=== 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: b263019182a81f50cd70653ab7c48fb92fe52198 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, # Cray-FFTW 3.3.8.10, COSMA 2.6.6, ELPA 2023.05.001, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, 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: 04.10.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/01 job id: 50294935 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/02 job id: 50294936 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/03 job id: 50294938 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/04 job id: 50294939 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/05 job id: 50294940 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/06 job id: 50294941 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/07 job id: 50294942 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/08 job id: 50294943 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/09 job id: 50294944 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/10 job id: 50294945 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/11 job id: 50294946 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/12 job id: 50294947 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/13 job id: 50294948 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/14 job id: 50294949 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/15 job id: 50294950 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/16 job id: 50294952 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/17 job id: 50294953 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/18 job id: 50294954 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/19 job id: 50294955 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/20 job id: 50294956 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/21 job id: 50294957 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/22 job id: 50294958 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/23 job id: 50294959 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/24 job id: 50294960 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/25 job id: 50294961 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/26 job id: 50294962 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/27 job id: 50294963 --- 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/b263019182a81f50cd70653ab7c48fb92fe52198_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 0.020 0.036 136.843 136.844 farming_run 1 2.0 135.825 135.826 136.809 136.811 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488081E+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 0.007 0.027 115.166 115.166 qs_energies 1 2.0 0.000 0.000 114.923 114.925 mp2_main 1 3.0 0.000 0.000 112.673 112.675 mp2_gpw_main 1 4.0 0.042 0.054 111.529 111.531 mp2_ri_gpw_compute_in 1 5.0 0.202 0.215 92.503 92.799 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.585 54.869 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.165 41.005 46.024 get_2c_integrals 1 6.0 0.008 0.009 36.873 37.713 integrate_v_rspace 273 8.0 0.436 0.454 24.619 29.525 pw_transfer 6555 10.6 0.381 0.397 26.825 27.276 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 25.446 25.991 grid_integrate_task_list 273 9.0 20.530 25.891 20.530 25.891 fft_wrap_pw1pw2_100 2178 12.4 0.028 0.030 22.988 23.520 compute_2c_integrals 1 7.0 0.002 0.002 19.223 19.269 rpa_ri_compute_en 1 5.0 0.019 0.022 18.907 19.082 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.635 18.908 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.431 18.633 18.906 cp_fm_cholesky_decompose 12 8.2 17.680 18.550 17.680 18.550 cholesky_decomp 1 7.0 0.000 0.000 16.495 17.376 fft3d_s 5443 13.4 16.140 16.530 16.161 16.552 ao_to_mo_and_store_B_mult_1 272 7.0 10.756 15.306 10.756 15.306 calculate_wavefunction 272 8.0 5.388 5.508 12.297 12.998 rpa_num_int 1 6.0 0.000 0.001 10.765 10.775 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.734 10.753 calc_mat_Q 8 8.0 0.000 0.000 9.502 9.600 contract_S_to_Q 8 9.0 0.000 0.000 8.922 9.023 calc_potential_gpw 544 9.5 0.004 0.005 8.239 8.672 parallel_gemm_fm 14 9.1 0.000 0.000 8.498 8.606 parallel_gemm_fm_cosma 14 10.1 8.498 8.606 8.498 8.606 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.088 8.346 potential_pw2rs 545 10.0 0.107 0.109 7.490 8.097 create_integ_mat 1 6.0 0.016 0.034 7.764 7.774 collocate_single_gaussian 272 10.0 0.039 0.041 7.352 7.568 array2fm 1 7.0 0.000 0.000 6.627 7.209 pw_scatter_s 2720 13.7 4.334 4.431 4.334 4.431 pw_gather_s 2722 13.2 3.424 3.758 3.424 3.758 array2fm_buffer_send 1 8.0 2.924 3.126 2.924 3.126 pw_poisson_solve 545 10.5 1.126 1.193 2.207 2.353 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.530994, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 0.025 0.039 423.760 423.761 farming_run 1 2.0 422.730 422.734 423.720 423.722 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 897827141120 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 249788822 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 1.247687E+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 0.012 0.033 218.579 218.580 qs_energies 1 2.0 0.000 0.000 218.102 218.112 scf_env_do_scf 1 3.0 0.000 0.000 115.361 115.362 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.389 114.397 rebuild_ks_matrix 4 6.0 0.000 0.000 114.388 114.396 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.063 114.388 114.396 hfx_ks_matrix 4 8.0 0.001 0.001 114.032 114.037 integrate_four_center 4 9.0 0.153 0.471 114.031 114.036 integrate_four_center_main 4 10.0 0.129 0.491 102.119 105.413 integrate_four_center_bin 264 11.0 101.990 105.191 101.990 105.191 mp2_main 1 3.0 0.000 0.000 102.440 102.450 mp2_gpw_main 1 4.0 0.050 0.070 101.467 101.480 init_scf_loop 1 4.0 0.000 0.000 97.125 97.125 mp2_ri_gpw_compute_in 1 5.0 0.071 0.085 74.469 75.630 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.962 55.120 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.156 41.606 46.515 integrate_v_rspace 95 8.0 0.396 0.562 28.095 32.880 pw_transfer 2240 10.6 0.146 0.180 29.476 29.985 fft_wrap_pw1pw2 1868 11.4 0.017 0.023 28.485 28.962 mp2_ri_gpw_compute_en 1 5.0 0.060 0.072 26.847 28.734 grid_integrate_task_list 95 9.0 23.509 28.497 23.509 28.497 ao_to_mo_and_store_B_mult_1 91 7.0 10.678 26.971 10.678 26.971 fft_wrap_pw1pw2_100 730 12.4 0.012 0.014 26.238 26.712 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.845 1.894 25.016 25.026 get_2c_integrals 1 6.0 0.000 0.000 20.394 20.444 compute_2c_integrals 1 7.0 0.002 0.003 19.373 19.389 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.581 18.902 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.897 18.580 18.902 fft3d_s 1823 13.4 18.505 18.828 18.519 18.841 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.233 18.233 calculate_wavefunction 91 8.0 2.007 2.035 9.612 9.875 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.582 8.749 9.291 potential_pw2rs 186 10.0 0.033 0.034 8.415 8.987 local_gemm 172 8.0 8.192 8.739 8.192 8.739 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.123 8.491 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.521 8.037 8.391 calc_potential_gpw 182 9.5 0.002 0.002 7.885 8.172 collocate_single_gaussian 91 10.0 0.017 0.022 7.735 7.956 mp_sync 37 10.5 3.904 7.864 3.904 7.864 integrate_four_center_load 4 10.0 0.000 0.000 6.747 6.751 hfx_load_balance 1 11.0 0.000 0.000 6.747 6.751 mp_sendrecv_dm3 2068 8.0 6.075 6.436 6.075 6.436 mp2_ri_gpw_compute_en_ener 172 7.0 6.339 6.390 6.339 6.390 pw_gather_s 912 13.2 4.461 5.022 4.461 5.022 pw_scatter_s 910 13.7 3.857 4.382 3.857 4.382 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.461238, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 452.927488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 60.004 60.006 qs_mol_dyn_low 1 2.0 0.003 0.003 55.931 59.675 qs_forces 11 3.9 0.002 0.003 55.826 55.827 qs_energies 11 4.9 0.002 0.002 54.301 54.315 scf_env_do_scf 11 5.9 0.000 0.001 47.841 47.841 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 45.673 45.674 dbcsr_multiply_generic 2286 12.5 0.098 0.105 34.516 35.157 qs_scf_new_mos 108 7.5 0.000 0.001 34.793 35.045 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.792 35.045 ot_scf_mini 108 9.5 0.002 0.002 33.028 33.230 multiply_cannon 2286 13.5 0.190 0.204 26.907 28.652 multiply_cannon_loop 2286 14.5 1.792 