613 3.681 10.613 make_images_data 5058 15.6 0.066 0.076 8.341 10.375 hybrid_alltoall_any 5245 16.5 0.353 1.522 7.242 9.626 wfi_extrapolate 11 7.9 0.006 0.045 8.433 8.433 fft_wrap_pw1pw2 1301 11.7 0.019 0.027 8.120 8.294 density_rs2pw 129 9.7 0.009 0.037 7.261 7.802 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.012 7.792 grid_integrate_task_list 129 12.3 7.235 7.740 7.235 7.740 cp_fm_cholesky_decompose 22 10.9 7.332 7.403 7.332 7.403 fft_wrap_pw1pw2_140 527 12.2 0.253 0.275 7.050 7.196 mp_sum_l 8016 12.9 4.825 6.873 4.825 6.873 fft3d_ps 1301 13.7 2.785 2.956 6.547 6.724 calculate_dm_sparse 129 9.5 0.001 0.001 6.578 6.721 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.271 6.287 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.510 5.582 potential_pw2rs 129 12.3 0.016 0.031 5.366 5.478 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.272 5.458 grid_collocate_task_list 129 9.7 5.040 5.422 5.040 5.422 mp_alltoall_d11v 2429 14.1 4.540 5.352 4.540 5.352 mp_allgather_i34 2529 14.6 2.086 4.991 2.086 4.991 mp_sum_d 4513 12.1 3.068 4.432 3.068 4.432 dbcsr_complete_redistribute 397 12.7 0.792 0.906 3.362 4.240 transfer_rs2pw 527 10.6 0.007 0.008 3.584 4.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=200.147000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.818182, yerr=2.724241 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 993.517568E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.127 184.482 184.487 qs_mol_dyn_low 1 2.0 0.044 0.119 183.632 183.645 qs_forces 11 3.9 0.006 0.013 183.303 183.344 qs_energies 11 4.9 0.054 0.418 176.329 176.664 scf_env_do_scf 11 5.9 0.001 0.004 159.153 159.155 scf_env_do_scf_inner_loop 116 6.6 0.010 0.057 122.644 122.644 velocity_verlet 10 3.0 0.068 0.172 114.191 114.210 qs_scf_new_mos 116 7.6 0.001 0.001 84.879 85.252 qs_scf_loop_do_ot 116 8.6 0.001 0.002 84.878 85.251 dbcsr_multiply_generic 2485 12.5 0.192 0.197 83.420 84.723 ot_scf_mini 116 9.6 0.004 0.007 80.650 81.113 multiply_cannon 2485 13.5 0.502 0.520 63.230 66.885 multiply_cannon_loop 2485 14.5 1.117 1.178 59.918 63.090 ot_mini 116 10.6 0.001 0.001 43.503 43.970 init_scf_loop 11 6.9 0.004 0.035 36.406 36.407 mp_waitall_1 169034 16.6 25.164 34.759 25.164 34.759 prepare_preconditioner 11 7.9 0.000 0.000 32.247 32.292 make_preconditioner 11 8.9 0.005 0.039 32.247 32.292 rebuild_ks_matrix 127 8.3 0.001 0.001 30.886 31.339 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.030 30.886 31.339 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.736 31.226 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.875 28.293 multiply_cannon_multrec 19880 15.5 13.006 15.966 22.852 25.785 multiply_cannon_metrocomm3 19880 15.5 0.059 0.062 15.162 24.617 qs_ot_get_derivative 116 11.6 0.001 0.002 24.109 24.584 qs_ot_get_p 127 10.4 0.002 0.003 22.420 22.969 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.522 20.514 apply_single 127 13.6 0.001 0.001 19.522 20.513 ot_diis_step 116 11.6 0.018 0.018 19.288 19.288 qs_ot_p2m_diag 82 11.4 0.262 0.269 17.868 17.880 cp_dbcsr_syevd 82 12.4 0.042 0.303 16.772 16.773 make_m2s 4970 13.5 0.076 0.080 15.010 16.452 make_images 4970 14.5 1.137 1.237 14.776 16.214 multiply_cannon_sync_h2d 19880 15.5 13.659 15.681 13.659 15.681 cp_fm_cholesky_invert 11 10.9 15.172 15.181 15.172 15.181 sum_up_and_integrate 127 10.3 0.002 0.003 14.166 14.191 integrate_v_rspace 127 11.3 0.003 0.004 14.106 14.134 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.734 13.798 calculate_rho_elec 127 8.7 0.131 0.145 13.733 13.797 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.121 13.178 cp_fm_diag_elpa_base 82 14.4 12.446 12.620 12.873 12.894 init_scf_run 11 5.9 0.000 0.001 11.540 11.540 scf_env_initial_rho_setup 11 6.9 0.003 0.008 11.540 11.540 make_images_data 4970 15.5 0.061 0.072 8.894 10.882 hybrid_alltoall_any 5155 16.4 0.447 2.043 7.808 10.052 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.411 9.747 dbcsr_mm_accdrv_process 41158 16.2 6.013 6.581 9.305 9.521 multiply_cannon_metrocomm4 17395 15.5 0.066 0.077 3.453 9.403 mp_irecv_dv 49801 16.2 3.324 9.144 3.324 9.144 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 7.812 7.930 cp_fm_cholesky_decompose 22 10.9 7.757 7.857 7.757 7.857 grid_integrate_task_list 127 12.3 7.232 7.775 7.232 7.775 wfi_extrapolate 11 7.9 0.001 0.001 7.576 7.576 density_rs2pw 127 9.7 0.006 0.010 7.046 7.488 cp_fm_upper_to_full 104 14.8 5.746 7.392 5.746 7.392 fft_wrap_pw1pw2_140 519 12.2 0.283 0.293 6.716 6.864 dbcsr_complete_redistribute 393 12.7 1.179 1.263 4.961 6.699 fft3d_ps 1281 13.7 2.742 2.969 6.132 6.209 calculate_dm_sparse 127 9.5 0.001 0.001 5.968 6.086 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.978 5.990 mp_alltoall_d11v 2401 14.1 4.609 5.754 4.609 5.754 grid_collocate_task_list 127 9.7 5.108 5.592 5.108 5.592 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.778 5.502 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.757 5.494 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.948 5.082 mp_sum_l 7884 12.9 3.390 4.843 3.390 4.843 mp_allgather_i34 2485 14.5 1.822 4.793 1.822 4.793 potential_pw2rs 127 12.3 0.020 0.022 4.605 4.643 qs_energies_init_hamiltonians 11 5.9 0.007 0.036 4.173 4.216 transfer_fm_to_dbcsr 11 9.9 0.019 0.025 2.485 4.185 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.992 4.031 mp_alltoall_i22 712 14.1 2.114 3.986 2.114 3.986 transfer_rs2pw 519 10.6 0.007 0.007 3.445 3.966 mp_sum_d 4454 12.1 2.799 3.882 2.799 3.882 calculate_first_density_matrix 1 7.0 0.003 0.025 3.750 3.