=== 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: e112afaa9cdfdd1e40549c16c19aedffb6738e2c ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # 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: 21.06.2023 # # \ 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-libvdwxc --with-pexsi --with-plumed; \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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_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_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)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib 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__HDF5 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 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/01 job id: 47558785 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/02 job id: 47558786 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/03 job id: 47558789 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/04 job id: 47558790 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/05 job id: 47558791 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/06 job id: 47558792 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/07 job id: 47558793 --- 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] = 5 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/08 job id: 47558794 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/09 job id: 47558795 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/10 job id: 47558796 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/11 job id: 47558797 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/12 job id: 47558798 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/13 job id: 47558800 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/14 job id: 47558803 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/15 job id: 47558804 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/16 job id: 47558805 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/17 job id: 47558806 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/18 job id: 47558807 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/19 job id: 47558809 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/20 job id: 47558810 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/21 job id: 47558811 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/22 job id: 47558812 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/23 job id: 47558813 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/24 job id: 47558817 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/25 job id: 47558818 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/26 job id: 47558819 --- 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] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/27 job id: 47558821 --- 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/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 177869. 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 0.021 0.036 134.141 134.142 farming_run 1 2.0 133.619 133.622 134.114 134.117 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458323E+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 0.006 0.025 114.817 114.817 qs_energies 1 2.0 0.000 0.000 114.629 114.631 mp2_main 1 3.0 0.000 0.000 112.595 112.596 mp2_gpw_main 1 4.0 0.019 0.026 111.610 111.611 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.660 93.171 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.884 55.394 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 41.340 46.664 get_2c_integrals 1 6.0 0.008 0.009 37.052 37.603 integrate_v_rspace 273 8.0 0.437 0.451 24.724 29.740 pw_transfer 6555 10.6 0.375 0.388 27.473 28.059 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.107 26.574 grid_integrate_task_list 273 9.0 20.534 26.051 20.534 26.051 fft_wrap_pw1pw2_100 2178 12.4 1.225 1.483 23.637 24.107 compute_2c_integrals 1 7.0 0.002 0.002 19.789 19.790 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.967 19.528 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.440 18.964 19.524 rpa_ri_compute_en 1 5.0 0.019 0.021 18.840 18.980 cp_fm_cholesky_decompose 12 8.2 17.195 17.746 17.195 17.746 cholesky_decomp 1 7.0 0.000 0.000 16.112 16.667 fft3d_s 5443 13.4 16.184 16.596 16.206 16.618 ao_to_mo_and_store_B_mult_1 272 7.0 10.718 15.290 10.718 15.290 calculate_wavefunction 272 8.0 5.403 5.551 12.539 13.152 rpa_num_int 1 6.0 0.000 0.000 10.615 10.615 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.593 10.599 calc_mat_Q 8 8.0 0.000 0.000 9.462 9.549 contract_S_to_Q 8 9.0 0.000 0.000 8.884 8.970 calc_potential_gpw 544 9.5 0.005 0.006 8.266 8.704 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.256 8.604 parallel_gemm_fm 14 9.1 0.000 0.000 8.470 8.567 parallel_gemm_fm_cosma 14 10.1 8.470 8.567 8.470 8.567 potential_pw2rs 545 10.0 0.107 0.110 7.718 8.368 collocate_single_gaussian 272 10.0 0.039 0.042 7.515 7.821 create_integ_mat 1 6.0 0.022 0.026 7.621 7.621 array2fm 1 7.0 0.000 0.000 6.642 7.068 pw_scatter_s 2720 13.7 4.424 4.559 4.424 4.559 pw_gather_s 2722 13.2 3.872 4.133 3.872 4.133 array2fm_buffer_send 1 8.0 2.962 3.092 2.962 3.092 pw_poisson_solve 545 10.5 1.107 1.155 2.177 2.367 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.608401, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2800.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 205321. 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 0.027 0.038 396.304 396.305 farming_run 1 2.0 395.528 395.536 396.270 396.273 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.229206E+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 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 0.009 0.029 209.153 209.153 qs_energies 1 2.0 0.000 0.000 208.895 208.911 scf_env_do_scf 1 3.0 0.000 0.000 106.080 106.080 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.219 105.227 rebuild_ks_matrix 4 6.0 0.000 0.000 105.218 105.226 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.218 105.226 hfx_ks_matrix 4 8.0 0.001 0.001 104.840 104.844 integrate_four_center 4 9.0 0.143 0.462 104.839 104.843 mp2_main 1 3.0 0.000 0.000 102.527 102.543 mp2_gpw_main 1 4.0 0.032 0.047 101.691 101.709 integrate_four_center_main 4 10.0 0.075 0.401 96.583 98.541 integrate_four_center_bin 265 11.0 96.507 98.488 96.507 98.488 init_scf_loop 1 4.0 0.000 0.000 91.740 91.740 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.585 75.680 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.066 55.160 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.165 41.672 46.749 integrate_v_rspace 95 8.0 0.399 0.579 28.054 33.125 pw_transfer 2240 10.6 0.144 0.183 29.913 30.592 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.914 29.637 ao_to_mo_and_store_B_mult_1 91 7.0 10.719 29.178 10.719 29.178 mp2_ri_gpw_compute_en 1 5.0 0.055 0.075 26.958 28.838 grid_integrate_task_list 95 9.0 23.351 28.614 23.351 28.614 fft_wrap_pw1pw2_100 730 12.4 1.297 1.530 26.620 27.249 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.897 25.219 25.229 get_2c_integrals 1 6.0 0.000 0.000 20.436 20.455 compute_2c_integrals 1 7.0 0.002 0.003 19.411 19.413 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.901 19.291 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.874 18.900 19.290 fft3d_s 1823 13.4 18.441 18.734 18.454 18.747 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.338 14.338 calculate_wavefunction 91 8.0 2.016 2.048 9.734 9.923 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.584 8.771 9.476 potential_pw2rs 186 10.0 0.034 0.036 8.637 9.203 local_gemm 172 8.0 8.212 8.903 8.212 8.903 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.516 8.219 8.692 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.273 8.598 calc_potential_gpw 182 9.5 0.002 0.002 7.932 8.295 collocate_single_gaussian 91 10.0 0.017 0.022 7.893 8.128 mp_sendrecv_dm3 2068 8.0 6.244 6.713 6.244 6.713 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.421 6.348 6.421 mp_sync 37 10.5 2.899 5.620 2.899 5.620 pw_gather_s 912 13.2 4.869 5.522 4.869 5.522 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.682879, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1516.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 451.743744E+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 588782. 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 0.010 0.026 53.559 53.560 qs_mol_dyn_low 1 2.0 0.003 0.003 52.976 53.356 qs_forces 11 3.9 0.002 0.002 52.911 52.913 qs_energies 11 4.9 0.001 0.004 51.446 51.459 scf_env_do_scf 11 5.9 0.000 0.001 44.558 44.559 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.581 42.581 dbcsr_multiply_generic 2286 12.5 0.094 0.099 33.368 33.787 qs_scf_new_mos 108 7.5 0.000 0.001 32.665 32.949 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.664 32.948 ot_scf_mini 108 9.5 0.002 0.002 31.057 31.254 multiply_cannon 2286 13.5 0.184 0.189 26.583 27.838 multiply_cannon_loop 2286 14.5 1.480 1.582 25.964 27.199 velocity_verlet 10 3.0 0.001 0.002 25.070 25.070 ot_mini 108 10.5 0.001 0.001 18.699 18.942 qs_ot_get_derivative 108 11.5 0.001 0.001 15.924 16.119 mp_waitall_1 245248 16.5 7.874 14.656 7.874 14.656 multiply_cannon_metrocomm3 54864 15.5 0.069 0.073 5.748 12.432 multiply_cannon_multrec 54864 15.5 4.314 6.809 8.704 12.315 qs_ot_get_p 119 10.4 0.001 0.001 7.844 8.129 rebuild_ks_matrix 119 8.3 0.000 0.000 7.751 7.890 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.750 7.889 multiply_cannon_sync_h2d 54864 15.5 6.018 7.195 6.018 7.195 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.825 6.944 mp_sum_l 7287 12.8 4.978 6.728 4.978 6.728 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.489 5.925 dbcsr_mm_accdrv_process 76910 16.1 1.129 1.801 4.311 5.772 init_scf_run 11 5.9 0.000 0.001 5.589 5.589 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.589 5.589 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.291 5.334 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.997 5.094 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.476 4.476 sum_up_and_integrate 119 10.3 0.012 0.014 4.419 4.426 integrate_v_rspace 119 11.3 0.002 0.003 4.407 4.416 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.291 4.291 cp_fm_redistribute_end 50 14.0 2.190 4.269 2.196 4.272 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.128 4.216 calculate_rho_elec 119 8.7 0.012 0.017 4.127 4.215 cp_fm_diag_elpa_base 50 14.0 2.071 4.159 2.074 4.167 jit_kernel_multiply 13 15.8 3.122 3.885 3.122 3.885 calculate_dm_sparse 119 9.5 0.000 0.001 3.659 3.797 calculate_first_density_matrix 1 7.0 0.000 0.003 3.422 3.427 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.860 3.046 apply_single 119 13.6 0.000 0.000 2.860 3.046 rs_pw_transfer 974 11.9 0.012 0.013 2.658 2.759 ot_diis_step 108 11.5 0.006 0.006 2.565 2.565 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.525 2.488 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.325 2.372 density_rs2pw 119 9.7 0.004 0.005 2.124 2.234 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.208 2.232 acc_transpose_blocks 54864 15.5 0.228 0.254 1.740 2.222 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.165 2.167 grid_integrate_task_list 119 12.3 2.021 2.111 2.021 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.094 2.094 init_scf_loop 11 6.9 0.000 0.001 1.962 1.962 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.902 1.961 potential_pw2rs 119 12.3 0.004 0.004 1.779 1.786 mp_sum_d 4135 12.0 1.138 1.719 1.138 1.719 pw_transfer 1439 11.6 0.052 0.057 1.644 1.709 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.568 1.636 make_m2s 4572 13.5 0.053 0.054 1.494 1.536 make_images 4572 14.5 0.132 0.138 1.413 1.454 mp_waitany 12084 13.8 1.252 1.449 1.252 1.449 grid_collocate_task_list 119 9.7 1.354 1.428 1.354 1.428 mp_alltoall_d11v 2130 13.8 1.173 1.361 1.173 1.361 fft_wrap_pw1pw2_140 487 13.2 0.189 0.206 1.213 1.282 fft3d_ps 1201 14.6 0.367 0.471 1.206 1.271 parallel_gemm_fm 81 9.0 0.000 0.000 1.107 1.111 parallel_gemm_fm_cosma 81 10.0 1.107 1.111 1.107 1.111 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.560000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 485.498880E+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 1773715. 