1.904 26.166 27.863 velocity_verlet 10 3.0 0.002 0.002 27.275 27.276 ot_mini 108 10.5 0.001 0.001 19.792 20.027 qs_ot_get_derivative 108 11.5 0.001 0.001 16.778 16.968 mp_waitall_1 245248 16.5 8.881 15.614 8.881 15.614 multiply_cannon_metrocomm3 54864 15.5 0.073 0.077 6.218 13.237 multiply_cannon_multrec 54864 15.5 3.715 5.704 8.123 11.662 qs_ot_get_p 119 10.4 0.001 0.002 8.376 8.654 rebuild_ks_matrix 119 8.3 0.000 0.000 8.054 8.214 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.053 8.213 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.088 7.221 mp_sum_l 7287 12.8 5.405 7.139 5.405 7.139 multiply_cannon_sync_h2d 54864 15.5 5.153 6.770 5.153 6.770 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.812 6.274 dbcsr_mm_accdrv_process 76910 16.1 1.768 2.819 4.321 6.193 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.382 5.425 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.164 5.262 init_scf_run 11 5.9 0.000 0.001 5.147 5.148 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.147 5.147 qs_rho_update_rho_low 119 7.7 0.000 0.000 4.541 4.647 calculate_rho_elec 119 8.7 0.011 0.017 4.541 4.646 sum_up_and_integrate 119 10.3 0.002 0.002 4.617 4.624 integrate_v_rspace 119 11.3 0.002 0.002 4.606 4.614 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.563 4.563 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.327 4.328 cp_fm_redistribute_end 50 14.0 2.205 4.291 2.214 4.295 cp_fm_diag_elpa_base 50 14.0 2.074 4.169 2.079 4.180 md_write_output 11 3.9 0.001 0.036 0.041 3.825 md_output 10 3.0 0.000 0.000 0.042 3.824 write_trajectory 44 4.9 0.001 0.009 0.040 3.788 mp_max_l 33 2.8 3.746 3.785 3.746 3.785 write_particle_coordinates 11 5.9 0.039 3.780 0.039 3.780 calculate_dm_sparse 119 9.5 0.000 0.000 3.155 3.295 jit_kernel_multiply 13 15.8 2.493 3.289 2.493 3.289 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.928 3.168 apply_single 119 13.6 0.000 0.000 2.928 3.167 multiply_cannon_metrocomm1 54864 15.5 0.056 0.061 1.754 3.034 ot_diis_step 108 11.5 0.006 0.006 2.742 2.742 acc_transpose_blocks 54864 15.5 0.218 0.241 2.200 2.738 calculate_first_density_matrix 1 7.0 0.000 0.001 2.722 2.727 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.609 2.695 density_rs2pw 119 9.7 0.004 0.004 2.520 2.646 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.382 2.384 wfi_extrapolate 11 7.9 0.001 0.001 2.189 2.189 init_scf_loop 11 6.9 0.000 0.000 2.146 2.146 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.079 2.132 pw_transfer 1439 11.6 0.052 0.057 2.037 2.110 grid_integrate_task_list 119 12.3 2.013 2.105 2.013 2.105 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.962 2.036 mp_sum_d 4137 12.0 1.376 2.013 1.376 2.013 potential_pw2rs 119 12.3 0.004 0.004 1.968 1.986 make_m2s 4572 13.5 0.053 0.055 1.884 1.932 make_images 4572 14.5 0.133 0.139 1.803 1.851 fft3d_ps 1201 14.6 0.373 0.480 1.676 1.747 transfer_rs2pw 487 10.6 0.005 0.006 1.459 1.610 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.560 1.590 mp_alltoall_d11v 2130 13.8 1.446 1.588 1.446 1.588 mp_waitany 12084 13.8 1.274 1.503 1.274 1.503 fft_wrap_pw1pw2_140 487 13.2 0.107 0.113 1.390 1.464 transfer_pw2rs 487 13.2 0.006 0.006 1.413 1.429 grid_collocate_task_list 119 9.7 1.347 1.416 1.347 1.416 acc_transpose_blocks_sync 164592 16.5 1.163 1.410 1.163 1.410 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.006000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.454545, yerr=0.890724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 488.960000E+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 1031941. 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 0.014 0.046 43.591 43.593 qs_mol_dyn_low 1 2.0 0.004 0.013 43.183 43.192 qs_forces 11 3.9 0.002 0.003 43.109 43.112 qs_energies 11 4.9 0.002 0.002 41.003 41.008 scf_env_do_scf 11 5.9 0.000 0.001 35.435 35.436 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 32.521 32.521 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.303 23.776 qs_scf_new_mos 108 7.5 0.001 0.001 22.815 23.043 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.814 23.043 ot_scf_mini 108 9.5 0.002 0.003 21.841 22.003 velocity_verlet 10 3.0 0.001 0.002 20.294 20.295 multiply_cannon 2286 13.5 0.208 0.216 17.370 19.119 multiply_cannon_loop 2286 14.5 1.185 1.250 16.017 17.870 ot_mini 108 10.5 0.001 0.001 13.147 13.381 mp_waitall_1 200699 16.5 6.301 11.500 6.301 11.500 qs_ot_get_derivative 108 11.5 0.001 0.001 10.695 10.860 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.285 9.793 multiply_cannon_multrec 27432 15.5 1.817 4.132 6.203 9.106 rebuild_ks_matrix 119 8.3 0.000 0.000 7.466 7.632 qs_ks_build_kohn_sham_matrix 119 9.3 0.022 0.093 7.465 7.631 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.620 6.771 dbcsr_mm_accdrv_process 47894 16.0 3.531 5.713 4.305 6.492 qs_ot_get_p 119 10.4 0.001 0.001 5.622 5.848 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.071 4.908 mp_sum_l 7287 12.8 2.410 4.555 2.410 4.555 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.220 4.258 calculate_rho_elec 119 8.7 0.021 0.024 4.220 4.258 sum_up_and_integrate 119 10.3 0.002 0.002 4.239 4.245 integrate_v_rspace 119 11.3 0.002 0.003 4.225 4.231 init_scf_run 11 5.9 0.000 0.001 4.223 4.224 scf_env_initial_rho_setup 11 6.9 0.011 0.027 4.223 4.224 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.037 4.132 apply_single 119 13.6 0.000 0.000 3.036 4.132 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.699 3.726 make_m2s 4572 13.5 0.052 0.053 3.084 3.456 make_images 4572 14.5 0.206 0.247 2.994 3.369 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.258 3.259 init_scf_loop 11 6.9 0.000 0.000 2.883 2.884 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.781 2.782 cp_fm_redistribute_end 50 14.0 1.404 2.731 1.409 2.734 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.610 2.696 cp_fm_diag_elpa_base 50 14.0 1.285 2.608 1.316 2.644 density_rs2pw 119 9.7 0.004 0.004 2.495 2.599 calculate_first_density_matrix 1 7.0 0.000 0.002 2.481 2.492 ot_diis_step 108 11.5 0.011 0.011 2.381 2.382 pw_transfer 1439 11.6 0.065 0.069 2.218 2.284 calculate_dm_sparse 119 9.5 0.000 0.001 2.178 2.261 multiply_cannon_sync_h2d 27432 15.5 1.650 2.250 1.650 2.250 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.126 2.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.065 2.068 jit_kernel_multiply 10 16.3 0.716 2.049 0.716 2.049 acc_transpose_blocks 27432 15.5 0.114 0.119 1.643 1.977 potential_pw2rs 119 12.3 0.006 0.006 1.950 1.972 grid_integrate_task_list 119 12.3 1.819 1.913 1.819 1.913 prepare_preconditioner 11 7.9 0.000 0.000 1.837 1.862 make_preconditioner 11 8.9 0.000 0.000 1.837 1.862 make_images_data 4572 15.5 0.047 0.054 1.369 1.839 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.782 1.826 fft3d_ps 1201 14.6 0.525 0.583 1.759 1.819 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.729 1.785 hybrid_alltoall_any 4725 16.4 0.055 0.118 1.227 1.706 wfi_extrapolate 11 7.9 0.001 0.001 1.663 1.663 mp_alltoall_d11v 2130 13.8 1.488 1.659 1.488 1.659 transfer_rs2pw 487 10.6 0.005 0.005 1.487 1.607 fft_wrap_pw1pw2_140 487 13.2 0.127 0.131 1.531 1.592 mp_sum_d 4137 12.0 1.018 1.542 1.018 1.542 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.312 1.362 grid_collocate_task_list 119 9.7 1.289 1.338 1.289 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.325 1.336 transfer_pw2rs 487 13.2 0.004 0.005 1.266 1.291 mp_allgather_i34 2286 14.5 0.777 1.249 0.777 1.249 make_images_sizes 4572 15.5 0.005 0.005 0.851 1.142 mp_alltoall_i44 4572 16.5 0.846 1.138 0.846 1.138 build_core_hamiltonian_matrix_ 11 4.9 0.029 0.230 0.748 1.115 mp_alltoall_z22v 1201 16.6 0.920 1.001 0.920 1.001 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.967 0.970 acc_transpose_blocks_sync 82296 16.5 0.813 0.939 0.813 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.908 0.922 acc_transpose_blocks_kernels 27432 16.5 0.188 0.278 0.691 0.