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=184.487000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=932.454545, yerr=25.349784 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.150898E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58189. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.125 194.188 194.191 qs_mol_dyn_low 1 2.0 0.020 0.140 193.371 193.382 qs_forces 11 3.9 0.004 0.005 193.050 193.063 qs_energies 11 4.9 0.024 0.165 185.866 185.879 scf_env_do_scf 11 5.9 0.006 0.040 166.906 166.925 velocity_verlet 10 3.0 0.028 0.043 124.197 124.218 scf_env_do_scf_inner_loop 116 6.6 0.024 0.167 118.697 118.698 qs_scf_new_mos 116 7.6 0.001 0.001 82.925 83.216 qs_scf_loop_do_ot 116 8.6 0.001 0.001 82.924 83.216 dbcsr_multiply_generic 2485 12.5 0.195 0.200 80.449 81.318 ot_scf_mini 116 9.6 0.003 0.004 78.403 78.704 multiply_cannon 2485 13.5 0.546 0.575 55.654 59.268 multiply_cannon_loop 2485 14.5 1.815 1.893 51.819 53.911 init_scf_loop 11 6.9 0.016 0.130 48.082 48.096 ot_mini 116 10.6 0.001 0.001 42.941 43.243 prepare_preconditioner 11 7.9 0.000 0.000 42.686 42.772 make_preconditioner 11 8.9 0.010 0.079 42.686 42.772 make_full_inverse_cholesky 11 9.9 0.010 0.023 36.376 41.311 multiply_cannon_multrec 29820 15.5 13.491 18.796 26.384 31.341 rebuild_ks_matrix 127 8.3 0.001 0.001 30.263 30.626 qs_ks_build_kohn_sham_matrix 127 9.3 0.148 0.682 30.262 30.626 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.466 27.799 mp_waitall_1 146592 16.7 16.948 26.486 16.948 26.486 qs_ot_get_derivative 116 11.6 0.001 0.002 23.367 23.672 qs_ot_get_p 127 10.4 0.012 0.017 21.690 22.023 make_m2s 4970 13.5 0.092 0.097 20.196 21.429 make_images 4970 14.5 1.920 2.195 19.887 21.117 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.946 19.562 apply_single 127 13.6 0.001 0.001 18.945 19.562 ot_diis_step 116 11.6 0.018 0.018 19.444 19.446 qs_ot_p2m_diag 82 11.4 0.339 0.385 17.371 17.426 cp_fm_upper_to_full 104 14.8 11.351 16.665 11.351 16.665 cp_fm_cholesky_invert 11 10.9 16.358 16.367 16.358 16.367 cp_dbcsr_syevd 82 12.4 0.016 0.092 15.958 15.960 multiply_cannon_metrocomm3 29820 15.5 0.049 0.052 6.186 14.759 sum_up_and_integrate 127 10.3 0.025 0.185 14.170 14.238 integrate_v_rspace 127 11.3 0.003 0.004 14.088 14.180 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.679 13.753 calculate_rho_elec 127 8.7 0.173 0.188 13.679 13.752 dbcsr_complete_redistribute 393 12.7 1.526 1.695 9.362 13.200 dbcsr_mm_accdrv_process 61748 16.2 8.364 9.268 12.461 12.944 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.628 12.653 make_images_data 4970 15.5 0.064 0.070 10.768 12.637 cp_fm_diag_elpa_base 82 14.4 11.592 11.908 12.557 12.571 init_scf_run 11 5.9 0.000 0.001 12.363 12.364 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.363 12.364 multiply_cannon_sync_h2d 29820 15.5 10.477 11.854 10.477 11.854 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.975 11.782 hybrid_alltoall_any 5155 16.4 0.523 2.197 9.404 11.430 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.742 9.963 transfer_fm_to_dbcsr 11 9.9 0.002 0.012 6.262 9.927 mp_alltoall_i22 712 14.1 5.813 9.781 5.813 9.781 fft_wrap_pw1pw2 1281 11.7 0.019 0.024 8.262 8.377 cp_fm_cholesky_decompose 22 10.9 7.738 7.840 7.738 7.840 grid_integrate_task_list 127 12.3 7.462 7.808 7.462 7.808 wfi_extrapolate 11 7.9 0.003 0.014 7.611 7.611 fft_wrap_pw1pw2_140 519 12.2 0.326 0.341 7.333 7.475 density_rs2pw 127 9.7 0.011 0.051 6.562 6.941 multiply_cannon_metrocomm4 24850 15.5 0.078 0.088 2.788 6.918 mp_irecv_dv 75445 16.2 2.635 6.633 2.635 6.633 fft3d_ps 1281 13.7 2.930 2.965 6.446 6.560 calculate_dm_sparse 127 9.5 0.001 0.001 6.224 6.329 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.864 5.998 mp_alltoall_d11v 2401 14.1 5.007 5.598 5.007 5.598 grid_collocate_task_list 127 9.7 5.276 5.571 5.276 5.571 qs_energies_init_hamiltonians 11 5.9 0.003 0.011 4.643 4.652 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.542 4.633 potential_pw2rs 127 12.3 0.024 0.038 4.537 4.606 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.432 4.535 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.237 4.307 calculate_first_density_matrix 1 7.0 0.015 0.116 4.125 4.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=194.191000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1090.454545, yerr=15.417603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 5.910792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 1978768 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 145.661668E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.554715E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243264 MPI messages size (bytes): total size 1.342058E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.516879E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115488 60548972544 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9025 51. MP_Alltoall 9734 793691. MP_ISend 40500 2096702. MP_IRecv 40500 2095807. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57624. MP_Allreduce 11184 1163. MP_Sync 88 MP_Alltoall 1724 18848021. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.075 183.781 183.781 qs_mol_dyn_low 1 2.0 0.003 0.003 183.134 183.147 qs_forces 11 3.9 0.025 0.027 183.022 183.026 qs_energies 11 4.9 0.003 0.004 175.516 175.521 scf_env_do_scf 11 5.9 0.001 0.001 157.364 157.377 scf_env_do_scf_inner_loop 118 6.6 0.019 0.024 118.526 118.527 velocity_verlet 10 3.0 0.022 0.024 118.349 118.353 qs_scf_new_mos 118 7.6 0.001 0.001 80.907 80.994 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.906 80.994 ot_scf_mini 118 9.6 0.003 0.004 76.405 76.474 dbcsr_multiply_generic 2534 12.6 0.190 0.199 74.237 74.538 multiply_cannon 2534 13.6 0.567 0.599 54.543 58.373 multiply_cannon_loop 2534 14.6 0.822 0.853 51.481 52.133 ot_mini 118 10.6 0.001 0.001 38.977 39.015 init_scf_loop 11 6.9 0.000 0.000 38.679 38.681 prepare_preconditioner 11 7.9 0.000 0.000 34.760 34.783 make_preconditioner 11 8.9 0.000 0.000 34.760 34.783 make_full_inverse_cholesky 11 9.9 0.016 0.028 32.541 32.859 mp_waitall_1 127116 16.7 25.794 31.791 25.794 31.791 rebuild_ks_matrix 129 8.3 0.001 0.001 29.869 29.929 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 29.869 29.929 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.300 27.358 qs_ot_get_p 129 10.4 0.025 0.026 24.473 24.545 qs_ot_get_derivative 118 11.6 0.002 0.002 21.874 21.940 multiply_cannon_multrec 10136 15.6 10.466 14.862 18.233 21.600 qs_ot_p2m_diag 84 11.4 0.502 0.508 20.271 20.289 cp_fm_cholesky_invert 11 10.9 20.254 20.263 20.254 20.263 multiply_cannon_metrocomm3 10136 15.6 0.025 0.026 12.579 19.251 cp_dbcsr_syevd 84 12.4 0.005 0.006 18.956 18.958 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 17.125 17.387 apply_single 129 13.6 0.001 0.001 17.124 17.386 ot_diis_step 118 11.6 0.020 0.022 17.017 17.017 make_m2s 5068 13.6 0.065 0.070 15.542 16.362 make_images 5068 14.6 2.170 2.610 15.230 16.052 cp_fm_diag_elpa 84 13.4 0.000 0.000 15.514 15.522 cp_fm_diag_elpa_base 84 14.4 15.228 15.337 15.509 15.517 sum_up_and_integrate 129 10.3 0.002 0.002 14.402 14.448 integrate_v_rspace 129 11.3 0.004 0.004 14.341 14.389 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.077 14.112 calculate_rho_elec 129 8.7 0.259 0.270 14.076 14.111 multiply_cannon_sync_h2d 10136 15.6 10.916 11.283 10.916 11.283 init_scf_run 11 5.9 0.000 0.001 11.160 11.160 scf_env_initial_rho_setup 11 6.9 0.002 0.003 11.160 11.160 make_images_data 5068 15.6 0.056 0.069 8.884 10.239 hybrid_alltoall_any 5255 16.5 0.847 3.792 8.616 10.051 cp_fm_cholesky_decompose 22 10.9 8.707 8.854 8.707 8.854 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.757 8.794 fft_wrap_pw1pw2 1301 11.7 0.018 0.020 8.389 8.422 grid_integrate_task_list 129 12.3 7.828 8.227 7.828 8.227 dbcsr_mm_accdrv_process 20954 16.1 3.175 4.246 7.415 8.127 wfi_extrapolate 11 7.9 0.001 0.001 7.574 7.574 multiply_cannon_metrocomm1 10136 15.6 0.030 0.031 4.832 7.533 fft_wrap_pw1pw2_140 527 12.2 0.439 0.446 7.264 7.305 density_rs2pw 129 9.7 0.005 0.005 6.673 7.006 calculate_dm_sparse 129 9.5 0.001 0.001 6.382 6.483 fft3d_ps 1301 13.7 3.139 3.241 6.348 6.363 dbcsr_complete_redistribute 397 12.7 2.079 2.144 5.494 5.920 grid_collocate_task_list 129 9.7 5.610 5.869 5.610 5.869 mp_alltoall_d11v 2429 14.1 4.975 5.796 4.975 5.796 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.403 5.411 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.364 5.365 mp_allgather_i34 2534 14.6 1.236 4.783 1.236 4.783 multiply_cannon_metrocomm4 7602 15.6 0.025 0.027 1.796 4.553 mp_irecv_dv 29142 15.9 1.757 4.474 1.757 4.474 potential_pw2rs 129 12.3 0.026 0.027 4.402 4.415 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.281 4.324 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.583 3.945 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.597 3.907 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.748 3.824 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.739 3.811 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=183.781000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1480.090909, yerr=6.841777 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.148874E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3991 59293. MP_Allreduce 11055 1504. MP_Sync 86 MP_Alltoall 1700 36954339. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.130 287.882 287.883 qs_mol_dyn_low 1 2.0 0.003 0.003 287.092 287.129 qs_forces 11 3.9 0.009 0.010 286.987 286.989 qs_energies 11 4.9 0.002 0.003 278.302 278.306 scf_env_do_scf 11 5.9 0.001 0.001 255.381 255.388 velocity_verlet 10 3.0 0.018 0.019 207.952 207.960 scf_env_do_scf_inner_loop 116 6.6 0.012 0.016 130.165 130.167 init_scf_loop 11 6.9 0.000 0.000 124.931 124.935 prepare_preconditioner 11 7.9 0.000 0.000 120.132 120.156 make_preconditioner 11 8.9 0.000 0.000 120.132 120.155 make_full_inverse_cholesky 11 9.9 0.038 0.039 95.663 117.296 qs_scf_new_mos 116 7.6 0.001 0.001 88.473 88.567 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.472 88.567 ot_scf_mini 116 9.6 0.004 0.004 83.704 83.802 dbcsr_multiply_generic 2485 12.5 0.227 0.238 81.453 82.005 cp_fm_upper_to_full 104 14.8 53.507 76.973 53.507 76.973 multiply_cannon 2485 13.5 0.667 0.708 59.020 59.616 multiply_cannon_loop 2485 14.5 1.047 1.063 55.069 56.543 ot_mini 116 10.6 0.001 0.001 43.998 44.114 dbcsr_complete_redistribute 393 12.7 4.006 4.049 30.417 43.698 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.970 40.223 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.427 37.571 mp_alltoall_i22 712 14.1 22.274 35.330 22.274 35.330 cp_fm_cholesky_invert 11 10.9 33.116 33.122 33.116 33.122 rebuild_ks_matrix 127 8.3 0.001 0.001 33.033 33.095 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.019 33.032 33.094 mp_waitall_1 102768 16.8 27.435 31.997 27.435 31.997 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.712 30.777 qs_ot_get_p 127 10.4 0.024 0.033 24.915 25.047 qs_ot_get_derivative 116 11.6 0.002 0.002 24.633 24.730 qs_ot_p2m_diag 82 11.4 0.869 0.874 20.915 20.944 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 18.836 20.430 ot_diis_step 116 11.6 0.021 0.022 19.337 19.337 make_m2s 4970 13.5 0.073 0.075 17.871 19.300 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.138 19.140 make_images 4970 14.5 3.052 3.233 17.393 18.822 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.674 18.772 apply_single 127 13.6 0.001 0.001 18.674 18.772 multiply_cannon_multrec 9940 15.5 10.188 12.015 18.285 18.492 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.016 16.017 cp_fm_diag_elpa_base 82 14.4 11.665 13.245 16.013 16.013 sum_up_and_integrate 127 10.3 0.002 0.002 15.436 15.527 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.444 15.523 calculate_rho_elec 127 8.7 0.478 0.479 15.443 15.522 integrate_v_rspace 127 11.3 0.004 0.005 15.374 15.466 multiply_cannon_sync_h2d 9940 15.5 14.180 14.208 14.180 14.208 init_scf_run 11 5.9 0.000 0.001 12.696 12.697 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.696 12.697 hybrid_alltoall_any 5155 16.4 1.302 3.020 10.188 12.265 make_images_data 4970 15.5 0.062 0.067 9.872 11.991 dbcsr_mm_accdrv_process 20590 16.0 4.453 6.348 7.847 9.710 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.547 9.620 fft_wrap_pw1pw2 1281 11.7 0.021 0.021 9.598 9.605 cp_fm_cholesky_decompose 22 10.9 8.826 8.846 8.826 8.846 wfi_extrapolate 11 7.9 0.001 0.001 8.802 8.802 grid_integrate_task_list 127 12.3 8.496 8.682 8.496 8.682 fft_wrap_pw1pw2_140 519 12.2 0.818 0.819 8.503 8.517 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.067 8.067 calculate_dm_sparse 127 9.5 0.001 0.001 6.941 6.987 fft3d_ps 1281 13.7 3.954 3.961 6.921 6.953 mp_alltoall_d11v 2401 14.1 6.651 6.833 6.651 6.833 density_rs2pw 127 9.7 0.005 0.005 6.595 6.637 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.491 6.587 grid_collocate_task_list 127 9.7 6.363 6.433 6.363 6.433 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.080 6.159 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.883000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2804.727273, yerr=173.660113 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/21/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 2.766000E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 419739 0.0% 0.0% 100.0% average stack size 0.0 0.0 22952.9 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 1.263923E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 33980775. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.299 86.292 86.300 qs_energies 1 2.0 0.000 0.000 85.189 85.198 ls_scf 1 3.0 0.000 0.000 84.205 84.214 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.884 73.029 multiply_cannon 111 7.7 0.017 0.020 56.193 57.275 multiply_cannon_loop 111 8.7 0.227 0.242 52.769 54.034 ls_scf_main 1 4.0 0.000 0.000 52.489 52.494 density_matrix_trs4 2 5.0 0.002 0.003 46.887 46.985 ls_scf_init_scf 1 4.0 0.000 0.000 28.680 28.686 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.358 27.405 mp_waitall_1 11031 10.9 22.379 25.612 22.379 25.612 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.006 25.150 25.171 multiply_cannon_multrec 2664 9.7 8.143 8.906 15.543 17.265 multiply_cannon_sync_h2d 2664 9.7 13.751 15.079 13.751 15.079 make_m2s 222 7.7 0.009 0.010 13.123 13.707 make_images 222 8.7 0.099 0.108 13.101 13.687 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.648 12.121 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.397 8.252 make_images_data 222 9.7 0.004 0.005 7.672 8.138 dbcsr_mm_accdrv_process 4760 10.4 0.588 0.709 7.020 7.992 hybrid_alltoall_any 227 10.6 0.217 1.832 6.618 7.989 dbcsr_mm_accdrv_process_sort 4760 11.4 6.233 7.133 6.233 7.133 calculate_norms 4752 9.8 5.534 6.161 5.534 6.161 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.113 5.243 mp_sum_l 887 5.1 3.277 4.865 3.277 4.865 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.294 3.652 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.036 3.540 