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 0.009 0.026 39.753 39.754 qs_mol_dyn_low 1 2.0 0.003 0.003 39.585 39.593 qs_forces 11 3.9 0.002 0.003 39.525 39.526 qs_energies 11 4.9 0.001 0.001 37.762 37.766 scf_env_do_scf 11 5.9 0.000 0.001 31.284 31.285 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.901 28.901 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.169 22.525 qs_scf_new_mos 108 7.5 0.001 0.001 19.862 20.113 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.862 20.112 ot_scf_mini 108 9.5 0.002 0.003 18.978 19.156 multiply_cannon 2286 13.5 0.207 0.218 17.316 18.918 velocity_verlet 10 3.0 0.001 0.002 18.003 18.004 multiply_cannon_loop 2286 14.5 0.900 0.962 16.233 17.608 ot_mini 108 10.5 0.001 0.001 11.696 11.934 mp_waitall_1 200699 16.5 5.571 10.868 5.571 10.868 multiply_cannon_multrec 27432 15.5 2.001 4.409 6.870 9.734 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.155 9.659 qs_ot_get_derivative 108 11.5 0.001 0.001 9.232 9.414 rebuild_ks_matrix 119 8.3 0.000 0.000 7.087 7.221 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.087 7.221 dbcsr_mm_accdrv_process 47894 16.0 3.633 6.400 4.799 6.660 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.240 6.362 init_scf_run 11 5.9 0.000 0.001 5.092 5.093 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.092 5.092 qs_ot_get_p 119 10.4 0.001 0.001 4.393 4.628 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.599 4.457 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.064 4.142 apply_single 119 13.6 0.000 0.000 3.064 4.142 sum_up_and_integrate 119 10.3 0.025 0.028 4.094 4.101 integrate_v_rspace 119 11.3 0.002 0.003 4.069 4.080 mp_sum_l 7287 12.8 1.995 3.958 1.995 3.958 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.773 3.803 calculate_rho_elec 119 8.7 0.021 0.024 3.773 3.802 calculate_first_density_matrix 1 7.0 0.000 0.000 3.566 3.567 calculate_dm_sparse 119 9.5 0.000 0.001 2.974 3.048 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.010 3.029 jit_kernel_multiply 10 16.1 1.114 2.942 1.114 2.942 multiply_cannon_sync_h2d 27432 15.5 2.196 2.923 2.196 2.923 rs_pw_transfer 974 11.9 0.010 0.011 2.467 2.916 make_m2s 4572 13.5 0.052 0.053 2.404 2.640 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.584 2.584 make_images 4572 14.5 0.200 0.236 2.315 2.550 density_rs2pw 119 9.7 0.004 0.004 2.058 2.523 ot_diis_step 108 11.5 0.011 0.011 2.415 2.415 init_scf_loop 11 6.9 0.000 0.000 2.364 2.364 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.253 2.253 cp_fm_redistribute_end 50 14.0 1.142 2.228 1.145 2.231 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.196 2.204 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.088 2.181 cp_fm_diag_elpa_base 50 14.0 1.054 2.136 1.082 2.172 grid_integrate_task_list 119 12.3 1.834 1.961 1.834 1.961 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.925 1.927 pw_transfer 1439 11.6 0.065 0.070 1.881 1.909 potential_pw2rs 119 12.3 0.006 0.006 1.819 1.827 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.789 1.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.724 1.763 make_images_data 4572 15.5 0.045 0.051 1.130 1.532 acc_transpose_blocks 27432 15.5 0.110 0.115 1.212 1.508 prepare_preconditioner 11 7.9 0.000 0.000 1.481 1.508 make_preconditioner 11 8.9 0.000 0.000 1.481 1.508 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.985 1.468 wfi_extrapolate 11 7.9 0.001 0.001 1.467 1.467 fft_wrap_pw1pw2_140 487 13.2 0.203 0.214 1.420 1.451 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.387 1.445 grid_collocate_task_list 119 9.7 1.273 1.415 1.273 1.415 fft3d_ps 1201 14.6 0.517 0.571 1.339 1.364 mp_allgather_i34 2286 14.5 0.526 1.296 0.526 1.296 mp_alltoall_d11v 2130 13.8 1.155 1.274 1.155 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.172 1.223 mp_sum_d 4135 12.0 0.614 1.064 0.614 1.064 rs_pw_transfer_RS2PW_140 130 11.5 0.141 0.149 0.540 0.989 mp_waitany 5720 13.7 0.515 0.962 0.515 0.962 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.943 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.917 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.663 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.009 0.071 0.707 0.797 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.754000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=462.363636, yerr=1.610913 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 519.041024E+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 1426811. 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 0.010 0.026 33.074 33.074 qs_mol_dyn_low 1 2.0 0.003 0.003 32.810 32.818 qs_forces 11 3.9 0.002 0.002 32.741 32.743 qs_energies 11 4.9 0.001 0.001 31.180 31.183 scf_env_do_scf 11 5.9 0.000 0.001 25.210 25.210 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.757 22.757 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.228 17.325 velocity_verlet 10 3.0 0.001 0.002 15.029 15.031 multiply_cannon 2286 13.5 0.195 0.204 14.107 14.870 qs_scf_new_mos 108 7.5 0.001 0.001 14.649 14.665 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.649 14.664 multiply_cannon_loop 2286 14.5 0.632 0.663 13.344 14.119 ot_scf_mini 108 9.5 0.002 0.002 13.937 13.956 ot_mini 108 10.5 0.001 0.001 8.594 8.616 multiply_cannon_multrec 18288 15.5 1.954 2.814 7.801 8.039 qs_ot_get_derivative 108 11.5 0.001 0.001 7.118 7.137 dbcsr_mm_accdrv_process 38222 16.0 4.501 6.412 5.761 6.632 rebuild_ks_matrix 119 8.3 0.000 0.000 6.314 6.337 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.313 6.336 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.565 5.586 init_scf_run 11 5.9 0.000 0.001 4.801 4.801 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.801 4.801 sum_up_and_integrate 119 10.3 0.032 0.032 3.907 3.911 integrate_v_rspace 119 11.3 0.002 0.003 3.875 3.884 mp_waitall_1 158411 16.6 2.533 3.709 2.533 3.709 calculate_first_density_matrix 1 7.0 0.000 0.000 3.589 3.589 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.422 3.430 calculate_rho_elec 119 8.7 0.031 0.031 3.421 3.429 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.709 3.316 qs_ot_get_p 119 10.4 0.001 0.001 3.201 3.227 jit_kernel_multiply 10 15.8 1.211 2.897 1.211 2.897 calculate_dm_sparse 119 9.5 0.000 0.000 2.786 2.799 rs_pw_transfer 974 11.9 0.009 0.010 2.210 2.510 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.435 2.453 init_scf_loop 11 6.9 0.000 0.000 2.436 2.437 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.015 2.338 apply_single 119 13.6 0.000 0.000 2.015 2.338 density_rs2pw 119 9.7 0.004 0.004 1.914 2.205 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.169 2.174 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.165 2.170 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.911 1.912 grid_integrate_task_list 119 12.3 1.798 1.897 1.798 1.897 pw_transfer 1439 11.6 0.066 0.069 1.869 1.875 make_m2s 4572 13.5 0.044 0.045 1.712 1.857 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.775 1.784 make_images 4572 14.5 0.191 0.204 1.628 1.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.758 1.760 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.671 1.681 cp_fm_diag_elpa_base 50 14.0 1.649 1.663 1.669 1.680 potential_pw2rs 119 12.3 0.007 0.008 1.669 1.675 prepare_preconditioner 11 7.9 0.000 0.000 1.664 1.668 make_preconditioner 11 8.9 0.000 0.000 1.664 1.668 multiply_cannon_sync_h2d 18288 15.5 1.411 1.612 1.411 1.612 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.598 1.604 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.522 1.603 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.516 1.528 ot_diis_step 108 11.5 0.011 0.011 1.458 1.458 fft_wrap_pw1pw2_140 487 13.2 0.256 0.264 1.444 1.454 mp_sum_l 7287 12.8 1.017 1.407 1.017 1.407 grid_collocate_task_list 119 9.7 1.242 1.400 1.242 1.400 acc_transpose_blocks 18288 15.5 0.076 0.078 1.285 1.307 fft3d_ps 1201 14.6 0.528 0.544 1.261 1.271 wfi_extrapolate 11 7.9 0.001 0.001 1.158 1.158 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.955 0.956 make_images_data 4572 15.5 0.045 0.048 0.774 0.931 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.859 0.882 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.674 0.859 acc_transpose_blocks_kernels 18288 16.5 0.210 0.218 0.836 0.852 mp_waitany 9880 13.7 0.529 0.839 0.529 0.839 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.124 0.530 0.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.820 mp_alltoall_d11v 2130 13.8 0.644 0.814 0.644 0.814 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.642 0.727 cp_fm_cholesky_invert 11 10.9 0.691 0.694 0.691 0.694 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.074000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=494.636364, yerr=0.642824 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 555.560960E+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 0.011 0.029 38.539 38.539 qs_mol_dyn_low 1 2.0 0.003 0.004 38.327 38.335 qs_forces 11 3.9 0.002 0.002 38.269 38.270 qs_energies 11 4.9 0.001 0.002 36.514 36.521 scf_env_do_scf 11 5.9 0.000 0.001 29.114 29.115 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.531 25.532 dbcsr_multiply_generic 2286 12.5 0.099 0.103 20.229 20.416 velocity_verlet 10 3.0 0.001 0.002 18.484 18.486 multiply_cannon 2286 13.5 0.221 0.231 16.373 16.796 qs_scf_new_mos 108 7.5 0.001 0.001 16.558 16.615 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.558 16.614 multiply_cannon_loop 2286 14.5 0.942 0.970 15.358 15.671 ot_scf_mini 108 9.5 0.002 0.003 15.613 15.661 multiply_cannon_multrec 27432 15.5 2.367 3.086 9.641 10.309 ot_mini 108 10.5 0.001 0.001 9.393 9.457 dbcsr_mm_accdrv_process 47916 15.9 5.875 7.967 7.183 8.424 qs_ot_get_derivative 108 11.5 0.001 0.001 7.539 7.594 rebuild_ks_matrix 119 8.3 0.000 0.000 6.893 6.944 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.892 6.943 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.127 6.170 init_scf_run 11 5.9 0.000 0.001 6.063 6.064 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.063 6.063 calculate_first_density_matrix 1 7.0 0.000 0.000 4.600 4.629 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.866 4.492 calculate_rho_elec 119 8.7 0.040 0.046 3.865 4.491 sum_up_and_integrate 119 10.3 0.036 0.038 4.087 4.096 integrate_v_rspace 119 11.3 0.003 0.003 4.051 4.061 mp_waitall_1 137007 16.6 2.904 3.896 2.904 3.896 calculate_dm_sparse 119 9.5 0.000 0.000 3.669 3.756 qs_ot_get_p 119 10.4 0.001 0.001 3.546 3.612 init_scf_loop 11 6.9 0.000 0.000 3.562 3.564 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.821 3.218 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.844 2.880 jit_kernel_multiply 9 15.7 1.248 2.750 1.248 2.750 prepare_preconditioner 11 7.9 0.000 0.000 2.680 2.688 make_preconditioner 11 8.9 0.000 0.000 2.680 2.688 rs_pw_transfer 974 11.9 0.009 0.010 2.482 2.681 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.284 2.613 make_m2s 4572 13.5 0.054 0.055 2.414 2.558 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.545 2.546 make_images 4572 14.5 0.271 0.332 2.307 2.451 density_rs2pw 119 9.7 0.004 0.004 2.214 2.407 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.240 2.250 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 1.210 2.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.066 2.198 apply_single 119 13.6 0.000 0.000 2.066 2.198 pw_transfer 1439 11.6 0.066 0.071 2.073 2.116 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.981 2.028 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.985 2.010 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.960 1.970 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.914 1.914 grid_integrate_task_list 119 12.3 1.822 1.910 1.822 1.910 potential_pw2rs 119 12.3 0.008 0.010 1.823 1.835 ot_diis_step 108 11.5 0.012 0.012 1.809 1.809 fft_wrap_pw1pw2_140 487 13.2 0.289 0.301 1.615 1.665 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.649 1.659 acc_transpose_blocks 27432 15.5 0.110 0.112 1.487 1.658 cp_fm_diag_elpa_base 50 14.0 1.613 1.630 1.645 1.654 mp_alltoall_d11v 2130 13.8 0.785 1.581 0.785 1.581 dbcsr_complete_redistribute 329 12.2 0.530 0.556 1.197 1.472 fft3d_ps 1201 14.6 0.556 0.608 1.419 1.450 wfi_extrapolate 11 7.9 0.001 0.001 1.377 1.377 grid_collocate_task_list 119 9.7 1.250 1.363 1.250 1.363 mp_sum_l 7287 12.8 1.014 1.319 1.014 1.319 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.984 1.250 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.147 1.162 cp_fm_upper_to_full 72 14.2 0.812 1.157 0.812 1.157 make_images_data 4572 15.5 0.044 0.048 0.939 1.151 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.078 1.080 multiply_cannon_sync_h2d 27432 15.5 0.997 1.062 0.997 1.062 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.793 1.042 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.548 1.030 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.848 1.013 rs_scatter_matrices 130 9.7 0.116 0.128 0.358 0.985 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.855 0.860 cp_fm_cholesky_invert 11 10.9 0.835 0.838 0.835 0.838 mp_alltoall_z22v 1201 16.6 0.746 0.777 0.746 0.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.539000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.818182, yerr=3.537287 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 605.687808E+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 0.012 0.037 28.705 28.706 qs_mol_dyn_low 1 2.0 0.003 0.004 28.526 28.534 qs_forces 11 3.9 0.002 0.002 28.464 28.464 qs_energies 11 4.9 0.001 0.001 26.749 26.751 scf_env_do_scf 11 5.9 0.000 0.001 20.876 20.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.446 18.447 velocity_verlet 10 3.0 0.002 0.002 14.144 14.147 dbcsr_multiply_generic 2286 12.5 0.092 0.097 12.751 12.829 multiply_cannon 2286 13.5 0.230 0.238 10.340 10.787 qs_scf_new_mos 108 7.5 0.001 0.001 10.582 10.603 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.582 10.602 ot_scf_mini 108 9.5 0.002 0.002 9.923 9.955 multiply_cannon_loop 2286 14.5 0.329 0.340 9.459 9.657 multiply_cannon_multrec 9144 15.5 1.600 1.845 6.089 6.426 rebuild_ks_matrix 119 8.3 0.000 0.000 5.931 5.952 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.931 5.952 ot_mini 108 10.5 0.001 0.001 5.561 5.600 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.274 5.292 dbcsr_mm_accdrv_process 12550 15.8 3.585 4.452 4.389 4.495 init_scf_run 11 5.9 0.000 0.001 4.436 4.436 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.436 4.436 qs_ot_get_derivative 108 11.5 0.001 0.001 4.293 4.325 sum_up_and_integrate 119 10.3 0.038 0.041 3.638 3.643 integrate_v_rspace 119 11.3 0.003 0.003 3.600 3.605 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.469 3.476 calculate_rho_elec 119 8.7 0.060 0.061 3.468 3.475 calculate_first_density_matrix 1 7.0 0.000 0.000 3.295 3.297 qs_ot_get_p 119 10.4 0.001 0.001 2.624 2.666 init_scf_loop 11 6.9 0.000 0.000 2.410 2.412 calculate_dm_sparse 119 9.5 0.000 0.000 2.045 2.063 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.053 2.054 pw_transfer 1439 11.6 0.066 0.069 2.029 2.038 jit_kernel_multiply 9 16.0 0.767 1.977 0.767 1.977 grid_integrate_task_list 119 12.3 1.845 1.975 1.845 1.975 density_rs2pw 119 9.7 0.004 0.004 1.819 1.956 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.935 1.945 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.894 1.903 mp_waitall_1 115863 16.7 1.387 1.890 1.387 1.890 make_m2s 4572 13.5 0.034 0.036 1.659 1.819 rs_pw_transfer 974 11.9 0.008 0.009 1.647 1.765 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.737 1.739 make_images 4572 14.5 0.268 0.299 1.570 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.676 1.681 make_preconditioner 11 8.9 0.000 0.000 1.676 1.681 fft_wrap_pw1pw2_140 487 13.2 0.366 0.375 1.591 1.602 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.564 1.591 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.546 1.546 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.465 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.478 1.484 grid_collocate_task_list 119 9.7 1.296 1.412 1.296 1.412 acc_transpose_blocks 9144 15.5 0.038 0.041 1.383 1.392 potential_pw2rs 119 12.3 0.010 0.011 1.366 1.