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.593000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.880742 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 521.777152E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.035 37.571 37.573 qs_mol_dyn_low 1 2.0 0.003 0.003 36.656 36.666 qs_forces 11 3.9 0.023 0.033 36.577 36.578 qs_energies 11 4.9 0.004 0.006 34.933 34.955 scf_env_do_scf 11 5.9 0.000 0.001 29.394 29.394 scf_env_do_scf_inner_loop 108 6.5 0.005 0.098 26.383 26.383 dbcsr_multiply_generic 2286 12.5 0.097 0.100 18.942 19.039 qs_scf_new_mos 108 7.5 0.001 0.001 17.642 17.660 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.642 17.659 ot_scf_mini 108 9.5 0.002 0.002 16.854 16.869 velocity_verlet 10 3.0 0.001 0.001 16.767 16.769 multiply_cannon 2286 13.5 0.194 0.199 14.809 15.528 multiply_cannon_loop 2286 14.5 0.857 0.892 13.939 14.692 ot_mini 108 10.5 0.001 0.001 10.332 10.349 qs_ot_get_derivative 108 11.5 0.001 0.001 8.780 8.797 multiply_cannon_multrec 18288 15.5 2.009 3.030 7.347 7.677 rebuild_ks_matrix 119 8.3 0.000 0.000 6.525 6.539 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.525 6.538 dbcsr_mm_accdrv_process 38222 16.0 5.165 6.188 5.244 6.251 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.762 5.776 mp_waitall_1 158411 16.6 3.472 4.600 3.472 4.600 qs_ot_get_p 119 10.4 0.002 0.005 4.429 4.487 init_scf_run 11 5.9 0.000 0.001 4.128 4.128 scf_env_initial_rho_setup 11 6.9 0.016 0.026 4.127 4.128 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.463 4.093 sum_up_and_integrate 119 10.3 0.001 0.002 4.033 4.037 integrate_v_rspace 119 11.3 0.002 0.003 4.020 4.026 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.926 3.931 calculate_rho_elec 119 8.7 0.031 0.032 3.926 3.931 init_scf_loop 11 6.9 0.000 0.000 2.986 2.988 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.809 2.816 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.781 2.777 calculate_first_density_matrix 1 7.0 0.000 0.000 2.727 2.743 density_rs2pw 119 9.7 0.004 0.004 2.370 2.547 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.480 2.480 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.089 2.433 apply_single 119 13.6 0.000 0.000 2.089 2.433 make_m2s 4572 13.5 0.044 0.045 2.246 2.380 make_images 4572 14.5 0.192 0.203 2.159 2.294 pw_transfer 1439 11.6 0.065 0.068 2.201 2.223 acc_transpose_blocks 18288 15.5 0.079 0.082 1.927 2.174 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.139 2.150 cp_fm_diag_elpa_base 50 14.0 2.108 2.129 2.136 2.148 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.108 2.130 calculate_dm_sparse 119 9.5 0.000 0.001 2.064 2.069 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.006 2.007 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.892 1.901 prepare_preconditioner 11 7.9 0.000 0.000 1.898 1.899 make_preconditioner 11 8.9 0.000 0.000 1.898 1.899 grid_integrate_task_list 119 12.3 1.802 1.885 1.802 1.885 mp_sum_l 7287 12.8 1.497 1.861 1.497 1.861 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.744 1.831 potential_pw2rs 119 12.3 0.007 0.008 1.786 1.797 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.782 1.788 fft3d_ps 1201 14.6 0.551 0.571 1.664 1.686 fft_wrap_pw1pw2_140 487 13.2 0.179 0.183 1.582 1.603 ot_diis_step 108 11.5 0.011 0.012 1.533 1.534 transfer_rs2pw 487 10.6 0.005 0.005 1.346 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.380 1.385 wfi_extrapolate 11 7.9 0.001 0.001 1.324 1.325 grid_collocate_task_list 119 9.7 1.234 1.323 1.234 1.323 acc_transpose_blocks_kernels 18288 16.5 0.216 0.223 1.094 1.314 multiply_cannon_sync_h2d 18288 15.5 1.026 1.169 1.026 1.169 transfer_pw2rs 487 13.2 0.004 0.004 1.127 1.139 make_images_data 4572 15.5 0.047 0.050 0.943 1.122 qs_energies_init_hamiltonians 11 5.9 0.024 0.035 1.089 1.111 jit_kernel_transpose 5 15.6 0.878 1.092 0.878 1.092 mp_alltoall_d11v 2130 13.8 0.955 1.076 0.955 1.076 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.014 1.033 hybrid_alltoall_any 4725 16.4 0.060 0.118 0.813 1.012 mp_alltoall_z22v 1201 16.6 0.931 1.003 0.931 1.003 cp_fm_cholesky_invert 11 10.9 0.841 0.846 0.841 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.841 acc_transpose_blocks_sync 54864 16.5 0.734 0.789 0.734 0.789 multiply_cannon_metrocomm1 18288 15.5 0.030 0.031 0.478 0.781 make_images_sizes 4572 15.5 0.005 0.005 0.559 0.764 mp_alltoall_i44 4572 16.5 0.554 0.760 0.554 0.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=37.573000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=1.443137 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 555.438080E+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 0.007 0.026 37.528 37.529 qs_mol_dyn_low 1 2.0 0.003 0.003 37.349 37.357 qs_forces 11 3.9 0.002 0.002 37.271 37.272 qs_energies 11 4.9 0.002 0.002 35.555 35.560 scf_env_do_scf 11 5.9 0.000 0.001 29.800 29.801 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.225 26.225 dbcsr_multiply_generic 2286 12.5 0.102 0.105 20.667 20.785 velocity_verlet 10 3.0 0.001 0.001 18.644 18.646 qs_scf_new_mos 108 7.5 0.001 0.001 18.007 18.065 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.006 18.065 multiply_cannon 2286 13.5 0.218 0.225 16.946 17.774 ot_scf_mini 108 9.5 0.002 0.003 16.968 17.024 multiply_cannon_loop 2286 14.5 1.550 1.630 15.908 16.595 ot_mini 108 10.5 0.001 0.001 10.389 10.455 multiply_cannon_multrec 27432 15.5 2.449 3.102 9.127 9.554 qs_ot_get_derivative 108 11.5 0.001 0.001 8.523 8.580 dbcsr_mm_accdrv_process 47916 15.9 6.084 7.564 6.576 8.232 rebuild_ks_matrix 119 8.3 0.000 0.000 6.402 6.454 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 6.402 6.454 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.673 5.717 init_scf_run 11 5.9 0.000 0.001 4.416 4.416 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.416 4.416 qs_ot_get_p 119 10.4 0.033 0.136 3.866 3.939 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.220 3.572 sum_up_and_integrate 119 10.3 0.001 0.001 3.560 3.567 integrate_v_rspace 119 11.3 0.003 0.003 3.549 3.556 init_scf_loop 11 6.9 0.000 0.000 3.553 3.554 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.445 3.489 calculate_rho_elec 119 8.7 0.040 0.046 3.444 3.488 acc_transpose_blocks 27432 15.5 0.120 0.124 2.923 3.087 calculate_first_density_matrix 1 7.0 0.000 0.000 3.009 3.013 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.056 2.660 apply_single 119 13.6 0.000 0.000 2.056 2.660 prepare_preconditioner 11 7.9 0.000 0.000 2.554 2.561 make_preconditioner 11 8.9 0.000 0.000 2.554 2.561 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.164 2.491 make_m2s 4572 13.5 0.054 0.055 2.338 2.475 mp_waitall_1 137007 16.6 1.940 2.428 1.940 2.428 calculate_dm_sparse 119 9.5 0.000 0.000 2.347 2.400 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.362 2.391 make_images 4572 14.5 0.273 0.334 2.230 2.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.360 2.361 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.265 2.277 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.186 2.199 pw_transfer 1439 11.6 0.065 0.069 2.031 2.066 density_rs2pw 119 9.7 0.004 0.004 1.910 2.024 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.939 1.977 grid_integrate_task_list 119 12.3 1.836 1.911 1.836 1.911 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.852 1.853 ot_diis_step 108 11.5 0.012 0.012 1.825 1.825 mp_sum_l 7287 12.8 0.954 1.718 0.954 1.718 fft_wrap_pw1pw2_140 487 13.2 0.227 0.233 1.597 1.636 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.599 1.609 cp_fm_diag_elpa_base 50 14.0 1.565 1.582 1.597 1.607 acc_transpose_blocks_sync 82296 16.5 1.472 1.595 1.472 1.595 fft3d_ps 1201 14.6 0.592 0.651 1.427 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.438 1.449 jit_kernel_multiply 6 16.1 0.428 1.439 0.428 1.439 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 1.303 1.384 multiply_cannon_metrocomm3 27432 15.5 0.040 0.041 0.879 1.363 potential_pw2rs 119 12.3 0.008 0.009 1.358 1.362 grid_collocate_task_list 119 9.7 1.247 1.359 1.247 1.359 wfi_extrapolate 11 7.9 0.001 0.001 1.352 1.352 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.311 1.326 cp_fm_upper_to_full 72 14.2 0.817 1.163 0.817 1.163 jit_kernel_transpose 5 15.6 1.033 1.107 1.033 1.107 transfer_rs2pw 487 10.6 0.004 0.005 0.963 1.078 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.070 1.071 dbcsr_complete_redistribute 329 12.2 0.117 0.148 0.767 1.045 make_images_data 4572 15.5 0.048 0.053 0.876 1.010 hybrid_alltoall_any 4725 16.4 0.068 0.157 0.750 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.800 0.878 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.562 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.823 mp_alltoall_d11v 2130 13.8 0.731 0.822 0.731 0.822 transfer_pw2rs 487 13.2 0.004 0.004 0.779 0.782 cp_fm_cholesky_invert 11 10.9 0.757 0.760 0.757 0.