mp_irecv_dv 6231 10.9 2.019 3.513 2.019 3.513 arnoldi_extremal 4 6.8 0.000 0.000 3.298 3.331 arnoldi_normal_ev 4 7.8 0.001 0.003 3.298 3.331 make_images_sizes 222 9.7 0.000 0.000 0.698 3.314 mp_alltoall_i44 222 10.7 0.698 3.314 0.698 3.314 build_subspace 16 8.4 0.009 0.012 3.205 3.207 ls_scf_post 1 4.0 0.000 0.000 3.036 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.851 2.889 dbcsr_special_finalize 555 9.7 0.005 0.006 2.397 2.790 dbcsr_merge_single_wm 555 10.7 0.462 0.592 2.388 2.782 make_images_pack 222 9.7 2.210 2.631 2.212 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.360 2.585 dbcsr_sort_data 658 11.4 2.184 2.515 2.184 2.515 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.452 2.067 2.454 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.301 2.379 buffer_matrices_ensure_size 222 8.7 1.751 2.047 1.751 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.869 1.875 rebuild_ks_matrix 3 7.3 0.000 0.000 1.859 1.865 qs_ks_build_kohn_sham_matrix 3 8.3 0.012 0.030 1.859 1.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.300000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1128.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/22/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 5.588524E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 368848 0.0% 0.0% 100.0% average stack size 0.0 0.0 26119.8 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.174337E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.218 91.839 91.849 qs_energies 1 2.0 0.000 0.004 90.911 90.931 ls_scf 1 3.0 0.009 0.035 89.357 89.377 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.200 74.470 multiply_cannon 111 7.7 0.028 0.043 52.917 57.188 ls_scf_main 1 4.0 0.003 0.039 54.566 54.593 multiply_cannon_loop 111 8.7 0.135 0.147 50.077 53.247 density_matrix_trs4 2 5.0 0.014 0.098 48.769 48.958 ls_scf_init_scf 1 4.0 0.004 0.016 31.080 31.112 mp_waitall_1 9105 10.9 20.637 29.825 20.637 29.825 ls_scf_init_matrix_S 1 5.0 0.012 0.064 28.964 29.081 multiply_cannon_multrec 1332 9.7 13.308 17.119 22.612 27.595 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.016 26.478 26.505 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.665 20.964 make_m2s 222 7.7 0.006 0.008 14.831 15.428 make_images 222 8.7 1.371 1.692 14.800 15.398 dbcsr_mm_accdrv_process 4041 10.4 0.355 0.598 8.902 10.425 dbcsr_mm_accdrv_process_sort 4041 11.4 8.407 9.893 8.407 9.893 make_images_data 222 9.7 0.004 0.005 8.443 9.339 hybrid_alltoall_any 227 10.6 0.542 2.552 7.817 9.267 mp_sum_l 887 5.1 5.085 8.133 5.085 8.133 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.195 7.613 mp_irecv_dv 3311 11.0 3.174 7.555 3.174 7.555 calculate_norms 2376 9.8 6.048 6.702 6.048 6.702 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.704 6.273 multiply_cannon_sync_h2d 1332 9.7 4.893 5.942 4.893 5.942 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.202 arnoldi_extremal 4 6.8 0.000 0.000 4.866 4.907 arnoldi_normal_ev 4 7.8 0.002 0.006 4.866 4.907 build_subspace 16 8.4 0.014 0.021 4.573 4.576 ls_scf_post 1 4.0 0.004 0.018 3.703 3.723 ls_scf_store_result 1 5.0 0.000 0.000 3.419 3.496 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.219 3.450 dbcsr_matrix_vector_mult_local 304 10.0 2.760 3.250 2.762 3.252 mp_allgather_i34 111 8.7 0.829 3.028 0.829 3.028 ls_scf_dm_to_ks 2 5.0 0.015 0.101 2.647 2.721 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.162 2.541 dbcsr_data_new 4174 10.1 2.112 2.406 2.112 2.406 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.324 2.357 rebuild_ks_matrix 3 7.3 0.000 0.000 2.311 2.344 qs_ks_build_kohn_sham_matrix 3 8.3 0.033 0.096 2.311 2.344 ls_scf_initial_guess 1 5.0 0.000 0.000 2.112 2.219 ls_scf_qs_atomic_guess 1 6.0 0.000 0.008 2.112 2.219 make_images_pack 222 9.7 1.830 2.117 1.832 2.119 dbcsr_sort_data 436 11.2 1.838 2.097 1.838 2.097 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.849000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1793.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/23/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 8.404608E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 353133 0.0% 0.0% 100.0% average stack size 0.0 0.0 27282.1 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 2.960896E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.063 0.145 93.433 93.441 qs_energies 1 2.0 0.002 0.008 92.413 92.468 ls_scf 1 3.0 0.006 0.023 90.723 90.727 dbcsr_multiply_generic 111 6.7 0.017 0.029 74.389 74.737 multiply_cannon 111 7.7 0.036 0.070 51.568 56.164 ls_scf_main 1 4.0 0.002 0.028 55.936 55.948 multiply_cannon_loop 111 8.7 0.117 0.139 48.844 52.403 density_matrix_trs4 2 5.0 0.010 0.051 49.936 50.140 mp_waitall_1 7281 11.0 23.035 32.683 23.035 32.683 ls_scf_init_scf 1 4.0 0.003 0.021 31.073 31.080 ls_scf_init_matrix_S 1 5.0 0.001 0.003 29.237 29.317 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.761 26.787 multiply_cannon_multrec 888 9.7 12.698 15.201 21.299 24.426 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.530 22.270 make_m2s 222 7.7 0.006 0.008 16.462 17.146 make_images 222 8.7 1.590 1.850 16.424 17.105 make_images_data 222 