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.321 1.330 cp_fm_diag_elpa_base 50 14.0 1.295 1.312 1.320 1.328 fft3d_ps 1201 14.6 0.559 0.570 1.266 1.275 ot_diis_step 108 11.5 0.012 0.013 1.256 1.257 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.229 1.229 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.138 1.159 apply_single 119 13.6 0.000 0.000 1.138 1.159 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.138 1.147 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.123 1.140 wfi_extrapolate 11 7.9 0.001 0.001 1.088 1.088 jit_kernel_transpose 5 15.6 1.022 1.032 1.022 1.032 make_images_data 4572 15.5 0.039 0.042 0.766 0.952 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.704 0.946 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.872 0.921 cp_fm_cholesky_invert 11 10.9 0.833 0.835 0.833 0.835 multiply_cannon_sync_h2d 9144 15.5 0.710 0.798 0.710 0.798 mp_alltoall_d11v 2130 13.8 0.648 0.794 0.648 0.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.779 0.782 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.676 0.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.702 0.711 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.362 0.684 parallel_gemm_fm 81 9.0 0.000 0.000 0.643 0.644 parallel_gemm_fm_cosma 81 10.0 0.642 0.644 0.642 0.644 mp_alltoall_z22v 1201 16.6 0.582 0.615 0.582 0.615 make_basis_sm 11 9.8 0.000 0.000 0.606 0.607 mp_allgather_i34 2286 14.5 0.215 0.605 0.215 0.605 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.706000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.727273, yerr=6.916264 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 754.786304E+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 42.802 42.802 qs_mol_dyn_low 1 2.0 0.003 0.004 42.559 42.567 qs_forces 11 3.9 0.002 0.002 42.497 42.498 qs_energies 11 4.9 0.003 0.003 40.489 40.492 scf_env_do_scf 11 5.9 0.001 0.001 33.511 33.511 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.707 25.708 velocity_verlet 10 3.0 0.002 0.002 23.160 23.165 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.809 19.036 qs_scf_new_mos 108 7.5 0.001 0.001 16.065 16.166 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.064 16.165 multiply_cannon 2286 13.5 0.299 0.309 15.077 16.051 ot_scf_mini 108 9.5 0.002 0.002 14.995 15.097 multiply_cannon_loop 2286 14.5 0.343 0.350 13.829 14.836 multiply_cannon_multrec 9144 15.5 3.443 5.015 9.631 9.959 ot_mini 108 10.5 0.001 0.001 9.193 9.312 init_scf_loop 11 6.9 0.000 0.000 7.777 7.778 dbcsr_mm_accdrv_process 12550 15.8 5.112 7.378 6.065 7.678 rebuild_ks_matrix 119 8.3 0.000 0.000 7.219 7.363 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.218 7.362 qs_ot_get_derivative 108 11.5 0.001 0.001 7.173 7.276 prepare_preconditioner 11 7.9 0.000 0.000 6.823 6.836 make_preconditioner 11 8.9 0.000 0.000 6.823 6.836 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.343 6.715 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.512 6.642 init_scf_run 11 5.9 0.000 0.001 4.929 4.929 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.929 4.929 cp_fm_upper_to_full 72 14.2 3.195 4.619 3.195 4.619 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.190 4.195 calculate_rho_elec 119 8.7 0.118 0.121 4.189 4.195 sum_up_and_integrate 119 10.3 0.064 0.066 3.951 3.957 integrate_v_rspace 119 11.3 0.003 0.004 3.887 3.892 mp_waitall_1 94719 16.7 2.632 3.767 2.632 3.767 calculate_first_density_matrix 1 7.0 0.000 0.000 3.500 3.501 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.943 3.391 qs_ot_get_p 119 10.4 0.001 0.001 3.089 3.223 calculate_dm_sparse 119 9.5 0.000 0.000 3.178 3.196 dbcsr_complete_redistribute 329 12.2 0.291 0.293 2.011 2.879 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.214 2.705 apply_single 119 13.6 0.000 0.000 2.214 2.705 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.588 2.696 pw_transfer 1439 11.6 0.069 0.069 2.624 2.628 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.707 2.557 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.525 2.529 make_m2s 4572 13.5 0.037 0.038 2.189 2.364 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.474 2.316 mp_alltoall_i22 627 13.8 1.447 2.296 1.447 2.296 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.264 2.271 make_images 4572 14.5 0.350 0.379 2.069 2.244 fft_wrap_pw1pw2_140 487 13.2 0.618 0.622 2.134 2.140 density_rs2pw 119 9.7 0.004 0.004 2.118 2.130 grid_integrate_task_list 119 12.3 2.047 2.068 2.047 2.068 ot_diis_step 108 11.5 0.014 0.014 1.995 1.995 mp_sum_l 7287 12.8 0.998 1.913 0.998 1.913 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.876 1.878 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.860 1.862 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.775 1.776 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.652 1.706 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.619 1.659 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.607 1.608 fft3d_ps 1201 14.6 0.592 0.602 1.540 1.542 grid_collocate_task_list 119 9.7 1.497 1.504 1.497 1.504 rs_pw_transfer 974 11.9 0.009 0.009 1.431 1.443 cp_fm_cholesky_invert 11 10.9 1.410 1.413 1.410 1.413 potential_pw2rs 119 12.3 0.014 0.014 1.363 1.365 wfi_extrapolate 11 7.9 0.001 0.001 1.362 1.363 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.331 1.332 cp_fm_diag_elpa_base 50 14.0 1.183 1.238 1.329 1.329 jit_kernel_multiply 6 15.6 0.927 1.324 0.927 1.324 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.062 1.311 make_images_data 4572 15.5 0.042 0.045 1.018 1.233 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.123 1.140 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.100 1.117 acc_transpose_blocks 9144 15.5 0.038 0.039 1.000 1.076 mp_alltoall_d11v 2130 13.8 1.023 1.058 1.023 1.058 multiply_cannon_sync_h2d 9144 15.5 1.045 1.052 1.045 1.052 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.993 1.048 qs_create_task_list 11 7.9 0.000 0.001 0.931 0.944 generate_qs_task_list 11 8.9 0.368 0.388 0.931 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.918 0.932 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.802000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.909091, yerr=11.595097 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 501.014528E+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 2088483. 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 0.011 0.027 84.917 84.918 qs_mol_dyn_low 1 2.0 0.003 0.003 84.309 84.318 qs_forces 11 3.9 0.003 0.003 84.239 84.240 qs_energies 11 4.9 0.002 0.002 81.374 81.389 scf_env_do_scf 11 5.9 0.000 0.001 69.044 69.046 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.447 63.447 dbcsr_multiply_generic 2055 12.4 0.106 0.108 53.544 53.804 multiply_cannon 2055 13.4 0.181 0.187 45.542 46.423 qs_scf_new_mos 99 7.5 0.000 0.001 46.139 46.229 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.138 46.229 multiply_cannon_loop 2055 14.4 1.520 1.555 44.613 45.443 ot_scf_mini 99 9.5 0.002 0.002 43.795 43.869 velocity_verlet 10 3.0 0.001 0.002 42.066 42.067 ot_mini 99 10.5 0.001 0.001 25.738 25.824 multiply_cannon_multrec 49320 15.4 12.562 13.345 19.674 20.424 qs_ot_get_derivative 99 11.5 0.001 0.001 18.989 19.074 rebuild_ks_matrix 110 8.3 0.000 0.000 14.768 14.831 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.767 14.830 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.955 13.009 multiply_cannon_sync_h2d 49320 15.4 10.413 10.939 10.413 10.939 mp_waitall_1 220248 16.4 10.238 10.937 10.238 10.937 init_scf_run 11 5.9 0.000 0.001 10.296 10.296 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.296 10.296 qs_ot_get_p 110 10.4 0.001 0.001 9.660 9.783 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.189 7.747 apply_single 110 13.6 0.000 0.001 7.188 7.747 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.111 7.620 dbcsr_mm_accdrv_process 87628 16.1 1.908 1.972 6.986 7.244 sum_up_and_integrate 110 10.3 0.036 0.042 7.093 7.108 integrate_v_rspace 110 11.3 0.003 0.004 7.056 7.080 multiply_cannon_metrocomm3 49320 15.4 0.080 0.084 6.214 7.029 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.587 6.687 calculate_rho_elec 110 8.6 0.021 0.026 6.587 6.687 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.635 6.673 ot_diis_step 99 11.5 0.005 0.006 6.539 6.539 calculate_first_density_matrix 1 7.0 0.000 0.000 6.183 6.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.112 6.116 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.786 5.823 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.784 5.784 init_scf_loop 11 6.9 0.000 0.000 5.570 5.571 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.272 5.332 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.218 5.245 cp_fm_diag_elpa_base 48 14.0 5.205 5.233 5.216 5.243 jit_kernel_multiply 13 15.9 4.805 4.829 4.805 4.829 calculate_dm_sparse 110 9.5 0.001 0.001 4.703 4.789 mp_sum_l 6594 12.7 3.663 4.457 3.663 4.457 rs_pw_transfer 902 11.9 0.012 0.013 3.654 4.238 wfi_extrapolate 11 7.9 0.001 0.001 4.004 4.004 make_m2s 4110 13.4 0.060 0.065 3.880 3.977 density_rs2pw 110 9.6 0.004 0.005 3.380 3.971 make_images 4110 14.4 0.177 0.191 3.785 3.886 acc_transpose_blocks 49320 15.4 0.210 0.217 3.463 3.511 grid_integrate_task_list 110 12.3 3.258 3.411 3.258 3.411 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.193 3.239 prepare_preconditioner 11 7.9 0.000 0.000 3.205 3.222 make_preconditioner 11 8.9 0.000 0.000 3.205 3.222 pw_transfer 1331 11.6 0.055 0.065 3.123 3.200 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.106 3.177 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.035 3.113 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.999 3.044 fft_wrap_pw1pw2_140 451 13.1 0.459 0.504 2.591 2.687 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.635 2.669 potential_pw2rs 110 12.3 0.006 0.007 2.609 2.626 mp_alltoall_d11v 2046 13.8 1.979 2.464 1.979 2.464 acc_transpose_blocks_kernels 49320 16.4 0.416 0.431 2.421 2.451 mp_waitany 14300 13.8 1.823 2.418 1.823 2.418 grid_collocate_task_list 110 9.6 2.152 2.384 2.152 2.384 fft3d_ps 1111 14.6 0.783 0.868 2.220 2.274 jit_kernel_transpose 5 15.6 2.005 2.026 2.005 2.026 mp_sum_d 3889 11.9 1.415 1.962 1.415 1.962 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.942 1.960 make_images_data 4110 15.4 0.042 0.046 1.787 1.924 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.566 1.831 cp_fm_cholesky_invert 11 10.9 1.745 1.748 1.745 1.748 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.918000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.090909, yerr=2.391445 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 588.783616E+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 2701873. 