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.529000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.636364, yerr=3.023325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 621.805568E+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 0.029 0.034 31.190 31.191 qs_mol_dyn_low 1 2.0 0.003 0.003 30.741 30.751 qs_forces 11 3.9 0.002 0.003 30.610 30.611 qs_energies 11 4.9 0.002 0.002 28.882 28.885 scf_env_do_scf 11 5.9 0.000 0.001 23.570 23.570 scf_env_do_scf_inner_loop 108 6.5 0.016 0.223 20.468 20.469 velocity_verlet 10 3.0 0.069 0.080 15.275 15.277 dbcsr_multiply_generic 2286 12.5 0.094 0.098 13.705 13.826 qs_scf_new_mos 108 7.5 0.001 0.001 12.360 12.389 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.359 12.389 ot_scf_mini 108 9.5 0.002 0.002 11.642 11.665 multiply_cannon 2286 13.5 0.223 0.231 10.862 11.356 multiply_cannon_loop 2286 14.5 0.642 0.660 9.910 10.110 ot_mini 108 10.5 0.001 0.001 6.697 6.725 rebuild_ks_matrix 119 8.3 0.000 0.000 6.353 6.371 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.352 6.371 multiply_cannon_multrec 9144 15.5 1.688 1.877 6.048 6.321 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.697 5.714 qs_ot_get_derivative 108 11.5 0.001 0.001 5.374 5.399 dbcsr_mm_accdrv_process 12550 15.8 3.612 4.331 4.250 4.373 init_scf_run 11 5.9 0.000 0.001 3.778 3.778 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.778 3.778 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.688 3.709 calculate_rho_elec 119 8.7 0.060 0.061 3.688 3.708 sum_up_and_integrate 119 10.3 0.002 0.007 3.542 3.546 integrate_v_rspace 119 11.3 0.003 0.003 3.531 3.535 qs_ot_get_p 119 10.4 0.002 0.002 3.067 3.112 init_scf_loop 11 6.9 0.000 0.000 3.079 3.080 calculate_first_density_matrix 1 7.0 0.001 0.009 2.591 2.592 mp_waitall_1 115863 16.7 1.762 2.308 1.762 2.308 pw_transfer 1439 11.6 0.066 0.067 2.214 2.223 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.120 2.131 make_m2s 4572 13.5 0.033 0.034 1.920 2.066 make_images 4572 14.5 0.268 0.300 1.831 1.975 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.971 1.973 density_rs2pw 119 9.7 0.003 0.003 1.895 1.972 grid_integrate_task_list 119 12.3 1.868 1.936 1.868 1.936 prepare_preconditioner 11 7.9 0.000 0.000 1.839 1.844 make_preconditioner 11 8.9 0.000 0.000 1.838 1.844 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.811 1.834 calculate_dm_sparse 119 9.5 0.000 0.000 1.778 1.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.764 1.766 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.728 1.754 fft_wrap_pw1pw2_140 487 13.2 0.324 0.330 1.734 1.748 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.719 1.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.597 1.605 acc_transpose_blocks 9144 15.5 0.043 0.044 1.488 1.530 fft3d_ps 1201 14.6 0.652 0.665 1.473 1.486 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.433 1.441 cp_fm_diag_elpa_base 50 14.0 1.405 1.421 1.431 1.439 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.393 1.407 grid_collocate_task_list 119 9.7 1.297 1.372 1.297 1.372 potential_pw2rs 119 12.3 0.010 0.011 1.340 1.344 ot_diis_step 108 11.5 0.013 0.013 1.307 1.307 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.268 1.303 apply_single 119 13.6 0.000 0.000 1.268 1.303 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.260 1.260 jit_kernel_multiply 6 15.5 0.598 1.206 0.598 1.206 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.163 1.169 wfi_extrapolate 11 7.9 0.001 0.001 1.132 1.132 mp_alltoall_d11v 2130 13.8 0.989 1.062 0.989 1.062 hybrid_alltoall_any 4725 16.4 0.066 0.179 0.814 1.030 make_images_data 4572 15.5 0.042 0.045 0.851 1.013 cp_fm_cholesky_invert 11 10.9 0.928 0.931 0.928 0.931 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.877 0.925 transfer_rs2pw 487 10.6 0.004 0.004 0.850 0.924 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.850 0.856 qs_vxc_create 119 10.3 0.352 0.402 0.788 0.796 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.456 0.791 xc_vxc_pw_create 119 11.3 0.059 0.061 0.436 0.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.780 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.706 0.755 acc_transpose_blocks_sync 27432 16.5 0.717 0.740 0.717 0.740 mp_allgather_i34 2286 14.5 0.269 0.739 0.269 0.739 mp_alltoall_z22v 1201 16.6 0.694 0.734 0.694 0.734 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.713 0.733 mp_sum_d 4135 12.0 0.322 0.725 0.322 0.725 transfer_pw2rs 487 13.2 0.003 0.004 0.701 0.704 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.430 0.662 mp_sum_l 7287 12.8 0.476 0.636 0.476 0.636 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.191000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=589.272727, yerr=5.986210 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 772.603904E+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 0.033 43.344 43.345 qs_mol_dyn_low 1 2.0 0.003 0.003 43.140 43.148 qs_forces 11 3.9 0.002 0.003 43.075 43.076 qs_energies 11 4.9 0.020 0.031 41.082 41.085 scf_env_do_scf 11 5.9 0.001 0.001 34.956 34.956 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.129 27.131 velocity_verlet 10 3.0 0.002 0.002 23.906 23.912 dbcsr_multiply_generic 2286 12.5 0.104 0.106 19.214 19.444 qs_scf_new_mos 108 7.5 0.001 0.001 17.478 17.577 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.478 17.576 ot_scf_mini 108 9.5 0.002 0.002 16.321 16.421 multiply_cannon 2286 13.5 0.297 0.305 15.156 16.292 multiply_cannon_loop 2286 14.5 0.859 0.879 13.919 15.088 ot_mini 108 10.5 0.001 0.001 10.115 10.230 multiply_cannon_multrec 9144 15.5 3.395 4.813 8.882 9.211 qs_ot_get_derivative 108 11.5 0.001 0.001 8.079 8.180 init_scf_loop 11 6.9 0.000 0.000 7.801 7.803 rebuild_ks_matrix 119 8.3 0.000 0.000 7.241 7.383 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.241 7.382 dbcsr_mm_accdrv_process 12550 15.8 4.564 6.546 5.349 6.876 prepare_preconditioner 11 7.9 0.000 0.000 6.730 6.743 make_preconditioner 11 8.9 0.000 0.000 6.730 6.743 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.519 6.648 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.274 6.625 cp_fm_upper_to_full 72 14.2 3.152 4.557 3.152 4.557 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.193 4.198 calculate_rho_elec 119 8.7 0.117 0.120 4.193 4.198 sum_up_and_integrate 119 10.3 0.001 0.001 3.844 3.850 integrate_v_rspace 119 11.3 0.003 0.003 3.834 3.839 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.996 3.625 init_scf_run 11 5.9 0.000 0.001 3.582 3.582 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.581 3.581 mp_waitall_1 94719 16.7 2.261 3.550 2.261 3.550 qs_ot_get_p 119 10.4 0.032 0.032 3.408 3.543 pw_transfer 1439 11.6 0.068 0.069 3.009 3.014 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.411 2.916 apply_single 119 13.6 0.000 0.000 2.411 2.916 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.910 2.916 dbcsr_complete_redistribute 329 12.2 0.282 0.288 1.979 2.781 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.378 2.568 fft_wrap_pw1pw2_140 487 13.2 0.648 0.653 2.489 2.497 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.684 2.486 make_m2s 4572 13.5 0.037 0.037 2.303 2.447 make_images 4572 14.5 0.356 0.393 2.182 2.327 mp_alltoall_i22 627 13.8 1.430 2.271 1.430 2.271 calculate_dm_sparse 119 9.5 0.000 0.000 2.232 2.248 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.450 2.246 density_rs2pw 119 9.7 0.003 0.004 2.214 2.232 qs_energies_init_hamiltonians 11 5.9 0.028 0.030 2.189 2.191 mp_sum_l 7287 12.8 1.207 2.140 1.207 2.140 grid_integrate_task_list 119 12.3 2.084 2.123 2.084 2.123 calculate_first_density_matrix 1 7.0 0.000 0.000 2.117 2.119 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.039 2.091 ot_diis_step 108 11.5 0.014 0.015 2.010 2.010 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.947 1.949 acc_transpose_blocks 9144 15.5 0.044 0.045 1.831 1.