9.7 0.004 0.004 9.564 10.555 hybrid_alltoall_any 227 10.6 0.642 2.944 9.031 10.226 dbcsr_mm_accdrv_process 3754 10.4 0.328 0.535 8.125 9.399 dbcsr_mm_accdrv_process_sort 3754 11.4 7.680 8.864 7.680 8.864 mp_sum_l 887 5.1 4.696 7.931 4.696 7.931 multiply_cannon_sync_h2d 888 9.7 6.030 7.388 6.030 7.388 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.584 6.951 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.461 6.855 mp_irecv_dv 2335 11.1 2.446 6.816 2.446 6.816 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.524 6.333 arnoldi_extremal 4 6.8 0.000 0.000 5.238 5.269 arnoldi_normal_ev 4 7.8 0.001 0.006 5.238 5.269 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.743 4.989 build_subspace 16 8.4 0.014 0.020 4.846 4.853 calculate_norms 1584 9.8 4.411 4.834 4.411 4.834 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.466 3.797 ls_scf_post 1 4.0 0.001 0.004 3.707 3.713 mp_allgather_i34 111 8.7 0.816 3.707 0.816 3.707 dbcsr_matrix_vector_mult_local 304 10.0 3.054 3.636 3.056 3.638 ls_scf_store_result 1 5.0 0.000 0.000 3.429 3.518 ls_scf_dm_to_ks 2 5.0 0.003 0.035 2.867 2.985 dbcsr_data_new 4116 9.9 2.112 2.460 2.112 2.460 make_images_sizes 222 9.7 0.000 0.000 1.061 2.372 mp_alltoall_i44 222 10.7 1.060 2.371 1.060 2.371 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.201 2.208 rebuild_ks_matrix 3 7.3 0.000 0.000 2.183 2.190 qs_ks_build_kohn_sham_matrix 3 8.3 0.014 0.026 2.183 2.190 dbcsr_sort_data 325 11.1 1.870 2.169 1.870 2.169 ls_scf_initial_guess 1 5.0 0.000 0.000 1.832 1.934 ls_scf_qs_atomic_guess 1 6.0 0.000 0.011 1.832 1.934 dbcsr_finalize 304 7.8 0.026 0.032 1.623 1.932 make_images_pack 222 9.7 1.627 1.892 1.630 1.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.441000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2286.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/24/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 10.747127E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 369794 0.0% 0.0% 100.0% average stack size 0.0 0.0 26053.0 marketing flops 1.742116E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 3.435770E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.076 96.725 96.726 qs_energies 1 2.0 0.000 0.000 95.988 95.992 ls_scf 1 3.0 0.000 0.000 94.307 94.313 dbcsr_multiply_generic 111 6.7 0.016 0.018 77.489 77.728 ls_scf_main 1 4.0 0.000 0.000 58.218 58.241 multiply_cannon 111 7.7 0.050 0.124 51.030 55.409 density_matrix_trs4 2 5.0 0.002 0.003 52.067 52.187 multiply_cannon_loop 111 8.7 0.151 0.165 45.915 49.383 ls_scf_init_scf 1 4.0 0.000 0.000 32.848 32.910 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.542 31.614 matrix_sqrt_Newton_Schulz 2 6.5 0.007 0.017 28.910 28.924 mp_waitall_1 6369 11.0 21.660 28.437 21.660 28.437 multiply_cannon_multrec 1332 9.7 14.159 17.538 22.085 25.061 make_m2s 222 7.7 0.006 0.008 20.817 22.223 make_images 222 8.7 3.147 3.605 20.767 22.175 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 8.380 16.440 make_images_data 222 9.7 0.004 0.004 11.539 13.136 hybrid_alltoall_any 227 10.6 0.800 3.825 11.017 12.743 dbcsr_mm_accdrv_process 3641 10.4 0.297 0.488 7.567 9.087 dbcsr_mm_accdrv_process_sort 3641 11.4 7.121 8.590 7.121 8.590 mp_sum_l 887 5.1 4.065 7.546 4.065 7.546 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.117 6.098 mp_irecv_dv 3229 10.9 2.092 6.010 2.092 6.010 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.097 5.961 multiply_cannon_sync_h2d 1332 9.7 5.380 5.823 5.380 5.823 arnoldi_extremal 4 6.8 0.000 0.000 5.445 5.483 arnoldi_normal_ev 4 7.8 0.021 0.042 5.445 5.482 build_subspace 16 8.4 0.014 0.021 5.073 5.080 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.570 5.021 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.532 4.730 mp_allgather_i34 111 8.7 2.205 4.653 2.205 4.653 calculate_norms 2376 9.8 4.227 4.603 4.227 4.603 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.710 3.983 dbcsr_matrix_vector_mult_local 304 10.0 3.242 3.745 3.244 3.747 dbcsr_sort_data 658 11.4 3.094 3.367 3.094 3.367 ls_scf_post 1 4.0 0.000 0.000 3.240 3.244 dbcsr_special_finalize 555 9.7 0.006 0.007 2.836 3.134 dbcsr_merge_single_wm 555 10.7 0.542 0.659 2.828 3.126 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.993 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.939 2.994 dbcsr_data_release 10477 10.7 1.592 2.479 1.592 2.479 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.039 2.099 rebuild_ks_matrix 3 7.3 0.000 0.000 2.016 2.076 qs_ks_build_kohn_sham_matrix 3 8.3 0.041 0.101 2.016 2.076 dbcsr_finalize 304 7.8 0.049 0.061 1.813 1.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.726000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/25/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 15.383312E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 336818 0.0% 0.0% 100.0% average stack size 0.0 0.0 28603.