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 0.012 0.030 74.381 74.382 qs_mol_dyn_low 1 2.0 0.003 0.004 73.694 73.836 qs_forces 11 3.9 0.003 0.003 73.506 73.507 qs_energies 11 4.9 0.001 0.002 70.132 70.136 scf_env_do_scf 11 5.9 0.000 0.001 58.020 58.024 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.219 50.219 dbcsr_multiply_generic 2055 12.4 0.115 0.119 42.223 42.433 multiply_cannon 2055 13.4 0.222 0.243 35.667 37.070 velocity_verlet 10 3.0 0.001 0.002 36.246 36.250 multiply_cannon_loop 2055 14.4 0.919 0.950 34.374 35.437 qs_scf_new_mos 99 7.5 0.001 0.001 33.858 34.019 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.857 34.018 ot_scf_mini 99 9.5 0.003 0.003 32.208 32.362 ot_mini 99 10.5 0.001 0.001 19.239 19.399 multiply_cannon_multrec 24660 15.4 7.635 10.091 17.141 19.389 rebuild_ks_matrix 110 8.3 0.000 0.000 14.096 14.274 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.096 14.274 qs_ot_get_derivative 99 11.5 0.001 0.001 13.446 13.602 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.413 12.573 mp_waitall_1 176588 16.5 7.554 10.032 7.554 10.032 init_scf_run 11 5.9 0.000 0.001 9.737 9.738 scf_env_initial_rho_setup 11 6.9 0.000 0.001 9.737 9.738 dbcsr_mm_accdrv_process 52282 16.1 6.378 8.693 9.344 9.683 multiply_cannon_sync_h2d 24660 15.4 7.020 8.404 7.020 8.404 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.102 7.870 init_scf_loop 11 6.9 0.000 0.000 7.765 7.765 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.484 7.341 apply_single 110 13.6 0.000 0.001 6.484 7.341 sum_up_and_integrate 110 10.3 0.052 0.059 6.662 6.675 integrate_v_rspace 110 11.3 0.003 0.003 6.610 6.624 qs_ot_get_p 110 10.4 0.001 0.001 6.361 6.534 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.212 6.221 calculate_rho_elec 110 8.6 0.040 0.048 6.212 6.221 calculate_first_density_matrix 1 7.0 0.000 0.000 6.113 6.114 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 5.878 5.880 ot_diis_step 99 11.5 0.010 0.010 5.748 5.748 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.539 5.607 prepare_preconditioner 11 7.9 0.000 0.000 5.474 5.500 make_preconditioner 11 8.9 0.000 0.000 5.474 5.500 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.636 5.363 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.052 5.214 jit_kernel_multiply 10 16.2 2.618 5.119 2.618 5.119 make_m2s 4110 13.4 0.057 0.059 4.136 4.636 make_images 4110 14.4 0.401 0.446 4.028 4.524 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.502 4.521 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.076 4.076 calculate_dm_sparse 110 9.5 0.001 0.001 3.938 3.965 density_rs2pw 110 9.6 0.004 0.005 3.281 3.903 pw_transfer 1331 11.6 0.066 0.075 3.648 3.782 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.541 3.677 rs_pw_transfer 902 11.9 0.013 0.014 2.950 3.596 wfi_extrapolate 11 7.9 0.001 0.001 3.526 3.526 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.511 3.521 cp_fm_diag_elpa_base 48 14.0 3.463 3.483 3.509 3.518 grid_integrate_task_list 110 12.3 3.172 3.390 3.172 3.390 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.175 3.254 fft_wrap_pw1pw2_140 451 13.1 0.524 0.538 3.033 3.179 acc_transpose_blocks 24660 15.4 0.111 0.113 2.920 2.973 make_images_data 4110 15.4 0.046 0.050 2.328 2.807 hybrid_alltoall_any 4261 16.3 0.103 0.440 2.073 2.784 fft3d_ps 1111 14.6 1.104 1.322 2.509 2.653 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.526 2.538 cp_fm_cholesky_invert 11 10.9 2.528 2.535 2.528 2.535 grid_collocate_task_list 110 9.6 2.106 2.524 2.106 2.524 mp_sum_l 6594 12.7 1.790 2.523 1.790 2.523 potential_pw2rs 110 12.3 0.008 0.009 2.429 2.441 acc_transpose_blocks_kernels 24660 16.4 0.308 0.318 2.289 2.338 mp_alltoall_d11v 2046 13.8 1.776 2.103 1.776 2.103 jit_kernel_transpose 5 15.6 1.981 2.025 1.981 2.025 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.937 1.957 mp_waitany 10164 13.8 1.222 1.869 1.222 1.869 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.863 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.839 1.860 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.796 1.720 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.218 0.962 1.614 mp_allgather_i34 2055 14.4 0.622 1.611 0.622 1.611 mp_irecv_dv 57340 16.2 0.667 1.600 0.667 1.600 mp_sum_d 3889 11.9 1.079 1.560 1.079 1.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.382000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.090909, yerr=7.561320 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 658.333696E+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 0.057 0.073 62.882 62.883 qs_mol_dyn_low 1 2.0 0.003 0.004 62.504 62.517 qs_forces 11 3.9 0.003 0.003 62.392 62.392 qs_energies 11 4.9 0.008 0.056 59.130 59.133 scf_env_do_scf 11 5.9 0.000 0.001 49.020 49.020 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.105 40.105 velocity_verlet 10 3.0 0.015 0.025 32.700 32.703 dbcsr_multiply_generic 2055 12.4 0.106 0.112 30.508 30.770 multiply_cannon 2055 13.4 0.212 0.221 24.246 25.463 qs_scf_new_mos 99 7.5 0.001 0.001 24.936 25.044 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.936 25.043 multiply_cannon_loop 2055 14.4 0.613 0.625 23.051 24.181 ot_scf_mini 99 9.5 0.002 0.003 23.714 23.841 ot_mini 99 10.5 0.001 0.001 13.461 13.583 rebuild_ks_matrix 110 8.3 0.000 0.000 12.456 12.616 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.022 12.455 12.615 multiply_cannon_multrec 16440 15.4 3.889 4.973 10.897 11.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.278 11.464 mp_waitall_1 139946 16.5 7.033 10.304 7.033 10.304 qs_ot_get_derivative 99 11.5 0.001 0.001 8.958 9.084 init_scf_loop 11 6.9 0.000 0.000 8.880 8.882 init_scf_run 11 5.9 0.000 0.001 7.631 7.631 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.631 7.631 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.371 7.407 dbcsr_mm_accdrv_process 34862 16.1 5.992 6.545 6.861 7.030 prepare_preconditioner 11 7.9 0.000 0.000 6.733 6.751 make_preconditioner 11 8.9 0.000 0.000 6.733 6.751 sum_up_and_integrate 110 10.3 0.061 0.061 6.578 6.594 integrate_v_rspace 110 11.3 0.003 0.004 6.518 6.533 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.059 6.414 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.077 6.086 calculate_rho_elec 110 8.6 0.059 0.060 6.077 6.086 qs_ot_get_p 110 10.4 0.001 0.001 5.438 5.599 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.021 5.475 apply_single 110 13.6 0.000 0.000 5.021 5.475 make_m2s 4110 13.4 0.049 0.051 4.181 4.564 calculate_first_density_matrix 1 7.0 0.000 0.000 4.538 4.539 ot_diis_step 99 11.5 0.011 0.011 4.474 4.475 make_images 4110 14.4 0.395 0.514 4.067 4.450 density_rs2pw 110 9.6 0.004 0.005 3.135 4.419 multiply_cannon_sync_h2d 16440 15.4 3.678 4.351 3.678 4.351 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.060 4.061 rs_pw_transfer 902 11.9 0.011 0.012 2.696 3.968 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.827 3.835 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.728 3.772 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.047 3.694 pw_transfer 1331 11.6 0.066 0.073 3.660 3.671 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.553 3.567 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.481 3.481 grid_integrate_task_list 110 12.3 3.188 3.415 3.188 3.415 calculate_dm_sparse 110 9.5 0.001 0.001 3.379 3.410 fft_wrap_pw1pw2_140 451 13.1 0.641 0.653 3.074 3.093 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 make_images_data 4110 15.4 0.043 0.048 2.453 2.955 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.939 2.950 cp_fm_diag_elpa_base 48 14.0 2.871 2.902 2.937 2.948 hybrid_alltoall_any 4261 16.3 0.106 0.380 2.241 2.836 cp_fm_cholesky_invert 11 10.9 2.613 2.620 2.613 2.620 grid_collocate_task_list 110 9.6 2.119 2.529 2.119 2.529 mp_waitany 17072 13.8 1.186 2.526 1.186 2.526 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.886 2.428 fft3d_ps 1111 14.6 1.093 1.105 2.370 2.385 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.300 2.376 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.347 2.359 mp_irecv_dv 48980 15.7 0.816 2.299 0.816 2.299 potential_pw2rs 110 12.3 0.011 0.011 2.284 2.289 rs_pw_transfer_RS2PW_140 121 11.5 0.179 0.184 0.908 2.180 mp_alltoall_d11v 2046 13.8 1.805 2.153 1.805 2.153 mp_sum_l 6594 12.7 1.473 2.124 1.473 2.124 acc_transpose_blocks 16440 15.4 0.074 0.077 2.108 2.122 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.958 1.958 dbcsr_complete_redistribute 325 12.2 0.329 0.371 1.480 1.928 cp_fm_upper_to_full 70 14.2 1.382 1.760 1.382 1.760 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.698 1.717 acc_transpose_blocks_kernels 16440 16.4 0.207 0.210 1.675 1.681 mp_allgather_i34 2055 14.4 0.516 1.624 0.516 1.624 cp_fm_cholesky_decompose 22 10.9 1.577 1.596 1.577 1.596 jit_kernel_transpose 5 15.6 1.468 1.473 1.468 1.473 build_core_hamiltonian_matrix_ 11 4.9 0.007 0.050 1.365 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.996 1.449 mp_sum_d 3887 11.9 0.986 1.285 0.986 1.285 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.256 1.267 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.883000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.090909, yerr=8.240196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 737.255424E+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 0.066 0.089 67.601 67.602 qs_mol_dyn_low 1 2.0 0.003 0.004 67.167 67.177 qs_forces 11 3.9 0.003 0.003 67.076 67.077 qs_energies 11 4.9 0.007 0.042 63.574 63.578 scf_env_do_scf 11 5.9 0.001 0.002 52.897 52.899 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 41.352 41.353 velocity_verlet 10 3.0 0.010 0.016 36.915 36.918 dbcsr_multiply_generic 2055 12.4 0.113 0.118 31.413 31.596 qs_scf_new_mos 99 7.5 0.001 0.001 26.200 26.305 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.199 26.305 multiply_cannon 2055 13.4 0.245 0.267 24.394 25.560 ot_scf_mini 99 9.5 0.003 0.003 24.581 24.692 multiply_cannon_loop 2055 14.4 0.886 0.908 23.052 23.615 multiply_cannon_multrec 24660 15.4 4.252 6.926 13.875 14.955 ot_mini 99 10.5 0.001 0.001 14.046 14.169 rebuild_ks_matrix 110 8.3 0.000 0.000 12.276 12.389 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.275 12.389 init_scf_loop 11 6.9 0.001 0.003 11.504 11.504 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.821 10.921 dbcsr_mm_accdrv_process 52304 16.0 7.726 9.929 9.476 10.448 qs_ot_get_derivative 99 11.5 0.001 0.001 9.884 9.994 prepare_preconditioner 11 7.9 0.000 0.000 9.744 9.760 make_preconditioner 11 8.9 0.000 0.002 9.744 9.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.989 9.445 init_scf_run 11 5.9 0.000 0.001 7.833 7.834 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.833 7.834 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.601 6.611 calculate_rho_elec 110 8.6 0.078 0.082 6.600 6.610 sum_up_and_integrate 110 10.3 0.068 0.070 6.535 6.547 integrate_v_rspace 110 11.3 0.003 0.003 6.467 6.482 mp_waitall_1 121746 16.5 4.286 6.197 4.286 6.197 make_m2s 4110 13.4 0.058 0.059 5.459 5.783 qs_ot_get_p 110 10.4 0.001 0.001 5.507 5.661 make_images 4110 14.4 0.576 0.701 5.320 5.640 cp_fm_upper_to_full 70 14.2 3.333 4.836 3.333 4.836 calculate_first_density_matrix 1 7.0 0.001 0.002 4.339 4.342 pw_transfer 1331 11.6 0.066 0.076 4.164 4.193 ot_diis_step 99 11.5 0.011 0.011 4.125 4.126 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.032 4.119 apply_single 110 13.6 0.000 0.000 4.032 4.119 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.058 4.089 density_rs2pw 110 9.6 0.004 0.006 3.457 3.974 calculate_dm_sparse 110 9.5 0.001 0.001 3.805 3.834 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.823 3.825 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.776 3.791 dbcsr_complete_redistribute 325 12.2 0.420 0.464 2.606 3.753 fft_wrap_pw1pw2_140 451 13.1 0.672 0.694 3.543 3.578 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.476 3.521 grid_integrate_task_list 110 12.3 3.286 3.456 3.286 3.456 multiply_cannon_sync_h2d 24660 15.4 3.213 3.392 3.213 3.392 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.244 3.310 cp_dbcsr_syevd 48 12.0 0.003 0.005 3.301 3.301 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.093 3.223 make_images_data 4110 15.4 0.045 0.050 2.824 3.163 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.388 3.128 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.351 3.113 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.959 3.020 wfi_extrapolate 11 7.9 0.001 0.001 2.982 2.982 jit_kernel_multiply 10 16.0 1.419 2.921 1.419 2.921 rs_pw_transfer 902 11.9 0.010 0.012 2.388 2.917 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.746 2.864 fft3d_ps 1111 14.6 1.082 1.114 2.812 2.824 cp_fm_diag_elpa 48 13.0 0.000 0.001 2.782 2.795 cp_fm_diag_elpa_base 48 14.0 2.631 2.690 2.777 2.791 mp_alltoall_i22 605 13.7 1.636 2.788 1.636 2.788 cp_fm_cholesky_invert 11 10.9 2.646 2.654 2.646 2.654 grid_collocate_task_list 110 9.6 2.223 2.465 2.223 2.465 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.424 2.440 acc_transpose_blocks 24660 15.4 0.104 0.105 2.359 2.379 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.289 2.291 potential_pw2rs 110 12.3 0.013 0.013 2.156 2.163 mp_alltoall_d11v 2046 13.8 1.732 2.047 1.732 2.047 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.694 1.733 acc_transpose_blocks_kernels 24660 16.4 0.304 0.309 1.704 1.718 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.697 1.709 build_core_hamiltonian_matrix_ 11 4.9 0.016 0.065 1.630 1.706 cp_fm_cholesky_decompose 22 10.9 1.616 1.667 1.616 1.667 mp_alltoall_z22v 1111 16.6 1.417 1.581 1.417 1.581 mp_waitany 13376 13.8 1.067 1.538 1.067 1.538 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.856 1.500 mp_sum_l 6594 12.7 0.886 1.443 0.886 1.443 mp_allgather_i34 2055 14.4 0.438 1.434 0.438 1.434 mp_irecv_dv 62702 16.1 0.755 1.414 0.755 1.414 jit_kernel_transpose 5 15.6 1.401 1.414 1.401 1.414 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.179 0.832 1.369 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.602000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.090909, yerr=10.891653 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 833.884160E+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 0.013 0.033 59.755 59.756 qs_mol_dyn_low 1 2.0 0.003 0.004 59.510 59.522 qs_forces 11 3.9 0.003 0.003 57.767 57.768 qs_energies 11 4.9 0.001 0.002 54.019 54.024 scf_env_do_scf 11 5.9 0.011 0.041 45.095 45.096 scf_env_do_scf_inner_loop 99 6.5 0.007 0.027 37.254 37.255 velocity_verlet 10 3.0 0.002 0.003 32.855 32.962 dbcsr_multiply_generic 2055 12.4 0.104 0.109 24.860 24.997 qs_scf_new_mos 99 7.5 0.001 0.001 22.085 22.154 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.085 22.154 ot_scf_mini 99 9.5 0.002 0.002 20.854 20.918 multiply_cannon 2055 13.4 0.261 0.275 19.373 20.727 multiply_cannon_loop 2055 14.4 0.320 0.334 18.043 18.297 ot_mini 99 10.5 0.001 0.001 12.167 12.227 rebuild_ks_matrix 110 8.3 0.000 0.000 12.088 12.142 qs_ks_build_kohn_sham_matrix 110 9.3 0.017 0.029 12.088 12.142 multiply_cannon_multrec 8220 15.4 4.450 6.441 9.326 10.786 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.695 10.743 qs_ot_get_derivative 99 11.5 0.001 0.001 8.338 8.401 mp_waitall_1 103326 16.6 6.238 7.934 6.238 7.934 init_scf_loop 11 6.9 0.000 0.000 7.784 7.796 dbcsr_mm_accdrv_process 17442 15.9 3.379 6.341 4.745 6.863 sum_up_and_integrate 110 10.3 0.079 0.081 6.599 6.612 integrate_v_rspace 110 11.3 0.003 0.005 6.520 6.533 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.372 6.384 calculate_rho_elec 110 8.6 0.114 0.115 6.372 6.383 prepare_preconditioner 11 7.9 0.000 0.000 6.019 6.029 make_preconditioner 11 8.9 0.000 0.000 6.019 6.029 init_scf_run 11 5.9 0.000 0.001 5.741 5.741 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.741 5.741 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.592 5.672 qs_ot_get_p 110 10.4 0.001 0.001 4.747 4.822 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.210 4.675 make_m2s 4110 13.4 0.039 0.040 4.187 4.431 make_images 4110 14.4 0.637 0.696 4.057 4.302 pw_transfer 1331 11.6 0.066 0.072 3.961 3.966 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.853 3.863 ot_diis_step 99 11.5 0.012 0.012 3.808 3.808 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.722 3.790 apply_single 110 13.6 0.000 0.000 3.722 3.789 grid_integrate_task_list 110 12.3 3.357 3.553 3.357 3.553 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.432 3.435 density_rs2pw 110 9.6 0.004 0.004 3.039 3.413 fft_wrap_pw1pw2_140 451 13.1 0.837 0.848 3.323 3.336 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.127 3.127 multiply_cannon_sync_h2d 8220 15.4 2.883 3.101 2.883 3.101 cp_fm_cholesky_invert 11 10.9 2.890 2.894 2.890 2.894 calculate_first_density_matrix 1 7.0 0.000 0.001 2.870 2.873 calculate_dm_sparse 110 9.5 0.001 0.001 2.745 2.795 wfi_extrapolate 11 7.9 0.001 0.001 2.767 2.768 make_images_data 4110 15.4 0.038 0.043 2.361 2.765 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.343 2.709 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.679 2.680 jit_kernel_multiply 10 15.7 1.057 2.650 1.057 2.650 rs_pw_transfer 902 11.9 0.010 0.011 2.217 2.624 grid_collocate_task_list 110 9.6 2.316 2.577 2.316 2.577 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.563 2.570 cp_fm_diag_elpa_base 48 14.0 2.506 2.525 2.561 2.568 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.564 2.566 fft3d_ps 1111 14.6 1.134 1.164 2.394 2.