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.843 1.844 fft3d_ps 1201 14.6 0.868 0.886 1.831 1.834 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.596 1.636 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.591 1.592 grid_collocate_task_list 119 9.7 1.525 1.542 1.525 1.542 qs_env_update_s_mstruct 11 6.9 0.386 0.388 1.491 1.509 wfi_extrapolate 11 7.9 0.001 0.001 1.395 1.395 cp_fm_cholesky_invert 11 10.9 1.379 1.382 1.379 1.382 potential_pw2rs 119 12.3 0.014 0.014 1.348 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.300 1.319 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.314 1.314 cp_fm_diag_elpa_base 50 14.0 1.169 1.222 1.312 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.249 hybrid_alltoall_any 4725 16.4 0.093 0.152 1.039 1.230 mp_alltoall_d11v 2130 13.8 1.156 1.187 1.156 1.187 make_images_data 4572 15.5 0.046 0.049 0.960 1.131 acc_transpose_blocks_sync 27432 16.5 1.100 1.127 1.100 1.127 jit_kernel_multiply 6 15.6 0.757 1.117 0.757 1.117 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.995 1.023 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.930 0.944 qs_create_task_list 11 7.9 0.000 0.001 0.925 0.936 generate_qs_task_list 11 8.9 0.367 0.386 0.924 0.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.345000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=724.090909, yerr=16.406509 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 504.479744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.132 83.365 83.368 qs_mol_dyn_low 1 2.0 0.003 0.007 82.491 82.554 qs_forces 11 3.9 0.008 0.030 82.410 82.413 qs_energies 11 4.9 0.002 0.006 79.549 79.575 scf_env_do_scf 11 5.9 0.001 0.002 69.792 69.794 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.309 64.310 dbcsr_multiply_generic 2055 12.4 0.106 0.121 50.895 51.202 qs_scf_new_mos 99 7.5 0.000 0.001 47.275 47.375 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.275 47.375 ot_scf_mini 99 9.5 0.002 0.003 44.931 45.007 multiply_cannon 2055 13.4 0.183 0.189 42.368 43.107 velocity_verlet 10 3.0 0.001 0.002 42.762 42.763 multiply_cannon_loop 2055 14.4 1.784 1.834 41.385 42.149 ot_mini 99 10.5 0.001 0.001 26.489 26.559 qs_ot_get_derivative 99 11.5 0.001 0.001 19.726 19.790 multiply_cannon_multrec 49320 15.4 11.412 12.092 17.378 18.054 rebuild_ks_matrix 110 8.3 0.000 0.000 14.497 14.621 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.016 14.497 14.620 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.726 12.845 mp_waitall_1 220248 16.4 10.610 11.619 10.610 11.619 multiply_cannon_sync_h2d 49320 15.4 9.635 10.189 9.635 10.189 qs_ot_get_p 110 10.4 0.001 0.002 10.006 10.110 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.509 8.052 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.252 7.776 apply_single 110 13.6 0.000 0.000 7.252 7.776 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.417 7.591 init_scf_run 11 5.9 0.000 0.001 7.194 7.195 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.194 7.195 sum_up_and_integrate 110 10.3 0.002 0.003 6.993 7.006 integrate_v_rspace 110 11.3 0.003 0.003 6.968 6.987 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.859 6.881 ot_diis_step 99 11.5 0.006 0.006 6.584 6.584 qs_rho_update_rho_low 110 7.6 0.000 0.002 6.321 6.436 calculate_rho_elec 110 8.6 0.021 0.026 6.320 6.435 dbcsr_mm_accdrv_process 87628 16.1 3.028 3.123 5.834 6.108 cp_dbcsr_syevd 48 12.0 0.044 0.337 6.048 6.049 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.467 5.516 init_scf_loop 11 6.9 0.001 0.005 5.453 5.454 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.138 5.163 cp_fm_diag_elpa_base 48 14.0 5.122 5.147 5.135 5.161 mp_sum_l 6594 12.7 4.032 4.769 4.032 4.769 make_m2s 4110 13.4 0.063 0.088 4.068 4.187 make_images 4110 14.4 0.180 0.201 3.970 4.096 wfi_extrapolate 11 7.9 0.001 0.001 4.049 4.049 calculate_dm_sparse 110 9.5 0.001 0.001 3.811 3.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.628 3.632 density_rs2pw 110 9.6 0.004 0.004 3.376 3.617 grid_integrate_task_list 110 12.3 3.243 3.442 3.243 3.442 multiply_cannon_metrocomm1 49320 15.4 0.065 0.067 2.315 3.364 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.267 3.327 prepare_preconditioner 11 7.9 0.000 0.000 3.275 3.298 make_preconditioner 11 8.9 0.000 0.001 3.275 3.298 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.136 3.189 pw_transfer 1331 11.6 0.055 0.066 3.101 3.177 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.064 3.109 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.013 3.093 calculate_first_density_matrix 1 7.0 0.015 0.114 3.037 3.042 fft_wrap_pw1pw2_140 451 13.1 0.310 0.332 2.554 2.640 acc_transpose_blocks 49320 15.4 0.211 0.218 2.557 2.627 potential_pw2rs 110 12.3 0.005 0.006 2.597 2.613 jit_kernel_multiply 13 15.9 2.531 2.555 2.531 2.555 mp_alltoall_d11v 2046 13.8 2.030 2.531 2.030 2.531 fft3d_ps 1111 14.6 0.798 0.887 2.306 2.362 grid_collocate_task_list 110 9.6 2.161 2.269 2.161 2.269 transfer_rs2pw 451 10.6 0.005 0.006 1.948 2.176 mp_waitany 14300 13.8 1.864 2.143 1.864 2.143 make_images_data 4110 15.4 0.043 0.048 1.801 1.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.902 1.923 mp_sum_d 3891 11.9 1.306 1.808 1.306 1.808 hybrid_alltoall_any 4261 16.3 0.084 0.480 1.540 1.806 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.760 1.802 cp_fm_cholesky_invert 11 10.9 1.781 1.785 1.781 1.785 transfer_pw2rs 451 13.1 0.006 0.007 1.758 1.768 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.368000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=3.599816 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 592.486400E+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 2515781. 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 0.018 0.036 71.948 71.955 qs_mol_dyn_low 1 2.0 0.003 0.003 71.085 71.125 qs_forces 11 3.9 0.003 0.003 71.013 71.014 qs_energies 11 4.9 0.001 0.002 67.659 67.663 scf_env_do_scf 11 5.9 0.000 0.001 58.731 58.734 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 50.714 50.714 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.616 38.872 velocity_verlet 10 3.0 0.001 0.002 36.995 36.996 qs_scf_new_mos 99 7.5 0.001 0.001 34.513 34.668 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.512 34.667 ot_scf_mini 99 9.5 0.003 0.003 32.824 32.955 multiply_cannon 2055 13.4 0.224 0.244 31.721 32.760 multiply_cannon_loop 2055 14.4 1.162 1.188 30.370 31.308 ot_mini 99 10.5 0.001 0.001 19.112 19.260 multiply_cannon_multrec 24660 15.4 6.942 8.058 14.294 15.481 rebuild_ks_matrix 110 8.3 0.000 0.000 13.700 13.824 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.699 13.824 qs_ot_get_derivative 99 11.5 0.001 0.001 13.261 13.395 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.050 12.159 mp_waitall_1 176588 16.5 7.770 10.128 7.770 10.128 init_scf_loop 11 6.9 0.000 0.000 7.976 7.977 multiply_cannon_metrocomm3 24660 15.4 0.074 0.077 5.252 7.683 dbcsr_mm_accdrv_process 52282 16.1 5.721 6.473 7.182 7.471 multiply_cannon_sync_h2d 24660 15.4 6.338 7.389 6.338 7.389 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.548 7.189 apply_single 110 13.6 0.000 0.001 6.548 7.189 qs_ot_get_p 110 10.4 0.009 0.024 7.047 7.185 sum_up_and_integrate 110 10.3 0.002 0.003 6.555 6.566 integrate_v_rspace 110 11.3 0.002 0.003 6.528 6.541 init_scf_run 11 5.9 0.000 0.001 6.458 6.459 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.458 6.459 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.118 6.126 calculate_rho_elec 110 8.6 0.039 0.047 6.117 6.126 prepare_preconditioner 11 7.9 0.000 0.000 5.921 5.939 make_preconditioner 11 8.9 0.000 0.000 5.921 5.939 ot_diis_step 99 11.5 0.010 0.010 5.798 5.798 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.502 5.660 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.847 5.594 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.009 5.030 make_m2s 4110 13.4 0.057 0.059 4.372 4.800 make_images 4110 14.4 0.404 0.463 4.260 4.684 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.522 4.522 pw_transfer 1331 11.6 0.066 0.071 3.754 3.904 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.886 3.897 cp_fm_diag_elpa_base 48 14.0 3.836 3.860 3.883 3.894 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.647 3.798 wfi_extrapolate 11 7.9 0.001 0.001 3.569 3.570 density_rs2pw 110 9.6 0.004 0.004 3.329 3.509 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.416 3.484 grid_integrate_task_list 110 12.3 3.157 3.324 3.157 3.324 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.233 3.235 fft_wrap_pw1pw2_140 451 13.1 0.358 0.378 3.041 3.194 calculate_dm_sparse 110 9.5 0.001 0.001 2.999 3.026 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.878 2.927 fft3d_ps 1111 14.6 1.115 1.344 2.741 2.899 cp_fm_cholesky_invert 11 10.9 2.813 2.822 2.813 2.822 calculate_first_density_matrix 1 7.0 0.002 0.006 2.788 2.790 make_images_data 4110 15.4 0.048 0.053 2.341 2.789 hybrid_alltoall_any 4261 16.3 0.106 0.455 2.085 2.764 mp_sum_l 6594 12.7 1.931 2.652 1.931 2.652 potential_pw2rs 110 12.3 0.008 0.009 