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 4.798054E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.070 93.733 93.734 qs_energies 1 2.0 0.000 0.000 92.671 92.673 ls_scf 1 3.0 0.000 0.000 90.728 90.733 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.844 72.019 ls_scf_main 1 4.0 0.000 0.000 57.369 57.372 multiply_cannon 111 7.7 0.080 0.116 53.223 56.559 multiply_cannon_loop 111 8.7 0.087 0.094 50.574 52.212 density_matrix_trs4 2 5.0 0.002 0.003 50.327 50.425 ls_scf_init_scf 1 4.0 0.000 0.000 29.947 29.951 mp_waitall_1 5436 11.0 25.011 29.331 25.011 29.331 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.468 28.498 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.336 26.351 multiply_cannon_multrec 444 9.7 13.757 16.058 20.881 22.559 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.938 16.168 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.424 15.267 make_m2s 222 7.7 0.005 0.005 13.681 14.636 make_images 222 8.7 2.038 2.474 13.614 14.567 hybrid_alltoall_any 227 10.6 0.800 3.811 8.277 9.941 make_images_data 222 9.7 0.003 0.004 8.462 9.811 multiply_cannon_sync_h2d 444 9.7 6.700 8.338 6.700 8.338 dbcsr_mm_accdrv_process 3003 10.4 0.377 0.499 6.820 7.965 dbcsr_mm_accdrv_process_sort 3003 11.4 6.427 7.560 6.427 7.560 arnoldi_extremal 4 6.8 0.000 0.000 5.985 5.993 arnoldi_normal_ev 4 7.8 0.002 0.005 5.985 5.993 build_subspace 16 8.4 0.015 0.020 5.580 5.590 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.518 4.739 mp_sum_l 887 5.1 2.951 4.602 2.951 4.602 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.274 4.456 dbcsr_matrix_vector_mult_local 304 10.0 3.755 4.201 3.757 4.203 mp_allgather_i34 111 8.7 1.207 3.850 1.207 3.850 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.561 3.721 calculate_norms 792 9.8 3.601 3.705 3.601 3.705 mp_irecv_dv 1241 11.2 1.548 3.668 1.548 3.668 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.510 3.590 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.050 3.586 ls_scf_post 1 4.0 0.000 0.000 3.411 3.413 make_images_sizes 222 9.7 0.000 0.000 0.834 3.299 mp_alltoall_i44 222 10.7 0.834 3.299 0.834 3.299 ls_scf_store_result 1 5.0 0.000 0.000 3.185 3.221 dbcsr_finalize 304 7.8 0.061 0.077 2.201 2.322 dbcsr_data_new 4608 9.7 1.805 2.263 1.805 2.263 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.223 2.228 rebuild_ks_matrix 3 7.3 0.000 0.000 2.190 2.195 qs_ks_build_kohn_sham_matrix 3 8.3 0.056 0.064 2.190 2.195 dbcsr_merge_all 275 8.9 0.479 0.528 2.058 2.159 qs_energies_init_hamiltonians 1 3.0 0.002 0.014 1.928 1.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.734000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3755.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/26/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 23 x 23 x 23 234439235724792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 234.439236E+12 0.0% 0.0% 100.0% flops max/rank 30.358840E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 9634225188 0.0% 0.0% 100.0% number of processed stacks 339931 0.0% 0.0% 100.0% average stack size 0.0 0.0 28341.7 marketing flops 1.742118E+15 ------------------------------------------------------------------------------- # multiplications 111 max memory usage/rank 9.056072E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.125 106.397 106.398 qs_energies 1 2.0 0.000 0.000 104.625 104.628 ls_scf 1 3.0 0.000 0.000 101.659 101.663 dbcsr_multiply_generic 111 6.7 0.024 0.026 74.812 74.965 ls_scf_main 1 4.0 0.000 0.000 63.734 63.734 density_matrix_trs4 2 5.0 0.002 0.003 54.897 54.958 multiply_cannon 111 7.7 0.113 0.213 48.511 51.035 multiply_cannon_loop 111 8.7 0.098 0.099 45.432 45.749 ls_scf_init_scf 1 4.0 0.000 0.000 34.081 34.082 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.401 32.424 matrix_sqrt_Newton_Schulz 2 6.5 0.013 0.014 29.662 29.667 mp_waitall_1 4527 11.1 21.710 25.560 21.710 25.560 make_m2s 222 7.7 0.005 0.005 22.604 23.742 make_images 222 8.7 3.570 3.867 22.497 23.631 multiply_cannon_multrec 444 9.7 17.843 18.423 22.482 23.083 hybrid_alltoall_any 227 10.6 1.652 3.621 12.747 15.748 make_images_data 222 9.7 0.003 0.004 12.960 15.186 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.170 10.583 multiply_cannon_sync_h2d 444 9.7 8.786 8.836 8.786 8.836 arnoldi_extremal 4 6.8 0.000 0.000 7.711 7.714 arnoldi_normal_ev 4 7.8 0.011 0.017 7.711 7.714 build_subspace 16 8.4 0.026 0.036 7.158 7.171 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.694 5.836 dbcsr_matrix_vector_mult_local 304 10.0 5.177 5.454 5.179 5.457 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.909 5.163 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.952 5.044 dbcsr_mm_accdrv_process 1814 10.4 0.290 0.354 4.463 4.595 dbcsr_mm_accdrv_process_sort 1814 11.4 4.126 4.262 4.126 4.262 ls_scf_post 1 4.0 0.000 0.000 3.845 3.849 make_images_sizes 222 9.7 0.000 0.000 1.517 3.807 mp_alltoall_i44 222 10.7 1.516 3.807 1.516 3.807 mp_allgather_i34 111 8.7 1.172 3.646 1.172 3.646 ls_scf_store_result 1 5.0 0.000 0.000 3.523 3.558 calculate_norms 792 9.8 3.233 3.274 3.233 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.073 3.125 qs_energies_init_hamiltonians 1 3.0 0.021 0.029 2.935 2.935 dbcsr_merge_all 275 8.9 0.885 0.913 2.856 2.902 dbcsr_complete_redistribute 5 7.6 1.438 1.485 2.769 2.883 matrix_ls_to_qs 2 6.0 0.000 0.000 2.408 2.542 dbcsr_sort_data 325 11.1 2.434 2.501 2.434 2.501 dbcsr_new_transposed 4 7.5 0.312 0.394 2.425 2.438 dbcsr_data_new 6591 9.6 1.885 2.395 1.885 2.395 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.386 2.387 rebuild_ks_matrix 3 7.3 0.000 0.000 2.317 2.318 qs_ks_build_kohn_sham_matrix 3 8.3 0.056 0.060 2.317 2.318 dbcsr_frobenius_norm 74 6.6 2.055 2.127 2.186 2.219 dbcsr_add_d 103 6.2 0.000 0.000 2.129 2.204 dbcsr_add_anytype 103 7.2 0.858 0.890 2.128 2.204 dbcsr_data_release 12724 10.6 1.972 2.176 1.972 2.176 dbcsr_redistribute 4 8.5 1.367 1.418 2.085 2.