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.203 2.219 potential_pw2rs 110 12.3 0.016 0.016 2.110 2.114 mp_alltoall_d11v 2046 13.8 1.758 2.108 1.758 2.108 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.784 1.996 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.919 1.956 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.787 1.818 mp_sum_dm 438 4.9 1.618 1.734 1.618 1.734 md_write_output 11 3.9 0.111 1.711 0.133 1.731 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.717 1.730 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.721 1.728 md_output 10 3.0 0.000 0.000 0.110 1.707 update_particle_set 20 4.0 0.000 0.000 1.594 1.701 cp_fm_cholesky_decompose 22 10.9 1.685 1.699 1.685 1.699 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.501 1.621 mp_allgather_i34 2055 14.4 0.502 1.607 0.502 1.607 dbcsr_complete_redistribute 325 12.2 0.564 0.587 1.448 1.544 mp_waitany 9240 13.8 1.092 1.531 1.092 1.531 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.481 1.405 mp_irecv_dv 24056 15.7 0.457 1.359 0.457 1.359 qs_create_task_list 11 7.9 0.001 0.001 1.216 1.321 generate_qs_task_list 11 8.9 0.375 0.443 1.215 1.320 rs_gather_matrices 110 12.3 0.324 0.363 0.963 1.275 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.204 1.228 rs_pw_transfer_RS2PW_140 121 11.5 0.164 0.168 0.811 1.224 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.820 1.217 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.756000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.909091, yerr=13.027625 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 1.353449E+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 0.016 0.033 88.098 88.099 qs_mol_dyn_low 1 2.0 0.003 0.003 87.794 87.808 qs_forces 11 3.9 0.003 0.003 87.711 87.711 qs_energies 11 4.9 0.001 0.001 83.493 83.494 scf_env_do_scf 11 5.9 0.000 0.001 72.948 72.948 velocity_verlet 10 3.0 0.002 0.002 56.279 56.287 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.183 44.184 dbcsr_multiply_generic 2055 12.4 0.118 0.122 28.929 29.060 init_scf_loop 11 6.9 0.000 0.000 28.693 28.695 prepare_preconditioner 11 7.9 0.000 0.000 26.684 26.697 make_preconditioner 11 8.9 0.000 0.000 26.684 26.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.674 26.162 qs_scf_new_mos 99 7.5 0.001 0.001 26.019 26.102 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.018 26.101 ot_scf_mini 99 9.5 0.002 0.002 24.261 24.329 multiply_cannon 2055 13.4 0.347 0.367 21.909 22.616 multiply_cannon_loop 2055 14.4 0.339 0.342 20.107 20.451 cp_fm_upper_to_full 70 14.2 12.924 18.667 12.924 18.667 rebuild_ks_matrix 110 8.3 0.000 0.001 14.054 14.142 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.053 14.141 ot_mini 99 10.5 0.001 0.001 13.563 13.631 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.705 12.785 dbcsr_complete_redistribute 325 12.2 1.028 1.043 7.584 10.964 multiply_cannon_multrec 8220 15.4 4.415 4.566 9.869 9.968 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.558 9.934 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.995 9.330 mp_alltoall_i22 605 13.7 5.641 9.038 5.641 9.038 qs_ot_get_derivative 99 11.5 0.001 0.001 8.884 8.953 mp_waitall_1 84994 16.7 7.439 8.404 7.439 8.404 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.686 7.719 calculate_rho_elec 110 8.6 0.224 0.224 7.685 7.719 sum_up_and_integrate 110 10.3 0.150 0.151 7.293 7.306 integrate_v_rspace 110 11.3 0.004 0.004 7.143 7.155 init_scf_run 11 5.9 0.000 0.001 6.267 6.267 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.267 6.267 make_m2s 4110 13.4 0.042 0.043 5.236 5.774 qs_ot_get_p 110 10.4 0.001 0.001 5.556 5.636 make_images 4110 14.4 0.876 0.926 5.049 5.586 dbcsr_mm_accdrv_process 11614 15.7 3.291 3.655 5.312 5.560 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.758 5.288 apply_single 110 13.6 0.000 0.000 4.758 5.288 cp_fm_cholesky_invert 11 10.9 5.236 5.241 5.236 5.241 pw_transfer 1331 11.6 0.075 0.076 5.002 5.015 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.704 4.991 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.885 4.897 ot_diis_step 99 11.5 0.015 0.016 4.659 4.659 fft_wrap_pw1pw2_140 451 13.1 1.334 1.343 4.263 4.271 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.994 4.002 multiply_cannon_sync_h2d 8220 15.4 3.950 3.955 3.950 3.955 grid_integrate_task_list 110 12.3 3.676 3.735 3.676 3.735 density_rs2pw 110 9.6 0.004 0.004 3.688 3.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.675 3.675 hybrid_alltoall_any 4261 16.3 0.257 0.553 2.870 3.603 calculate_dm_sparse 110 9.5 0.001 0.001 3.561 3.582 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.557 3.558 make_images_data 4110 15.4 0.041 0.043 2.856 3.499 wfi_extrapolate 11 7.9 0.001 0.001 3.325 3.325 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.834 3.305 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.969 2.970 cp_fm_diag_elpa_base 48 14.0 2.422 2.627 2.967 2.967 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.937 2.943 fft3d_ps 1111 14.6 1.301 1.307 2.818 2.825 calculate_first_density_matrix 1 7.0 0.000 0.000 2.821 2.822 grid_collocate_task_list 110 9.6 2.668 2.685 2.668 2.685 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.418 2.444 potential_pw2rs 110 12.3 0.021 0.022 2.396 2.403 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.362 2.398 rs_pw_transfer 902 11.9 0.011 0.011 2.249 2.265 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.181 2.242 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.115 2.192 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.098 2.112 jit_kernel_multiply 10 15.3 1.817 2.061 1.817 2.061 mp_alltoall_d11v 2046 13.8 1.939 2.043 1.939 2.043 cp_fm_cholesky_decompose 22 10.9 1.978 2.000 1.978 2.000 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.950 1.958 qs_create_task_list 11 7.9 0.000 0.001 1.888 1.935 generate_qs_task_list 11 8.9 0.734 0.786 1.888 1.934 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.786 1.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.099000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1220.545455, yerr=55.714146 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+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 6755942624 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 634.261504E+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 434.009031E+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 3175955597216 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 3526620. 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 0.022 0.055 208.355 208.357 qs_mol_dyn_low 1 2.0 0.004 0.012 207.773 207.789 qs_forces 11 3.9 0.006 0.007 207.650 207.652 qs_energies 11 4.9 0.002 0.002 201.921 201.939 scf_env_do_scf 11 5.9 0.001 0.001 184.713 184.717 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 164.294 164.296 dbcsr_multiply_generic 2507 12.6 0.176 0.179 124.848 125.395 qs_scf_new_mos 117 7.6 0.001 0.001 124.039 124.218 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.039 124.217 velocity_verlet 10 3.0 0.004 0.025 123.858 123.863 ot_scf_mini 117 9.6 0.003 0.003 117.400 117.543 multiply_cannon 2507 13.6 0.238 0.245 101.502 102.890 multiply_cannon_loop 2507 14.6 2.084 2.122 99.361 100.711 ot_mini 117 10.6 0.001 0.001 65.585 65.798 multiply_cannon_multrec 60168 15.6 33.409 35.332 41.806 43.714 qs_ot_get_derivative 117 11.6 0.001 0.003 40.845 40.960 rebuild_ks_matrix 128 8.3 0.001 0.001 33.797 34.025 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.796 34.025 mp_waitall_1 267128 16.5 28.063 30.554 28.063 30.554 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.316 30.521 qs_ot_get_p 128 10.4 0.001 0.001 29.868 30.178 multiply_cannon_sync_h2d 60168 15.6 27.623 29.100 27.623 29.100 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.360 25.089 apply_single 128 13.6 0.001 0.001 24.360 25.088 ot_diis_step 117 11.6 0.008 0.008 24.498 24.499 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.376 23.448 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.873 20.874 init_scf_loop 11 6.9 0.000 0.001 20.344 20.345 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.606 19.696 cp_fm_diag_elpa 83 13.4 0.000 0.001 17.911 17.937 cp_fm_diag_elpa_base 83 14.4 17.841 17.876 17.906 17.934 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.782 17.825 prepare_preconditioner 11 7.9 0.000 0.000 15.726 15.773 make_preconditioner 11 8.9 0.000 0.000 15.726 15.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.958 15.146 sum_up_and_integrate 128 10.3 0.089 0.108 14.281 14.298 integrate_v_rspace 128 11.3 0.004 0.005 14.191 14.212 make_m2s 5014 13.6 0.105 0.112 13.798 14.087 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.917 14.015 calculate_rho_elec 128 8.7 0.046 0.065 13.917 14.015 make_images 5014 14.6 0.404 0.423 13.621 13.922 init_scf_run 11 5.9 0.000 0.001 13.029 13.029 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.029 13.029 density_rs2pw 128 9.7 0.006 0.007 7.154 10.340 mp_sum_l 7950 12.9 8.291 9.572 8.291 9.572 wfi_extrapolate 11 7.9 0.001 0.001 9.167 9.167 rs_pw_transfer 1046 11.9 0.017 0.019 5.812 8.992 cp_fm_cholesky_invert 11 10.9 8.899 8.906 8.899 8.906 calculate_dm_sparse 128 9.5 0.001 0.001 8.521 8.634 dbcsr_mm_accdrv_process 124484 16.2 3.135 3.240 7.957 8.469 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.915 8.015 pw_transfer 1547 11.6 0.074 0.088 7.823 8.015 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.853 7.943 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.620 7.813 make_images_data 5014 15.6 0.066 0.071 6.761 7.717 grid_integrate_task_list 128 12.3 7.071 7.566 7.071 7.566 hybrid_alltoall_any 5200 16.5 0.292 2.271 5.921 7.268 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 7.187 7.197 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 5.769 7.114 fft_wrap_pw1pw2_140 523 13.2 1.278 1.324 6.663 6.843 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.326 6.394 grid_collocate_task_list 128 9.7 4.680 5.966 4.680 5.966 mp_waitany 16020 13.9 2.724 5.956 2.724 5.956 fft3d_ps 1291 14.7 2.136 2.780 5.378 5.720 mp_alltoall_d11v 2415 14.1 4.266 5.464 4.266 5.464 rs_pw_transfer_RS2PW_140 139 11.5 0.281 0.302 2.156 5.345 cp_fm_cholesky_decompose 22 10.9 4.646 4.660 4.646 4.660 potential_pw2rs 128 12.3 0.009 0.011 4.630 4.657 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.357000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.363636, yerr=7.113019 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 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 11528891191296 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 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 833.150976E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6583622. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.181 191.648 191.649 qs_mol_dyn_low 1 2.0 0.003 0.006 191.069 191.084 qs_forces 11 3.9 0.005 0.018 190.972 190.974 qs_energies 11 4.9 0.003 0.014 184.060 184.071 scf_env_do_scf 11 5.9 0.020 0.155 167.281 167.291 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.184 134.187 velocity_verlet 10 3.0 0.002 0.004 120.213 120.215 dbcsr_multiply_generic 2507 12.6 0.188 0.194 97.484 98.689 qs_scf_new_mos 117 7.6 0.001 0.001 93.943 94.461 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.942 94.460 ot_scf_mini 117 9.6 0.004 0.006 89.120 89.738 multiply_cannon 2507 13.6 0.486 0.593 77.345 81.775 multiply_cannon_loop 2507 14.6 1.239 1.271 74.063 76.542 ot_mini 117 10.6 0.002 0.010 49.645 50.203 mp_waitall_1 214728 16.6 24.468 37.542 24.468 37.542 multiply_cannon_multrec 30084 15.6 22.038 26.365 31.696 36.459 rebuild_ks_matrix 128 8.3 0.001 0.001 33.469 34.107 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.028 33.468 34.107 init_scf_loop 11 6.9 0.001 0.008 32.854 32.855 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.061 30.634 prepare_preconditioner 11 7.9 0.000 0.000 28.406 28.462 make_preconditioner 11 8.9 0.000 0.002 28.406 28.462 multiply_cannon_metrocomm3 30084 15.6 0.090 0.096 15.576 28.446 qs_ot_get_derivative 117 11.6 0.001 0.002 27.695 28.300 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.035 27.579 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.012 23.068 apply_single 128 13.6 0.001 0.001 22.012 23.067 multiply_cannon_sync_h2d 30084 15.6 19.244 21.930 19.244 21.930 qs_ot_get_p 128 10.4 0.002 0.010 21.224 21.894 ot_diis_step 117 11.6 0.015 0.022 21.774 21.776 cp_fm_cholesky_invert 11 10.9 16.598 16.610 16.598 16.610 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.415 16.449 make_m2s 5014 13.6 0.091 0.097 14.361 15.973 make_images 5014 14.6 1.164 1.343 14.150 15.760 cp_dbcsr_syevd 83 12.4 0.006 0.014 15.283 15.285 sum_up_and_integrate 128 10.3 0.117 0.135 14.775 14.806 integrate_v_rspace 128 11.3 0.004 0.005 14.658 14.695 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.212 14.247 calculate_rho_elec 128 8.7 0.089 0.106 14.212 14.246 cp_fm_diag_elpa 83 13.4 0.001 0.001 12.139 12.167 cp_fm_diag_elpa_base 83 14.4 11.882 11.999 12.132 12.162 init_scf_run 11 5.9 0.000 0.001 11.863 11.865 scf_env_initial_rho_setup 11 6.9 0.000 0.002 11.863 11.865 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.893 11.323 multiply_cannon_metrocomm4 27577 15.6 0.103 0.117 3.803 10.662 make_images_data 5014 15.6 0.064 0.073 8.799 10.615 mp_irecv_dv 69486 16.3 3.603 10.274 3.603 10.274 hybrid_alltoall_any 5200 16.5 0.344 1.541 7.480 9.925 density_rs2pw 128 9.7 0.007 0.008 7.478 9.922 dbcsr_mm_accdrv_process 62242 16.2 4.576 5.524 9.111 9.714 pw_transfer 1547 11.6 0.086 0.096 8.802 8.856 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.577 8.634 wfi_extrapolate 11 7.9 0.002 0.007 8.504 8.504 rs_pw_transfer 1046 11.9 0.014 0.017 5.804 8.272 fft_wrap_pw1pw2_140 523 13.2 1.339 1.369 7.576 7.648 grid_integrate_task_list 128 12.3 7.192 7.604 7.192 7.604 cp_fm_cholesky_decompose 22 10.9 6.892 6.973 6.892 6.973 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.201 6.967 calculate_dm_sparse 128 9.5 0.001 0.001 6.504 6.641 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.373 6.391 grid_collocate_task_list 128 9.7 4.783 5.946 4.783 5.946 fft3d_ps 1291 14.7 2.812 2.986 5.882 5.936 mp_sum_l 7950 12.9 4.077 5.745 4.077 5.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.411 5.574 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.304 5.371 mp_allgather_i34 2507 14.6 1.838 5.120 1.838 5.120 mp_waitany 11748 13.9 2.568 5.081 2.568 5.081 potential_pw2rs 128 12.3 0.015 0.017 4.989 5.009 mp_alltoall_d11v 2415 14.1 4.191 4.741 4.191 4.741 rs_pw_transfer_RS2PW_140 139 11.5 0.355 0.381 2.153 4.592 dbcsr_complete_redistribute 395 12.7 0.773 0.849 3.216 4.077 mp_sum_d 4465 12.1 2.658 3.973 2.658 3.973 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.806 3.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.649000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.181818, yerr=4.018552 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/17/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 2.906045E+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 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 958.144512E+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 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58278. MP_Allreduce 11070 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 0.021 0.082 174.657 174.658 qs_mol_dyn_low 1 2.0 0.003 0.005 174.156 174.191 qs_forces 11 3.9 0.005 0.010 174.053 174.060 qs_energies 11 4.9 0.002 0.006 167.316 167.326 scf_env_do_scf 11 5.9 0.001 0.003 150.988 150.990 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.438 116.439 velocity_verlet 10 3.0 0.003 0.004 111.227 111.229 dbcsr_multiply_generic 2485 12.5 0.177 0.183 81.669 82.797 qs_scf_new_mos 116 7.6 0.001 0.001 78.545 78.832 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.544 78.831 ot_scf_mini 116 9.6 0.003 0.004 74.447 74.803 multiply_cannon 2485 13.5 0.501 0.536 61.919 66.830 multiply_cannon_loop 2485 14.5 0.846 0.877 58.933 61.623 ot_mini 116 10.6 0.001 0.001 42.399 42.751 init_scf_loop 11 6.9 0.001 0.004 34.452 34.456 mp_waitall_1 169034 16.6 24.639 34.287 24.639 34.287 rebuild_ks_matrix 127 8.3 0.001 0.001 30.511 30.894 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.020 30.510 30.894 prepare_preconditioner 11 7.9 0.000 0.001 30.418 30.457 make_preconditioner 11 8.9 0.000 0.002 30.417 30.457 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.111 29.465 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.332 27.683 multiply_cannon_multrec 19880 15.5 13.591 16.342 22.630 25.543 multiply_cannon_metrocomm3 19880 15.5 0.057 0.060 14.800 24.392 qs_ot_get_derivative 116 11.6 0.001 0.002 22.839 23.188 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.749 20.851 apply_single 127 13.6 0.001 0.001 19.749 20.851 ot_diis_step 116 11.6 0.018 0.019 19.460 19.461 qs_ot_get_p 127 10.4 0.001 0.001 17.662 18.061 multiply_cannon_sync_h2d 19880 15.5 14.136 15.915 14.136 15.915 make_m2s 4970 13.5 0.080 0.087 14.639 15.543 make_images 4970 14.5 1.172 1.262 14.408 15.313 sum_up_and_integrate 127 10.3 0.133 0.143 14.529 14.562 integrate_v_rspace 127 11.3 0.004 0.005 14.396 14.427 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.356 14.391 calculate_rho_elec 127 8.7 0.131 0.146 14.356 14.391 cp_fm_cholesky_invert 11 10.9 14.193 14.202 14.193 14.202 qs_ot_p2m_diag 82 11.4 0.262 0.269 13.401 13.408 cp_dbcsr_syevd 82 12.4 0.005 0.006 12.427 12.428 init_scf_run 11 5.9 0.000 0.001 11.259 11.259 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.259 11.259 make_images_data 4970 15.5 0.058 0.066 9.081 10.374 hybrid_alltoall_any 5155 16.4 0.430 1.966 7.796 9.774 cp_fm_diag_elpa 82 13.4 0.000 0.001 9.382 9.399 cp_fm_diag_elpa_base 82 14.4 8.980 9.139 9.379 9.396 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.823 9.080 density_rs2pw 127 9.7 0.006 0.007 7.141 8.994 pw_transfer 1535 11.6 0.086 0.103 8.868 8.977 multiply_cannon_metrocomm4 17395 15.5 0.061 0.072 3.340 8.941 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.645 8.758 dbcsr_mm_accdrv_process 41158 16.2 5.256 5.665 8.488 8.754 mp_irecv_dv 49801 16.2 3.221 8.706 3.221 8.706 fft_wrap_pw1pw2_140 519 13.2 1.408 1.436 7.705 7.832 grid_integrate_task_list 127 12.3 7.232 7.747 7.232 7.747 wfi_extrapolate 11 7.9 0.001 0.001 7.438 7.438 cp_fm_cholesky_decompose 22 10.9 7.213 7.246 7.213 7.246 cp_fm_upper_to_full 104 14.8 5.662 7.144 5.662 7.144 rs_pw_transfer 1038 11.9 0.014 0.014 5.091 6.944 dbcsr_complete_redistribute 393 12.7 1.173 1.206 4.485 6.220 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.145 6.151 fft3d_ps 1281 14.7 2.721 2.957 5.783 5.878 calculate_dm_sparse 127 9.5 0.001 0.001 5.724 5.823 grid_collocate_task_list 127 9.7 4.970 5.680 4.970 5.680 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.606 5.406 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.231 5.365 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.350 5.084 mp_alltoall_d11v 2401 14.1 4.311 5.008 4.311 5.008 mp_sum_l 7884 12.9 3.335 4.811 3.335 4.811 potential_pw2rs 127 12.3 0.020 0.023 4.644 4.661 mp_allgather_i34 2485 14.5 1.512 4.532 1.512 4.532 mp_waitany 11660 13.9 2.390 4.283 2.390 4.283 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.955 3.983 transfer_fm_to_dbcsr 11 9.9 0.019 0.028 2.287 3.963 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.813 3.814 mp_alltoall_i22 712 14.1 1.931 3.789 1.931 3.789 rs_pw_transfer_RS2PW_140 138 11.5 0.337 0.374 1.942 3.764 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.675 3.709 calculate_first_density_matrix 1 7.0 0.000 0.003 3.592 3.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.658000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=899.636364, yerr=11.170356 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/18/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 4.387242E+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 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.162023E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57598. MP_Allreduce 11189 1079. MP_Sync 88 MP_Alltoall 1724 12509605. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 190.769 190.770 qs_mol_dyn_low 1 2.0 0.003 0.004 189.935 189.948 qs_forces 11 3.9 0.005 0.012 189.802 189.806 qs_energies 11 4.9 0.002 0.009 182.538 182.551 scf_env_do_scf 11 5.9 0.001 0.001 163.473 163.480 velocity_verlet 10 3.0 0.002 0.002 125.144 125.148 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.353 117.354 dbcsr_multiply_generic 2529 12.6 0.186 0.193 81.362 82.014 qs_scf_new_mos 118 7.6 0.001 0.001 79.516 79.833 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.516 79.832 ot_scf_mini 118 9.6 0.003 0.004 74.978 75.299 multiply_cannon 2529 13.6 0.565 0.595 56.963 59.883 multiply_cannon_loop 2529 14.6 1.192 1.222 53.166 54.715 init_scf_loop 11 6.9 0.001 0.005 46.001 46.002 ot_mini 118 10.6 0.001 0.001 42.804 43.118 prepare_preconditioner 11 7.9 0.000 0.000 41.888 41.908 make_preconditioner 11 8.9 0.000 0.001 41.888 41.908 make_full_inverse_cholesky 11 9.9 0.012 0.021 35.553 40.497 multiply_cannon_multrec 30348 15.6 14.233 20.284 27.093 32.662 rebuild_ks_matrix 129 8.3 0.001 0.001 30.149 30.464 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.149 30.464 mp_waitall_1 149172 16.7 17.576 27.527 17.576 27.527 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.142 27.432 qs_ot_get_derivative 118 11.6 0.002 0.002 22.979 23.308 make_m2s 5058 13.6 0.097 0.105 20.035 21.169 make_images 5058 14.6 1.966 2.304 19.727 20.861 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.235 19.788 apply_single 129 13.6 0.001 0.001 19.235 19.788 ot_diis_step 118 11.6 0.018 0.019 19.692 19.694 qs_ot_get_p 129 10.4 0.001 0.001 18.386 18.761 cp_fm_upper_to_full 106 14.8 11.330 16.655 11.330 16.655 cp_fm_cholesky_invert 11 10.9 16.080 16.089 16.080 16.089 multiply_cannon_metrocomm3 30348 15.6 0.048 0.051 6.841 15.627 sum_up_and_integrate 129 10.3 0.141 0.153 14.871 14.913 integrate_v_rspace 129 11.3 0.004 0.005 14.730 14.775 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.721 14.750 calculate_rho_elec 129 8.7 0.177 0.193 14.720 14.749 qs_ot_p2m_diag 84 11.4 0.347 0.394 14.364 14.417 dbcsr_mm_accdrv_process 62780 16.2 7.755 9.129 12.433 13.117 dbcsr_complete_redistribute 397 12.7 1.501 1.622 9.318 13.115 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.085 13.086 init_scf_run 11 5.9 0.000 0.001 12.997 12.999 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.996 12.998 multiply_cannon_sync_h2d 30348 15.6 11.761 12.807 11.761 12.807 make_images_data 5058 15.6 0.062 0.066 10.738 12.717 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.940 11.709 hybrid_alltoall_any 5245 16.5 0.529 2.201 9.762 11.463 cp_fm_diag_elpa 84 13.4 0.000 0.000 10.020 10.033 cp_fm_diag_elpa_base 84 14.4 9.057 9.374 10.012 10.029 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.311 9.959 mp_alltoall_i22 720 14.1 5.903 9.647 5.903 9.647 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.414 9.644 pw_transfer 1559 11.6 0.087 0.102 9.202 9.273 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.973 9.052 density_rs2pw 129 9.7 0.006 0.006 7.001 8.309 fft_wrap_pw1pw2_140 527 13.2 1.576 1.612 7.998 8.087 grid_integrate_task_list 129 12.3 7.569 8.023 7.569 8.023 wfi_extrapolate 11 7.9 0.001 0.001 7.743 7.743 cp_fm_cholesky_decompose 22 10.9 7.447 7.562 7.447 7.562 calculate_dm_sparse 129 9.5 0.001 0.001 7.110 7.246 multiply_cannon_metrocomm4 25290 15.6 0.081 0.091 2.811 7.062 mp_irecv_dv 76751 16.2 2.660 6.791 2.660 6.791 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.428 6.501 rs_pw_transfer 1054 12.0 0.013 0.014 4.710 6.088 fft3d_ps 1301 14.7 2.859 2.931 5.868 5.929 grid_collocate_task_list 129 9.7 5.204 5.868 5.204 5.868 mp_alltoall_d11v 2429 14.1 4.910 5.496 4.910 5.496 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.360 5.429 calculate_first_density_matrix 1 7.0 0.058 0.466 5.019 5.024 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.664 4.780 potential_pw2rs 129 12.3 0.024 0.024 4.561 4.574 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.502 4.502 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.205 4.