2.408 2.426 grid_collocate_task_list 110 9.6 2.171 2.342 2.171 2.342 jit_kernel_multiply 10 16.2 1.096 2.239 1.096 2.239 mp_alltoall_d11v 2046 13.8 1.900 2.184 1.900 2.184 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.016 2.034 acc_transpose_blocks 24660 15.4 0.116 0.119 1.971 2.003 qs_energies_init_hamiltonians 11 5.9 0.007 0.016 1.888 1.889 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.807 1.824 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.770 1.812 transfer_rs2pw 451 10.6 0.006 0.007 1.533 1.774 mp_allgather_i34 2055 14.4 0.661 1.679 0.661 1.679 mp_irecv_dv 57340 16.2 0.646 1.595 0.646 1.595 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.558 1.569 dbcsr_complete_redistribute 325 12.2 0.235 0.294 1.223 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.490 cp_fm_cholesky_decompose 22 10.9 1.462 1.467 1.462 1.467 mp_alltoall_z22v 1111 16.6 1.359 1.443 1.359 1.443 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.955000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=561.181818, yerr=7.208810 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 669.192192E+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 0.008 0.034 62.301 62.302 qs_mol_dyn_low 1 2.0 0.003 0.003 62.014 62.024 qs_forces 11 3.9 0.003 0.003 61.942 61.943 qs_energies 11 4.9 0.005 0.015 58.754 58.757 scf_env_do_scf 11 5.9 0.000 0.001 50.125 50.126 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.225 41.226 velocity_verlet 10 3.0 0.001 0.001 33.380 33.382 dbcsr_multiply_generic 2055 12.4 0.109 0.112 29.667 29.985 qs_scf_new_mos 99 7.5 0.001 0.001 26.383 26.507 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.382 26.506 ot_scf_mini 99 9.5 0.002 0.003 25.113 25.248 multiply_cannon 2055 13.4 0.211 0.221 23.035 24.347 multiply_cannon_loop 2055 14.4 0.817 0.858 21.891 23.155 ot_mini 99 10.5 0.001 0.001 14.248 14.384 rebuild_ks_matrix 110 8.3 0.000 0.000 12.260 12.432 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.260 12.431 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.805 10.962 multiply_cannon_multrec 16440 15.4 3.530 4.330 10.024 10.925 mp_waitall_1 139946 16.5 7.453 10.753 7.453 10.753 qs_ot_get_derivative 99 11.5 0.001 0.001 9.749 9.888 init_scf_loop 11 6.9 0.000 0.000 8.863 8.864 multiply_cannon_metrocomm3 16440 15.4 0.045 0.047 4.742 7.849 prepare_preconditioner 11 7.9 0.000 0.000 7.082 7.098 make_preconditioner 11 8.9 0.000 0.000 7.082 7.098 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.416 6.768 dbcsr_mm_accdrv_process 34862 16.1 5.614 5.896 6.341 6.511 sum_up_and_integrate 110 10.3 0.001 0.002 6.397 6.412 integrate_v_rspace 110 11.3 0.003 0.003 6.371 6.386 init_scf_run 11 5.9 0.000 0.001 6.195 6.195 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.194 6.195 qs_ot_get_p 110 10.4 0.001 0.001 5.950 6.115 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.859 5.871 calculate_rho_elec 110 8.6 0.059 0.059 5.859 5.870 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.042 5.553 apply_single 110 13.6 0.000 0.000 5.042 5.553 make_m2s 4110 13.4 0.049 0.051 4.275 4.602 make_images 4110 14.4 0.398 0.522 4.158 4.484 ot_diis_step 99 11.5 0.011 0.011 4.461 4.462 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.210 4.217 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.324 4.042 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.819 3.819 pw_transfer 1331 11.6 0.066 0.072 3.708 3.718 multiply_cannon_sync_h2d 16440 15.4 3.157 3.653 3.157 3.653 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.601 3.615 grid_integrate_task_list 110 12.3 3.188 3.414 3.188 3.414 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.245 3.255 cp_fm_diag_elpa_base 48 14.0 3.177 3.211 3.244 3.254 density_rs2pw 110 9.6 0.004 0.004 3.083 3.243 fft_wrap_pw1pw2_140 451 13.1 0.461 0.470 3.073 3.092 calculate_first_density_matrix 1 7.0 0.000 0.000 3.052 3.053 wfi_extrapolate 11 7.9 0.001 0.001 3.049 3.049 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.935 2.936 cp_fm_cholesky_invert 11 10.9 2.889 2.897 2.889 2.897 calculate_dm_sparse 110 9.5 0.001 0.001 2.823 2.849 make_images_data 4110 15.4 0.046 0.051 2.374 2.774 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.571 2.648 hybrid_alltoall_any 4261 16.3 0.110 0.382 2.072 2.646 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.573 2.628 fft3d_ps 1111 14.6 1.120 1.132 2.518 2.530 multiply_cannon_metrocomm4 14385 15.4 0.047 0.051 0.916 2.505 mp_sum_l 6594 12.7 1.735 2.439 1.735 2.439 grid_collocate_task_list 110 9.6 2.216 2.412 2.216 2.412 mp_irecv_dv 48980 15.7 0.841 2.365 0.841 2.365 potential_pw2rs 110 12.3 0.011 0.011 2.236 2.246 mp_alltoall_d11v 2046 13.8 1.838 2.126 1.838 2.126 dbcsr_complete_redistribute 325 12.2 0.326 0.387 1.523 1.970 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.968 1.968 cp_fm_upper_to_full 70 14.2 1.394 1.774 1.394 1.774 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.744 1.757 cp_fm_cholesky_decompose 22 10.9 1.679 1.705 1.679 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.649 1.667 acc_transpose_blocks 16440 15.4 0.079 0.081 1.582 1.647 mp_allgather_i34 2055 14.4 0.463 1.513 0.463 1.513 transfer_rs2pw 451 10.6 0.005 0.006 1.320 1.490 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.373 1.477 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.004 1.443 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.357 1.364 mp_waitany 17072 13.8 1.142 1.363 1.142 1.363 transfer_pw2rs 451 13.1 0.005 0.005 1.241 1.250 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.302000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.545455, yerr=8.116914 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 736.198656E+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 0.008 0.026 66.870 66.870 qs_mol_dyn_low 1 2.0 0.003 0.003 66.638 66.647 qs_forces 11 3.9 0.003 0.003 66.559 66.561 qs_energies 11 4.9 0.002 0.002 63.155 63.158 scf_env_do_scf 11 5.9 0.000 0.001 54.510 54.512 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.789 42.790 velocity_verlet 10 3.0 0.001 0.002 37.405 37.408 dbcsr_multiply_generic 2055 12.4 0.116 0.122 31.386 31.673 qs_scf_new_mos 99 7.5 0.001 0.001 28.108 28.220 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.107 28.219 ot_scf_mini 99 9.5 0.002 0.003 26.422 26.533 multiply_cannon 2055 13.4 0.236 0.250 24.014 25.058 multiply_cannon_loop 2055 14.4 1.435 1.478 22.582 23.296 ot_mini 99 10.5 0.001 0.001 15.105 15.237 multiply_cannon_multrec 24660 15.4 4.093 6.779 13.253 14.521 rebuild_ks_matrix 110 8.3 0.000 0.000 12.040 12.155 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.040 12.154 init_scf_loop 11 6.9 0.000 0.000 11.680 11.681 qs_ot_get_derivative 99 11.5 0.001 0.001 10.941 11.056 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.635 10.737 prepare_preconditioner 11 7.9 0.000 0.000 9.912 9.927 make_preconditioner 11 8.9 0.000 0.000 9.911 9.927 dbcsr_mm_accdrv_process 52304 16.0 7.993 9.264 9.000 9.873 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.120 9.581 qs_ot_get_p 110 10.4 0.001 0.001 6.175 6.348 sum_up_and_integrate 110 10.3 0.002 0.003 6.267 6.281 integrate_v_rspace 110 11.3 0.003 0.003 6.241 6.255 mp_waitall_1 121746 16.5 4.445 6.155 4.445 6.155 make_m2s 4110 13.4 0.059 0.063 5.749 6.106 make_images 4110 14.4 0.674 0.897 5.607 5.960 init_scf_run 11 5.9 0.000 0.001 5.922 5.922 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.921 5.922 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.861 5.869 calculate_rho_elec 110 8.6 0.078 0.081 5.860 5.869 cp_fm_upper_to_full 70 14.2 3.384 4.891 3.384 4.891 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.260 4.275 ot_diis_step 99 11.5 0.011 0.011 4.127 4.127 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.041 4.119 apply_single 110 13.6 0.000 0.000 4.041 4.119 dbcsr_complete_redistribute 325 12.2 0.458 0.544 2.728 3.876 pw_transfer 1331 11.6 0.065 0.074 3.790 3.825 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.683 3.721 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.509 3.564 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.490 3.490 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.393 3.459 grid_integrate_task_list 110 12.3 3.306 3.413 3.306 3.413 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.214 3.346 multiply_cannon_metrocomm3 24660 15.4 0.037 0.038 1.475 3.312 calculate_dm_sparse 110 9.5 0.001 0.001 3.224 3.258 fft_wrap_pw1pw2_140 451 13.1 0.531 0.543 3.164 3.203 hybrid_alltoall_any 4261 16.3 0.125 0.468 2.355 3.201 make_images_data 4110 15.4 0.048 0.052 2.795 3.157 density_rs2pw 110 9.6 0.004 0.004 2.990 3.099 wfi_extrapolate 11 7.9 0.001 0.001 3.020 3.020 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.947 2.958 cp_fm_diag_elpa_base 48 14.0 2.799 2.858 2.945 2.