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.398000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7199.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/41a40305d136800f944133e53e6b8ab400f7c868_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 583.921664E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 3686719. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.132 0.298 273.571 273.574 qs_mol_dyn_low 1 2.0 0.011 0.061 271.973 272.004 qs_forces 5 3.8 0.005 0.007 271.688 271.699 qs_energies 5 4.8 0.011 0.065 268.404 268.451 scf_env_do_scf 5 5.8 0.001 0.020 238.990 238.993 scf_env_do_scf_inner_loop 105 6.6 0.009 0.039 207.806 207.809 qs_scf_new_mos 105 7.6 0.000 0.001 162.268 162.493 qs_scf_loop_do_ot 105 8.6 0.001 0.009 162.267 162.493 ot_scf_mini 105 9.6 0.003 0.005 152.241 152.355 dbcsr_multiply_generic 1445 12.2 0.145 0.157 149.756 150.343 multiply_cannon 1445 13.2 0.276 0.289 128.818 130.617 multiply_cannon_loop 1445 14.2 2.848 2.999 126.678 128.543 velocity_verlet 4 3.0 0.044 0.187 115.356 115.360 ot_mini 105 10.6 0.001 0.001 62.316 62.445 qs_ot_get_p 112 10.4 0.001 0.001 59.385 59.743 qs_ot_p2m_diag 40 11.0 0.020 0.030 47.449 47.543 mp_waitall_1 488190 16.1 38.473 47.378 38.473 47.378 multiply_cannon_multrec 69360 15.2 29.683 34.236 40.454 45.228 cp_dbcsr_syevd 40 12.0 0.002 0.002 44.044 44.045 qs_ot_get_derivative 55 11.6 0.001 0.001 40.450 40.612 cp_fm_syevd 40 13.0 0.000 0.001 38.409 38.547 multiply_cannon_metrocomm3 69360 15.2 0.199 0.210 26.755 36.252 multiply_cannon_sync_h2d 69360 15.2 28.869 33.337 28.869 33.337 cp_fm_redistribute_end 40 14.0 16.622 33.163 16.628 33.167 cp_fm_syevd_base 40 14.0 16.524 33.065 16.524 33.065 rebuild_ks_matrix 110 8.4 0.000 0.000 30.971 31.241 qs_ks_build_kohn_sham_matrix 110 9.4 0.052 0.230 30.971 31.241 init_scf_loop 7 6.6 0.005 0.088 31.131 31.135 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.452 28.692 init_scf_run 5 5.8 0.000 0.001 26.478 26.479 scf_env_initial_rho_setup 5 6.8 0.001 0.003 26.478 26.479 prepare_preconditioner 7 7.6 0.000 0.000 25.779 25.844 make_preconditioner 7 8.6 0.004 0.107 25.779 25.844 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.098 23.425 apply_single 62 13.6 0.000 0.000 23.098 23.425 qs_rho_update_rho_low 110 7.6 0.000 0.001 21.124 21.530 calculate_rho_elec 110 8.6 0.030 0.057 21.123 21.529 calculate_first_density_matrix 1 7.0 0.003 0.063 21.096 21.132 ot_new_cg_direction 55 11.6 0.001 0.001 21.115 21.116 acc_transpose_blocks 69360 15.2 0.340 0.358 19.384 20.372 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 19.945 19.962 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 19.393 19.444 acc_transpose_blocks_kernels 69360 16.2 0.845 0.889 16.642 17.541 make_full_inverse_cholesky 7 9.6 0.000 0.000 17.148 17.213 jit_kernel_transpose 5 15.0 15.797 16.684 15.797 16.684 density_rs2pw 110 9.6 0.004 0.005 15.081 15.520 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.227 15.467 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.538 14.634 mp_sum_l 4764 12.2 12.945 13.873 12.945 13.873 multiply_cannon_metrocomm1 69360 15.2 0.094 0.101 5.275 12.542 calculate_dm_sparse 110 9.5 0.000 0.001 12.094 12.332 transfer_rs2pw 445 10.6 0.008 0.009 11.842 12.237 cp_fm_cholesky_invert 7 10.6 11.672 11.688 11.672 11.688 dbcsr_mm_accdrv_process 154766 15.8 6.231 6.399 10.640 11.490 fft_wrap_pw1pw2 1425 12.5 0.020 0.029 11.073 11.404 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.893 10.953 sum_up_and_integrate 60 10.3 0.002 0.041 10.537 10.550 integrate_v_rspace 60 11.3 0.002 0.002 10.518 10.532 check_diag 80 13.5 8.685 8.935 10.341 10.478 qs_vxc_create 110 10.4 0.002 0.002 10.027 10.054 fft_wrap_pw1pw2_240 915 14.0 0.208 0.236 9.573 9.842 fft3d_pb 915 15.0 2.277 2.533 8.512 8.815 make_full_single_inverse 7 9.6 0.003 0.026 8.320 8.388 make_m2s 2890 13.2 0.077 0.085 7.588 8.112 make_images 2890 14.2 0.239 0.259 7.481 8.004 xc_rho_set_and_dset_create 110 12.4 0.077 0.092 7.048 7.317 xc_vxc_pw_create 60 11.3 0.036 0.041 6.738 6.770 potential_pw2rs 60 12.3 0.002 0.003 6.587 6.624 transfer_rs2pw_30 110 11.6 1.254 1.314 5.708 6.190 mp_sendrecv_dv 168740 12.6 5.884 6.031 5.884 6.031 mp_alltoall_z22v 2340 16.7 5.655 5.961 5.655 5.961 xc_pw_derive 510 13.4 0.005 0.006 5.816 5.904 transfer_pw2rs 245 13.2 0.003 0.004 5.495 5.526 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=273.574000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.200000, yerr=1.600000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 41a40305d136800f944133e53e6b8ab400f7c868 Summary: empty Status: OK