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.770000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1094.454545, yerr=21.253293 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.518719E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 174.573 174.574 qs_mol_dyn_low 1 2.0 0.003 0.004 174.126 174.183 qs_forces 11 3.9 0.004 0.004 173.910 173.915 qs_energies 11 4.9 0.003 0.005 166.264 166.271 scf_env_do_scf 11 5.9 0.001 0.001 147.547 147.550 velocity_verlet 10 3.0 0.107 0.108 113.770 113.775 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 111.989 111.990 dbcsr_multiply_generic 2507 12.6 0.179 0.191 75.675 76.244 qs_scf_new_mos 117 7.6 0.001 0.001 73.577 73.701 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.577 73.700 ot_scf_mini 117 9.6 0.003 0.004 69.142 69.208 multiply_cannon 2507 13.6 0.584 0.636 55.787 60.285 multiply_cannon_loop 2507 14.6 0.445 0.455 51.028 52.022 ot_mini 117 10.6 0.001 0.001 39.724 39.795 init_scf_loop 11 6.9 0.000 0.000 35.413 35.414 mp_waitall_1 125778 16.7 25.904 33.487 25.904 33.487 prepare_preconditioner 11 7.9 0.000 0.000 31.456 31.480 make_preconditioner 11 8.9 0.000 0.000 31.456 31.480 rebuild_ks_matrix 128 8.3 0.001 0.001 29.873 29.986 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.873 29.985 make_full_inverse_cholesky 11 9.9 0.018 0.028 29.351 29.596 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.049 27.152 multiply_cannon_multrec 10028 15.6 10.397 14.270 18.965 21.706 ot_diis_step 117 11.6 0.020 0.021 20.003 20.004 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.523 19.938 apply_single 128 13.6 0.001 0.001 19.523 19.938 qs_ot_get_derivative 117 11.6 0.002 0.002 19.654 19.728 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.288 19.610 make_m2s 5014 13.6 0.067 0.071 16.362 18.910 make_images 5014 14.6 2.314 2.789 16.054 18.606 cp_fm_cholesky_invert 11 10.9 17.928 17.934 17.928 17.934 qs_ot_get_p 128 10.4 0.001 0.001 16.376 16.457 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.292 15.351 calculate_rho_elec 128 8.7 0.258 0.269 15.291 15.350 sum_up_and_integrate 128 10.3 0.179 0.189 14.907 14.957 integrate_v_rspace 128 11.3 0.004 0.004 14.727 14.786 qs_ot_p2m_diag 83 11.4 0.496 0.505 12.721 12.737 make_images_data 5014 15.6 0.051 0.060 10.038 12.505 hybrid_alltoall_any 5200 16.5 0.825 3.638 9.893 12.137 multiply_cannon_sync_h2d 10028 15.6 11.670 12.129 11.670 12.129 init_scf_run 11 5.9 0.000 0.001 11.681 11.682 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.681 11.682 cp_dbcsr_syevd 83 12.4 0.005 0.005 11.612 11.613 pw_transfer 1547 11.6 0.086 0.095 9.412 9.435 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.188 9.217 dbcsr_mm_accdrv_process 20762 16.1 2.634 3.523 8.194 8.847 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.655 8.668 cp_fm_diag_elpa_base 83 14.4 8.421 8.493 8.652 8.664 fft_wrap_pw1pw2_140 523 13.2 1.923 1.956 8.169 8.202 grid_integrate_task_list 128 12.3 7.764 8.073 7.764 8.073 cp_fm_cholesky_decompose 22 10.9 7.839 7.966 7.839 7.966 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.751 7.809 density_rs2pw 128 9.7 0.006 0.006 6.786 7.685 mp_allgather_i34 2507 14.6 2.952 7.608 2.952 7.608 wfi_extrapolate 11 7.9 0.001 0.001 7.557 7.558 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.362 7.533 calculate_dm_sparse 128 9.5 0.001 0.001 7.161 7.237 grid_collocate_task_list 128 9.7 5.468 6.143 5.468 6.143 fft3d_ps 1291 14.7 2.767 2.849 5.651 5.688 mp_alltoall_d11v 2415 14.1 4.927 5.666 4.927 5.666 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.488 5.488 dbcsr_complete_redistribute 395 12.7 2.126 2.183 5.089 5.402 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.347 5.373 rs_pw_transfer 1046 11.9 0.013 0.013 4.031 4.952 potential_pw2rs 128 12.3 0.027 0.027 4.349 4.360 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.282 4.313 calculate_first_density_matrix 1 7.0 0.000 0.000 3.896 3.905 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.588 3.872 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.852 3.753 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.728 3.740 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.692 3.711 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.669 3.699 mp_irecv_dv 28860 15.9 1.817 3.689 1.817 3.689 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.329 3.641 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.497 3.585 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.574000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1410.181818, yerr=43.621967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 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 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.111023E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 58741. MP_Allreduce 11152 1501. MP_Sync 87 MP_Alltoall 1724 36993717. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.039 304.560 304.562 qs_mol_dyn_low 1 2.0 0.003 0.004 303.933 303.957 qs_forces 11 3.9 0.004 0.005 303.835 303.837 qs_energies 11 4.9 0.005 0.005 294.449 294.454 scf_env_do_scf 11 5.9 0.001 0.001 268.378 268.392 velocity_verlet 10 3.0 0.006 0.006 217.149 217.157 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 141.934 141.936 init_scf_loop 11 6.9 0.000 0.000 126.197 126.200 prepare_preconditioner 11 7.9 0.000 0.000 121.067 121.101 make_preconditioner 11 8.9 0.000 0.000 121.067 121.101 make_full_inverse_cholesky 11 9.9 0.037 0.039 96.021 118.170 qs_scf_new_mos 118 7.6 0.001 0.001 90.829 91.010 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.829 91.009 dbcsr_multiply_generic 2527 12.6 0.212 0.223 88.149 89.062 ot_scf_mini 118 9.6 0.004 0.004 85.924 86.040 cp_fm_upper_to_full 105 14.8 54.064 78.124 54.064 78.124 multiply_cannon 2527 13.6 0.707 0.734 63.163 64.386 multiply_cannon_loop 2527 14.6 0.476 0.483 59.515 61.640 ot_mini 118 10.6 0.001 0.001 45.104 45.247 dbcsr_complete_redistribute 395 12.7 3.978 4.020 30.559 44.308 copy_fm_to_dbcsr 209 11.7 0.001 0.002 27.267 41.052 rebuild_ks_matrix 129 8.3 0.001 0.001 38.757 38.931 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 38.756 38.930 transfer_fm_to_dbcsr 11 9.9 0.029 0.034 25.001 38.668 mp_alltoall_i22 718 14.1 22.834 36.666 22.834 36.666 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.749 35.908 cp_fm_cholesky_invert 11 10.9 33.079 33.085 33.079 33.085 mp_waitall_1 104500 16.8 28.686 33.004 28.686 33.004 qs_ot_get_p 129 10.4 0.001 0.001 25.193 25.341 qs_ot_get_derivative 118 11.6 0.002 0.002 24.394 24.510 multiply_cannon_multrec 10108 15.6 10.662 12.487 21.250 21.371 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 19.194 21.180 qs_ot_p2m_diag 83 11.4 0.879 0.884 21.127 21.155 make_m2s 5054 13.6 0.076 0.080 20.108 21.039 ot_diis_step 118 11.6 0.022 0.023 20.664 20.664 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.551 20.567 calculate_rho_elec 129 8.7 0.484 0.485 20.551 20.567 make_images 5054 14.6 3.818 3.910 19.626 20.555 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.673 20.109 apply_single 129 13.6 0.001 0.001 19.672 20.108 sum_up_and_integrate 129 10.3 0.323 0.324 19.723 19.794 integrate_v_rspace 129 11.3 0.004 0.004 19.399 19.471 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.422 19.423 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.278 16.278 cp_fm_diag_elpa_base 83 14.4 11.781 13.488 16.274 16.274 init_scf_run 11 5.9 0.000 0.001 15.999 16.000 scf_env_initial_rho_setup 11 6.9 0.000 0.000 15.999 15.999 multiply_cannon_sync_h2d 10108 15.6 15.781 15.807 15.781 15.807 hybrid_alltoall_any 5240 16.5 1.309 3.069 10.913 12.783 make_images_data 5054 15.6 0.058 0.063 10.723 12.420 dbcsr_mm_accdrv_process 20926 16.1 3.854 5.759 10.347 12.219 pw_transfer 1559 11.6 0.094 0.095 11.342 11.353 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.104 11.113 fft_wrap_pw1pw2_140 527 13.2 3.092 3.131 9.893 9.904 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.365 9.451 mp_alltoall_d11v 2423 14.1 8.276 9.259 8.276 9.259 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 9.039 9.115 wfi_extrapolate 11 7.9 0.001 0.001 9.085 9.085 grid_integrate_task_list 129 12.3 8.659 8.852 8.659 8.852 cp_fm_cholesky_decompose 22 10.9 8.719 8.750 8.719 8.750 density_rs2pw 129 9.7 0.005 0.005 7.954 8.110 qs_energies_init_hamiltonians 11 5.9 0.009 0.013 8.040 8.040 calculate_dm_sparse 129 9.5 0.001 0.001 7.790 7.868 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.292 7.341 calculate_first_density_matrix 1 7.0 0.000 0.000 6.638 6.639 grid_collocate_task_list 129 9.7 6.487 6.556 6.487 6.556 rs_scatter_matrices 140 9.7 3.734 4.619 6.069 6.298 fft3d_ps 1301 14.7 2.869 2.887 6.148 6.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=304.562000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2815.727273, yerr=161.986582 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 1.259389E+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 24921484. 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 0.017 0.036 85.303 85.304 qs_energies 1 2.0 0.000 0.000 84.802 84.812 ls_scf 1 3.0 0.000 0.000 83.546 83.557 dbcsr_multiply_generic 111 6.7 0.016 0.016 72.456 72.584 multiply_cannon 111 7.7 0.018 0.020 55.962 56.854 multiply_cannon_loop 111 8.7 0.209 0.227 52.570 53.445 ls_scf_main 1 4.0 0.000 0.000 52.232 52.232 density_matrix_trs4 2 5.0 0.002 0.003 46.669 46.757 ls_scf_init_scf 1 4.0 0.000 0.000 28.292 28.292 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.106 27.155 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.024 25.047 mp_waitall_1 11031 10.9 22.193 24.981 22.193 24.981 multiply_cannon_multrec 2664 9.7 8.222 9.050 15.627 17.515 multiply_cannon_sync_h2d 2664 9.7 13.745 15.117 13.745 15.117 make_m2s 222 7.7 0.009 0.012 12.988 13.428 make_images 222 8.7 0.101 0.112 12.967 13.409 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.498 12.046 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.459 8.783 make_images_data 222 9.7 0.005 0.006 7.568 8.203 hybrid_alltoall_any 227 10.6 0.218 1.858 6.475 8.054 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.634 7.022 8.044 dbcsr_mm_accdrv_process_sort 4760 11.4 6.302 7.236 6.302 7.236 calculate_norms 4752 9.8 5.562 6.245 5.562 6.245 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.999 5.194 mp_sum_l 887 5.1 2.961 4.557 2.961 4.557 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.045 3.812 mp_irecv_dv 6231 10.9 2.029 3.782 2.029 3.782 make_images_sizes 222 9.7 0.000 0.000 0.767 3.559 mp_alltoall_i44 222 10.7 0.766 3.558 0.766 3.558 arnoldi_extremal 4 6.8 0.000 0.000 3.212 3.243 arnoldi_normal_ev 4 7.8 0.001 0.003 3.212 3.243 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.205 3.204 build_subspace 16 8.4 0.009 0.012 3.119 3.121 ls_scf_post 1 4.0 0.000 0.000 3.023 3.034 ls_scf_store_result 1 5.0 0.000 0.000 2.835 2.873 dbcsr_special_finalize 555 9.7 0.005 0.006 2.291 2.733 dbcsr_merge_single_wm 555 10.7 0.453 0.560 2.283 2.724 make_images_pack 222 9.7 2.213 2.645 2.215 2.646 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.312 2.551 dbcsr_sort_data 658 11.4 2.088 2.483 2.088 2.483 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.460 2.072 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.345 2.442 buffer_matrices_ensure_size 222 8.7 1.757 2.078 1.757 2.078 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.781 1.782 rebuild_ks_matrix 3 7.3 0.000 0.000 1.772 1.773 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.772 1.773 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.304000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 2.109739E+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 0.041 0.054 92.179 92.179 qs_energies 1 2.0 0.000 0.000 91.621 91.694 ls_scf 1 3.0 0.000 0.000 90.290 90.364 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.041 76.441 multiply_cannon 111 7.7 0.030 0.041 53.578 57.592 ls_scf_main 1 4.0 0.000 0.000 55.802 55.809 multiply_cannon_loop 111 8.7 0.115 0.122 50.261 53.713 density_matrix_trs4 2 5.0 0.002 0.003 50.010 50.197 ls_scf_init_scf 1 4.0 0.000 0.000 30.828 30.830 mp_waitall_1 9105 10.9 21.242 29.810 21.242 29.810 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.620 29.719 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.213 27.226 multiply_cannon_multrec 1332 9.7 13.411 16.373 22.727 26.845 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.708 20.540 make_m2s 222 7.7 0.008 0.009 15.614 16.367 make_images 222 8.7 1.591 1.965 15.583 16.338 dbcsr_mm_accdrv_process 4041 10.4 0.290 0.448 8.911 10.529 dbcsr_mm_accdrv_process_sort 4041 11.4 8.506 10.080 8.506 10.080 make_images_data 222 9.7 0.004 0.005 8.975 10.043 hybrid_alltoall_any 227 10.6 0.524 2.459 8.380 9.379 mp_sum_l 887 5.1 5.540 8.876 5.540 8.876 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.226 7.597 mp_irecv_dv 3311 11.0 3.206 7.525 3.206 7.525 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.298 7.235 calculate_norms 2376 9.8 5.997 6.723 5.997 6.723 multiply_cannon_sync_h2d 1332 9.7 4.813 6.181 4.813 6.181 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.043 5.258 arnoldi_extremal 4 6.8 0.000 0.000 4.670 4.688 arnoldi_normal_ev 4 7.8 0.001 0.005 4.670 4.688 build_subspace 16 8.4 0.014 0.021 4.421 4.423 ls_scf_post 1 4.0 0.000 0.000 3.659 3.733 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.178 3.412 ls_scf_store_result 1 5.0 0.000 0.000 3.209 3.365 dbcsr_matrix_vector_mult_local 304 10.0 2.776 3.268 2.778 3.270 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.638 2.724 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.199 2.719 mp_allgather_i34 111 8.7 1.004 2.543 1.004 2.543 make_images_pack 222 9.7 2.037 2.458 2.040 2.460 dbcsr_sort_data 436 11.2 1.847 2.038 1.847 2.038 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.869 1.871 rebuild_ks_matrix 3 7.3 0.000 0.000 1.856 1.858 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.856 1.858 dbcsr_data_new 4174 10.1 1.624 1.856 1.624 1.856 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.179000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1715.