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.902 2.903 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.782 2.901 mp_alltoall_i22 605 13.7 1.647 2.832 1.647 2.832 calculate_first_density_matrix 1 7.0 0.000 0.000 2.806 2.808 cp_fm_cholesky_invert 11 10.9 2.646 2.654 2.646 2.654 acc_transpose_blocks 24660 15.4 0.113 0.117 2.457 2.563 fft3d_ps 1111 14.6 1.150 1.190 2.509 2.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.461 2.507 multiply_cannon_sync_h2d 24660 15.4 2.388 2.507 2.388 2.507 grid_collocate_task_list 110 9.6 2.266 2.480 2.266 2.480 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.233 2.235 potential_pw2rs 110 12.3 0.013 0.013 2.060 2.065 mp_alltoall_d11v 2046 13.8 1.775 1.979 1.775 1.979 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.800 1.840 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.779 1.795 cp_fm_cholesky_decompose 22 10.9 1.675 1.731 1.675 1.731 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.616 1.716 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.645 1.659 acc_transpose_blocks_sync 73980 16.4 1.444 1.555 1.444 1.555 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.847 1.515 mp_sum_l 6594 12.7 0.934 1.444 0.934 1.444 mp_irecv_dv 62702 16.1 0.745 1.437 0.745 1.437 transfer_rs2pw 451 10.6 0.005 0.005 1.169 1.341 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.870000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.909091, yerr=9.109980 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 848.109568E+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 0.023 0.055 56.267 56.281 qs_mol_dyn_low 1 2.0 0.003 0.003 55.971 55.987 qs_forces 11 3.9 0.003 0.003 55.607 55.611 qs_energies 11 4.9 0.002 0.002 51.955 51.965 scf_env_do_scf 11 5.9 0.000 0.001 43.300 43.301 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.428 35.429 velocity_verlet 10 3.0 0.259 0.303 31.744 31.747 dbcsr_multiply_generic 2055 12.4 0.106 0.108 23.103 23.211 qs_scf_new_mos 99 7.5 0.001 0.001 20.704 20.744 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.703 20.743 ot_scf_mini 99 9.5 0.002 0.002 19.446 19.481 multiply_cannon 2055 13.4 0.235 0.249 17.432 18.855 multiply_cannon_loop 2055 14.4 0.605 0.623 16.151 16.467 rebuild_ks_matrix 110 8.3 0.000 0.000 11.752 11.773 qs_ks_build_kohn_sham_matrix 110 9.3 0.031 0.035 11.752 11.772 ot_mini 99 10.5 0.001 0.001 10.575 10.598 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.455 10.472 multiply_cannon_multrec 8220 15.4 3.158 4.401 7.544 8.495 init_scf_loop 11 6.9 0.000 0.000 7.823 7.824 mp_waitall_1 103326 16.6 5.733 7.337 5.733 7.337 qs_ot_get_derivative 99 11.5 0.001 0.001 6.893 6.926 sum_up_and_integrate 110 10.3 0.002 0.002 6.273 6.285 integrate_v_rspace 110 11.3 0.003 0.003 6.247 6.259 prepare_preconditioner 11 7.9 0.000 0.000 6.145 6.161 make_preconditioner 11 8.9 0.000 0.000 6.145 6.161 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.082 6.097 calculate_rho_elec 110 8.6 0.113 0.113 6.081 6.096 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.725 5.822 dbcsr_mm_accdrv_process 17442 15.9 3.079 4.140 4.248 5.203 init_scf_run 11 5.9 0.000 0.001 5.134 5.135 scf_env_initial_rho_setup 11 6.9 0.002 0.002 5.134 5.134 qs_ot_get_p 110 10.4 0.002 0.002 4.902 4.925 make_m2s 4110 13.4 0.037 0.039 4.296 4.576 make_images 4110 14.4 0.644 0.704 4.165 4.446 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.857 4.168 pw_transfer 1331 11.6 0.065 0.070 4.143 4.152 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.036 4.047 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.726 3.767 apply_single 110 13.6 0.000 0.000 3.726 3.767 ot_diis_step 99 11.5 0.012 0.012 3.659 3.659 grid_integrate_task_list 110 12.3 3.376 3.510 3.376 3.510 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.500 3.505 fft_wrap_pw1pw2_140 451 13.1 0.723 0.735 3.485 3.498 density_rs2pw 110 9.6 0.004 0.004 3.096 3.236 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.154 3.154 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.055 3.059 cp_fm_cholesky_invert 11 10.9 2.961 2.966 2.961 2.966 hybrid_alltoall_any 4261 16.3 0.201 0.856 2.270 2.799 wfi_extrapolate 11 7.9 0.001 0.001 2.750 2.751 make_images_data 4110 15.4 0.040 0.047 2.307 2.713 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.616 2.623 cp_fm_diag_elpa_base 48 14.0 2.558 2.580 2.614 2.621 fft3d_ps 1111 14.6 1.295 1.299 2.578 2.587 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.539 2.540 calculate_dm_sparse 110 9.5 0.001 0.001 2.480 2.519 multiply_cannon_sync_h2d 8220 15.4 2.383 2.456 2.383 2.456 grid_collocate_task_list 110 9.6 2.366 2.422 2.366 2.422 calculate_first_density_matrix 1 7.0 0.000 0.001 2.287 2.288 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.135 2.148 potential_pw2rs 110 12.3 0.015 0.016 2.066 2.073 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.042 2.072 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.787 1.993 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.882 1.989 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.959 1.972 mp_alltoall_d11v 2046 13.8 1.666 1.865 1.666 1.865 cp_fm_cholesky_decompose 22 10.9 1.733 1.758 1.733 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.628 mp_allgather_i34 2055 14.4 0.459 1.573 0.459 1.573 dbcsr_complete_redistribute 325 12.2 0.562 0.601 1.477 1.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.427 1.441 acc_transpose_blocks 8220 15.4 0.041 0.042 1.364 1.390 qs_create_task_list 11 7.9 0.001 0.001 1.213 1.313 generate_qs_task_list 11 8.9 0.376 0.444 1.212 1.313 transfer_rs2pw 451 10.6 0.005 0.005 1.143 1.307 mp_waitany 9240 13.8 1.048 1.219 1.048 1.219 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.157 1.183 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.717 1.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.281000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=803.181818, yerr=9.998347 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 1.387454E+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 0.024 0.045 91.028 91.030 qs_mol_dyn_low 1 2.0 0.003 0.003 90.702 90.714 qs_forces 11 3.9 0.003 0.003 90.370 90.371 qs_energies 11 4.9 0.002 0.002 86.181 86.182 scf_env_do_scf 11 5.9 0.000 0.001 75.111 75.111 velocity_verlet 10 3.0 0.002 0.003 57.579 57.618 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 46.232 46.234 dbcsr_multiply_generic 2055 12.4 0.122 0.126 30.652 30.772 init_scf_loop 11 6.9 0.000 0.000 28.804 28.806 qs_scf_new_mos 99 7.5 0.001 0.001 27.971 28.051 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.970 28.050 prepare_preconditioner 11 7.9 0.000 0.000 26.715 26.732 make_preconditioner 11 8.9 0.000 0.000 26.715 26.732 ot_scf_mini 99 9.5 0.002 0.002 26.148 26.222 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.915 26.175 multiply_cannon 2055 13.4 0.332 0.358 23.197 23.964 multiply_cannon_loop 2055 14.4 0.823 0.839 21.310 21.631 cp_fm_upper_to_full 70 14.2 12.685 18.194 12.685 18.194 ot_mini 99 10.5 0.001 0.001 14.666 14.736 rebuild_ks_matrix 110 8.3 0.000 0.001 14.230 14.318 qs_ks_build_kohn_sham_matrix 110 9.3 0.019 0.024 14.230 14.317 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.895 12.973 dbcsr_complete_redistribute 325 12.2 0.999 1.033 7.414 10.632 multiply_cannon_multrec 8220 15.4 4.065 4.257 9.931 10.246 qs_ot_get_derivative 99 11.5 0.001 0.001 9.814 9.889 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.371 9.586 mp_waitall_1 84994 16.7 8.206 9.336 8.206 9.336 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.784 8.971 mp_alltoall_i22 605 13.7 5.421 8.611 5.421 8.611 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.646 7.685 calculate_rho_elec 110 8.6 0.222 0.223 7.645 7.684 sum_up_and_integrate 110 10.3 0.002 0.002 7.076 7.090 integrate_v_rspace 110 11.3 0.003 0.004 7.048 7.062 init_scf_run 11 5.9 0.000 0.001 6.371 6.371 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.371 6.371 qs_ot_get_p 110 10.4 0.001 0.002 6.090 6.167 dbcsr_mm_accdrv_process 11614 15.7 3.975 4.809 5.717 6.134 make_m2s 4110 13.4 0.042 0.043 5.569 6.122 pw_transfer 1331 11.6 0.075 0.075 5.986 5.996 make_images 4110 14.4 0.878 0.932 5.379 5.932 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.869 5.879 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 5.340 5.682 cp_fm_cholesky_invert 11 10.9 5.608 5.613 5.608 5.613 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.963 5.460 apply_single 110 13.6 0.000 0.000 4.962 5.460 fft_wrap_pw1pw2_140 451 13.1 1.367 1.371 5.118 5.126 ot_diis_step 99 11.5 0.015 0.016 4.830 4.830 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.376 4.383 density_rs2pw 110 9.6 0.004 0.004 4.078 4.