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 2.729767E+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 0.037 0.086 95.752 95.753 qs_energies 1 2.0 0.000 0.000 94.854 94.861 ls_scf 1 3.0 0.000 0.003 93.339 93.345 dbcsr_multiply_generic 111 6.7 0.016 0.019 77.667 78.027 ls_scf_main 1 4.0 0.000 0.002 58.302 58.307 multiply_cannon 111 7.7 0.044 0.112 53.535 57.242 multiply_cannon_loop 111 8.7 0.100 0.108 49.967 54.179 density_matrix_trs4 2 5.0 0.016 0.157 52.243 52.391 mp_waitall_1 7281 11.0 24.228 34.643 24.228 34.643 ls_scf_init_scf 1 4.0 0.000 0.003 31.494 31.496 ls_scf_init_matrix_S 1 5.0 0.000 0.001 30.278 30.347 matrix_sqrt_Newton_Schulz 2 6.5 0.002 0.004 27.792 27.806 multiply_cannon_multrec 888 9.7 12.678 15.395 21.296 24.628 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.336 23.947 make_m2s 222 7.7 0.007 0.009 17.178 18.398 make_images 222 8.7 1.986 2.304 17.140 18.361 hybrid_alltoall_any 227 10.6 0.625 2.842 9.436 10.904 make_images_data 222 9.7 0.004 0.005 9.743 10.733 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.410 8.134 9.334 mp_sum_l 887 5.1 5.410 9.329 5.410 9.329 dbcsr_mm_accdrv_process_sort 3754 11.4 7.765 8.924 7.765 8.924 multiply_cannon_sync_h2d 888 9.7 6.053 7.504 6.053 7.504 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.195 7.371 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.465 6.802 mp_irecv_dv 2335 11.1 2.451 6.755 2.451 6.755 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.781 6.378 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.948 5.143 arnoldi_extremal 4 6.8 0.000 0.000 5.094 5.110 arnoldi_normal_ev 4 7.8 0.003 0.009 5.094 5.110 build_subspace 16 8.4 0.014 0.020 4.785 4.793 calculate_norms 1584 9.8 4.302 4.704 4.302 4.704 mp_allgather_i34 111 8.7 1.403 3.898 1.403 3.898 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.455 3.794 dbcsr_matrix_vector_mult_local 304 10.0 3.054 3.639 3.056 3.641 ls_scf_post 1 4.0 0.000 0.003 3.543 3.549 ls_scf_store_result 1 5.0 0.000 0.000 3.301 3.375 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.882 2.993 make_images_sizes 222 9.7 0.000 0.000 1.090 2.329 mp_alltoall_i44 222 10.7 1.089 2.329 1.089 2.329 dbcsr_sort_data 325 11.1 1.871 2.117 1.871 2.117 make_images_pack 222 9.7 1.823 2.090 1.826 2.093 dbcsr_data_release 9322 10.9 1.328 1.965 1.328 1.965 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.920 1.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.753000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2211.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 3.323953E+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 0.039 0.080 98.435 98.436 qs_energies 1 2.0 0.000 0.000 97.301 97.307 ls_scf 1 3.0 0.000 0.001 95.371 95.377 dbcsr_multiply_generic 111 6.7 0.021 0.047 78.999 79.328 ls_scf_main 1 4.0 0.037 0.297 59.461 59.462 multiply_cannon 111 7.7 0.050 0.137 51.994 56.252 density_matrix_trs4 2 5.0 0.002 0.009 53.242 53.361 multiply_cannon_loop 111 8.7 0.115 0.127 46.973 49.709 ls_scf_init_scf 1 4.0 0.000 0.002 32.677 32.679 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.452 31.518 mp_waitall_1 6369 11.0 22.917 29.754 22.917 29.754 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.966 28.980 multiply_cannon_multrec 1332 9.7 14.344 17.351 22.310 25.174 make_m2s 222 7.7 0.008 0.009 21.250 22.656 make_images 222 8.7 3.151 3.627 21.199 22.607 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.346 16.597 make_images_data 222 9.7 0.004 0.005 11.863 13.445 hybrid_alltoall_any 227 10.6 0.799 3.712 11.193 13.076 dbcsr_mm_accdrv_process 3641 10.4 0.235 0.409 7.601 9.101 dbcsr_mm_accdrv_process_sort 3641 11.4 7.226 8.695 7.226 8.695 mp_sum_l 887 5.1 4.168 7.433 4.168 7.433 multiply_cannon_sync_h2d 1332 9.7 5.554 6.283 5.554 6.283 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.068 5.968 mp_irecv_dv 3229 10.9 2.044 5.914 2.044 5.914 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.046 5.602 arnoldi_extremal 4 6.8 0.000 0.000 5.232 5.250 arnoldi_normal_ev 4 7.8 0.001 0.004 5.232 5.250 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.545 5.081 build_subspace 16 8.4 0.014 0.021 4.877 4.884 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.727 calculate_norms 2376 9.8 4.199 4.562 4.199 4.562 mp_allgather_i34 111 8.7 2.135 4.356 2.135 4.356 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.608 3.914 dbcsr_matrix_vector_mult_local 304 10.0 3.228 3.738 3.230 3.740 dbcsr_sort_data 658 11.4 3.107 3.471 3.107 3.471 dbcsr_special_finalize 555 9.7 0.007 0.022 2.857 3.264 dbcsr_merge_single_wm 555 10.7 0.539 0.665 2.848 3.255 ls_scf_post 1 4.0 0.000 0.001 3.234 3.240 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.041 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.973 3.036 dbcsr_data_release 10477 10.7 1.590 2.396 1.590 2.396 dbcsr_finalize 304 7.8 0.050 0.061 1.817 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.436000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2713.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 4.687389E+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 0.068 0.073 100.410 100.411 qs_energies 1 2.0 0.000 0.000 99.483 99.487 ls_scf 1 3.0 0.000 0.000 97.550 97.553 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.466 78.661 ls_scf_main 1 4.0 0.000 0.000 62.357 62.359 multiply_cannon 111 7.7 0.086 0.135 55.517 61.391 density_matrix_trs4 2 5.0 0.002 0.003 55.128 55.211 multiply_cannon_loop 111 8.7 0.069 0.075 51.089 52.717 mp_waitall_1 5436 11.0 26.937 32.179 26.937 32.179 ls_scf_init_scf 1 4.0 0.000 0.000 31.248 31.250 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.953 29.998 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.006 27.469 27.478 multiply_cannon_multrec 444 9.7 13.914 16.412 21.022 22.721 make_m2s 222 7.7 0.006 0.007 18.160 20.547 make_images 222 8.7 3.729 4.434 18.097 20.485 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.234 16.530 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.338 14.141 make_images_data 222 9.7 0.003 0.004 10.169 12.705 hybrid_alltoall_any 227 10.6 0.790 3.758 10.005 12.546 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.327 6.808 7.950 dbcsr_mm_accdrv_process_sort 3003 11.4 6.494 7.603 6.494 7.603 multiply_cannon_sync_h2d 444 9.7 6.568 7.547 6.568 7.547 mp_allgather_i34 111 8.7 2.649 7.379 2.649 7.379 arnoldi_extremal 4 6.8 0.000 0.000 5.887 5.897 arnoldi_normal_ev 4 7.8 0.002 0.005 5.887 5.897 build_subspace 16 8.4 0.015 0.019 5.507 5.515 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.906 5.032 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.697 4.664 mp_irecv_dv 1241 11.2 1.677 4.635 1.677 4.635 mp_sum_l 887 5.1 2.828 4.433 2.828 4.433 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.215 4.425 dbcsr_matrix_vector_mult_local 304 10.0 3.740 4.232 3.742 4.234 ls_scf_post 1 4.0 0.000 0.000 3.944 3.948 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.645 3.735 calculate_norms 792 9.8 3.571 3.675 3.571 3.675 make_images_sizes 222 9.7 0.000 0.000 1.157 3.489 mp_alltoall_i44 222 10.7 1.157 3.489 1.157 3.489 ls_scf_store_result 1 5.0 0.000 0.000 3.378 3.412 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.965 3.263 dbcsr_finalize 304 7.8 0.062 0.078 2.202 2.290 dbcsr_merge_all 275 8.9 0.474 0.522 2.051 2.126 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.090 2.092 rebuild_ks_matrix 3 7.3 0.000 0.000 2.056 2.058 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.056 2.058 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.411000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 8.776077E+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 0.074 0.094 109.040 109.041 qs_energies 1 2.0 0.000 0.000 107.268 107.280 ls_scf 1 3.0 0.000 0.000 104.315 104.326 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.727 77.821 ls_scf_main 1 4.0 0.000 0.000 66.115 66.116 density_matrix_trs4 2 5.0 0.002 0.003 56.822 56.881 multiply_cannon 111 7.7 0.154 0.255 49.928 51.690 multiply_cannon_loop 111 8.7 0.068 0.070 46.386 47.608 ls_scf_init_scf 1 4.0 0.000 0.000 34.480 34.481 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.913 32.925 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.730 29.737 mp_waitall_1 4527 11.1 22.141 25.930 22.141 25.930 make_m2s 222 7.7 0.007 0.007 24.025 25.043 make_images 222 8.7 4.609 4.974 23.915 24.931 multiply_cannon_multrec 444 9.7 17.871 18.505 22.582 23.164 make_images_data 222 9.7 0.003 0.004 13.215 15.661 hybrid_alltoall_any 227 10.6 1.667 3.631 12.935 15.647 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.357 10.933 multiply_cannon_sync_h2d 444 9.7 8.846 8.880 8.846 8.880 arnoldi_extremal 4 6.8 0.000 0.000 7.407 7.418 arnoldi_normal_ev 4 7.8 0.003 0.009 7.407 7.418 build_subspace 16 8.4 0.026 0.036 6.854 6.862 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.500 5.665 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.392 5.655 dbcsr_matrix_vector_mult_local 304 10.0 5.068 5.388 5.070 5.391 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.223 5.316 dbcsr_mm_accdrv_process 1814 10.4 0.233 0.317 4.534 4.758 dbcsr_mm_accdrv_process_sort 1814 11.4 4.183 4.313 4.183 4.313 ls_scf_post 1 4.0 0.000 0.000 3.720 3.732 mp_allgather_i34 111 8.7 1.153 3.607 1.153 3.607 make_images_sizes 222 9.7 0.000 0.000 1.427 3.564 mp_alltoall_i44 222 10.7 1.427 3.564 1.427 3.564 ls_scf_store_result 1 5.0 0.000 0.000 3.397 3.407 calculate_norms 792 9.8 3.242 3.275 3.242 3.275 dbcsr_finalize 304 7.8 0.082 0.090 3.103 3.170 dbcsr_merge_all 275 8.9 0.893 0.916 2.887 2.949 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.923 2.923 dbcsr_complete_redistribute 5 7.6 1.440 1.476 2.778 2.891 dbcsr_data_release 12724 10.6 2.334 2.869 2.334 2.869 matrix_ls_to_qs 2 6.0 0.000 0.000 2.439 2.552 dbcsr_sort_data 325 11.1 2.450 2.502 2.450 2.502 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.473 2.475 rebuild_ks_matrix 3 7.3 0.000 0.000 2.407 2.409 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.407 2.409 dbcsr_new_transposed 4 7.5 0.252 0.267 2.315 2.332 dbcsr_frobenius_norm 74 6.6 2.064 2.141 2.202 2.239 mp_sum_l 887 5.1 1.429 2.231 1.429 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.135 2.206 dbcsr_add_anytype 103 7.2 0.860 0.891 2.135 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.041000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6873.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e112afaa9cdfdd1e40549c16c19aedffb6738e2c_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 585.695232E+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 1070783569. 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 0.023 0.128 236.849 236.851 qs_mol_dyn_low 1 2.0 0.004 0.046 235.291 235.315 qs_forces 5 3.8 0.005 0.019 235.163 235.165 qs_energies 5 4.8 0.003 0.041 232.103 232.123 scf_env_do_scf 5 5.8 0.000 0.003 214.720 214.723 scf_env_do_scf_inner_loop 105 6.6 0.002 0.012 188.111 188.112 qs_scf_new_mos 105 7.6 0.000 0.001 145.813 146.016 qs_scf_loop_do_ot 105 8.6 0.001 0.001 145.812 146.016 ot_scf_mini 105 9.6 0.003 0.005 136.110 136.232 dbcsr_multiply_generic 1445 12.2 0.125 0.134 134.180 134.522 multiply_cannon 1445 13.2 0.274 0.287 115.310 117.490 multiply_cannon_loop 1445 14.2 2.393 2.506 113.433 114.701 velocity_verlet 4 3.0 0.003 0.030 106.657 106.658 ot_mini 105 10.6 0.001 0.001 59.585 59.721 qs_ot_get_p 112 10.4 0.001 0.002 46.808 47.129 multiply_cannon_multrec 69360 15.2 31.773 37.146 39.593 45.623 mp_waitall_1 488190 16.1 34.013 41.642 34.013 41.642 qs_ot_get_derivative 55 11.6 0.001 0.002 37.868 38.014 qs_ot_p2m_diag 40 11.0 0.020 0.030 35.792 35.892 multiply_cannon_sync_h2d 69360 15.2 30.714 34.974 30.714 34.974 multiply_cannon_metrocomm3 69360 15.2 0.197 0.207 24.675 32.939 cp_dbcsr_syevd 40 12.0 0.002 0.003 32.614 32.615 rebuild_ks_matrix 110 8.4 0.000 0.000 29.708 29.892 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.017 29.707 29.892 qs_ks_update_qs_env 112 7.6 0.001 0.001 27.277 27.442 cp_fm_syevd 40 13.0 0.000 0.001 27.113 27.276 init_scf_loop 7 6.6 0.000 0.006 26.577 26.578 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.932 23.146 apply_single 62 13.6 0.000 0.000 22.932 23.146 cp_fm_redistribute_end 40 14.0 11.190 22.328 11.195 22.329 cp_fm_syevd_base 40 14.0 11.125 22.263 11.125 22.263 prepare_preconditioner 7 7.6 0.000 0.001 21.663 21.699 make_preconditioner 7 8.6 0.000 0.003 21.663 21.699 ot_new_cg_direction 55 11.6 0.001 0.001 20.954 20.954 qs_rho_update_rho_low 110 7.6 0.001 0.011 19.001 19.387 calculate_rho_elec 110 8.6 0.036 0.389 19.000 19.387 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.502 14.715 init_scf_run 5 5.8 0.000 0.003 14.654 14.655 scf_env_initial_rho_setup 5 6.8 0.006 0.374 14.654 14.655 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.539 14.598 rs_pw_transfer 690 11.5 0.011 0.013 13.189 14.265 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.653 13.800 density_rs2pw 110 9.6 0.006 0.007 12.670 13.796 pw_transfer 1645 12.4 0.083 0.105 12.235 12.455 mp_sum_l 4764 12.2 11.758 12.398 11.758 12.398 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.093 12.319 calculate_dm_sparse 110 9.5 0.000 0.001 11.311 11.470 qs_vxc_create 110 10.4 0.002 0.008 10.492 10.549 fft_wrap_pw1pw2_240 915 15.0 1.192 1.280 10.277 10.509 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.156 10.230 cp_fm_cholesky_invert 7 10.6 10.124 10.133 10.124 10.133 dbcsr_mm_accdrv_process 154766 15.8 4.014 4.151 7.687 9.847 check_diag 80 13.5 8.589 8.868 9.706 9.847 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 9.710 9.727 acc_transpose_blocks 69360 15.2 0.364 0.385 8.933 9.411 calculate_first_density_matrix 1 7.0 0.000 0.004 9.327 9.335 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.237 9.290 sum_up_and_integrate 60 10.3 0.028 0.031 9.232 9.244 integrate_v_rspace 60 11.3 0.002 0.003 9.204 9.216 fft3d_pb 915 16.0 2.390 2.634 8.320 8.632 xc_rho_set_and_dset_create 110 12.4 0.077 0.096 7.350 7.593 acc_transpose_blocks_kernels 69360 16.2 0.852 0.893 6.998 7.399 make_m2s 2890 13.2 0.081 0.093 6.640 7.303 make_images 2890 14.2 0.241 0.262 6.530 7.195 xc_vxc_pw_create 60 11.3 0.039 0.049 7.027 7.083 make_full_single_inverse 7 9.6 0.001 0.002 6.853 6.888 jit_kernel_transpose 5 15.0 6.146 6.523 6.146 6.523 multiply_cannon_metrocomm1 69360 15.2 0.093 0.098 4.273 6.521 xc_pw_derive 510 13.4 0.005 0.006 6.269 6.332 mp_waitany 7680 13.5 4.506 5.825 4.506 5.825 mp_alltoall_z22v 2340 17.7 5.406 5.745 5.406 5.745 potential_pw2rs 60 12.3 0.003 0.003 5.387 5.428 multiply_cannon_metrocomm4 67915 15.2 0.182 0.207 2.065 5.109 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=236.851000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=558.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e112afaa9cdfdd1e40549c16c19aedffb6738e2c Summary: empty Status: OK