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.070 4.070 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.849 3.849 grid_integrate_task_list 110 12.3 3.697 3.771 3.697 3.771 hybrid_alltoall_any 4261 16.3 0.266 0.566 2.954 3.694 make_images_data 4110 15.4 0.045 0.048 2.977 3.678 fft3d_ps 1111 14.6 1.885 1.899 3.574 3.587 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.116 3.585 wfi_extrapolate 11 7.9 0.001 0.001 3.480 3.480 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.360 3.363 calculate_dm_sparse 110 9.5 0.001 0.001 3.219 3.244 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.227 3.228 cp_fm_diag_elpa_base 48 14.0 2.691 2.888 3.225 3.226 multiply_cannon_sync_h2d 8220 15.4 3.126 3.143 3.126 3.143 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.821 2.852 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.741 2.781 calculate_first_density_matrix 1 7.0 0.000 0.000 2.769 2.770 grid_collocate_task_list 110 9.6 2.684 2.709 2.684 2.709 potential_pw2rs 110 12.3 0.021 0.022 2.555 2.564 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.213 2.267 mp_alltoall_d11v 2046 13.8 2.158 2.226 2.158 2.226 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.097 2.167 cp_fm_cholesky_decompose 22 10.9 2.102 2.118 2.102 2.118 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.941 1.949 qs_create_task_list 11 7.9 0.005 0.005 1.888 1.934 generate_qs_task_list 11 8.9 0.730 0.784 1.883 1.928 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.812 1.861 jit_kernel_multiply 10 15.1 1.546 1.843 1.546 1.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=91.030000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1246.090909, yerr=44.029855 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 635.490304E+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 2893409. 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 0.012 0.033 207.747 207.748 qs_mol_dyn_low 1 2.0 0.003 0.003 207.291 207.307 qs_forces 11 3.9 0.005 0.005 207.205 207.207 qs_energies 11 4.9 0.002 0.003 201.596 201.613 scf_env_do_scf 11 5.9 0.001 0.001 184.836 184.840 scf_env_do_scf_inner_loop 117 6.6 0.003 0.013 164.326 164.328 dbcsr_multiply_generic 2507 12.6 0.179 0.183 126.231 126.935 qs_scf_new_mos 117 7.6 0.001 0.001 124.898 125.199 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.897 125.198 velocity_verlet 10 3.0 0.001 0.002 123.598 123.599 ot_scf_mini 117 9.6 0.003 0.003 118.233 118.537 multiply_cannon 2507 13.6 0.236 0.246 102.258 104.297 multiply_cannon_loop 2507 14.6 2.376 2.454 100.097 102.065 ot_mini 117 10.6 0.001 0.001 66.832 67.168 multiply_cannon_multrec 60168 15.6 32.050 33.870 41.787 43.908 qs_ot_get_derivative 117 11.6 0.001 0.001 41.828 42.141 rebuild_ks_matrix 128 8.3 0.001 0.001 33.627 33.883 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.627 33.882 mp_waitall_1 267128 16.5 29.168 32.562 29.168 32.562 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.189 30.443 qs_ot_get_p 128 10.4 0.011 0.041 29.362 29.694 multiply_cannon_sync_h2d 60168 15.6 26.647 28.516 26.647 28.516 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.493 25.176 apply_single 128 13.6 0.001 0.001 24.492 25.176 ot_diis_step 117 11.6 0.008 0.008 24.721 24.722 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.550 22.616 init_scf_loop 11 6.9 0.000 0.001 20.436 20.437 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.769 20.030 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.917 19.918 multiply_cannon_metrocomm3 60168 15.6 0.119 0.126 16.138 18.249 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.944 16.986 cp_fm_diag_elpa_base 83 14.4 16.871 16.917 16.939 16.983 prepare_preconditioner 11 7.9 0.000 0.000 15.681 15.727 make_preconditioner 11 8.9 0.000 0.000 15.681 15.727 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.933 15.093 make_m2s 5014 13.6 0.104 0.110 13.849 14.324 make_images 5014 14.6 0.399 0.416 13.664 14.146 sum_up_and_integrate 128 10.3 0.002 0.004 13.998 14.012 integrate_v_rspace 128 11.3 0.003 0.004 13.940 13.958 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.120 13.270 calculate_rho_elec 128 8.7 0.045 0.064 13.120 13.270 init_scf_run 11 5.9 0.000 0.001 12.550 12.551 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.550 12.550 mp_sum_l 7950 12.9 8.858 10.568 8.858 10.568 dbcsr_mm_accdrv_process 124484 16.2 4.585 4.787 9.299 10.137 wfi_extrapolate 11 7.9 0.001 0.001 9.085 9.085 cp_fm_cholesky_invert 11 10.9 8.884 8.892 8.884 8.892 multiply_cannon_metrocomm1 60168 15.6 0.096 0.104 6.570 8.638 calculate_dm_sparse 128 9.5 0.001 0.001 8.530 8.603 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.135 8.275 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.060 8.143 pw_transfer 1547 11.6 0.073 0.087 7.636 7.816 make_images_data 5014 15.6 0.066 0.073 6.741 7.647 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.435 7.609 grid_integrate_task_list 128 12.3 7.055 7.501 7.055 7.501 density_rs2pw 128 9.7 0.006 0.006 6.791 7.444 hybrid_alltoall_any 5200 16.5 0.294 2.260 5.893 7.134 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.810 6.822 fft_wrap_pw1pw2_140 523 13.2 0.858 0.904 6.493 6.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.787 5.893 fft3d_ps 1291 14.7 2.195 2.844 5.441 5.772 mp_alltoall_d11v 2415 14.1 4.284 5.565 4.284 5.565 grid_collocate_task_list 128 9.7 4.845 5.231 4.845 5.231 cp_fm_cholesky_decompose 22 10.9 4.628 4.643 4.628 4.643 potential_pw2rs 128 12.3 0.009 0.010 4.500 4.527 mp_sum_d 4481 12.1 3.502 4.264 3.502 4.264 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=207.748000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=599.090909, yerr=7.354090 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/16/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 2.183246E+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 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 845.221888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6608528. 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 0.011 0.033 190.928 190.930 qs_mol_dyn_low 1 2.0 0.003 0.003 190.443 190.465 qs_forces 11 3.9 0.004 0.005 190.355 190.357 qs_energies 11 4.9 0.002 0.003 183.588 183.600 scf_env_do_scf 11 5.9 0.001 0.001 167.151 167.160 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 133.911 133.914 velocity_verlet 10 3.0 0.001 0.002 119.581 119.583 dbcsr_multiply_generic 2507 12.6 0.187 0.194 97.706 98.933 qs_scf_new_mos 117 7.6 0.001 0.001 95.363 95.983 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.362 95.983 ot_scf_mini 117 9.6 0.004 0.004 90.555 91.246 multiply_cannon 2507 13.6 0.501 0.558 76.859 80.937 multiply_cannon_loop 2507 14.6 1.568 1.650 73.863 76.762 ot_mini 117 10.6 0.001 0.001 50.259 50.872 mp_waitall_1 214728 16.6 24.399 38.083 24.399 38.083 multiply_cannon_multrec 30084 15.6 20.863 24.869 31.456 35.990 init_scf_loop 11 6.9 0.000 0.000 33.148 33.149 rebuild_ks_matrix 128 8.3 0.001 0.001 32.226 32.961 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.027 32.225 32.960 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.979 29.666 multiply_cannon_metrocomm3 30084 15.6 0.100 0.107 15.835 29.036 qs_ot_get_derivative 117 11.6 0.001 0.002 28.297 28.986 prepare_preconditioner 11 7.9 0.000 0.000 28.759 28.842 make_preconditioner 11 8.9 0.000 0.000 28.759 28.842 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.448 27.998 qs_ot_get_p 128 10.4 0.012 0.022 22.469 23.276 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.106 23.181 apply_single 128 13.6 0.001 0.001 22.106 23.181 ot_diis_step 117 11.6 0.014 0.015 21.784 21.787 multiply_cannon_sync_h2d 30084 15.6 17.899 20.009 17.899 20.009 qs_ot_p2m_diag 83 11.4 0.189 0.217 17.414 17.458 cp_fm_cholesky_invert 11 10.9 16.818 16.831 16.818 16.831 make_m2s 5014 13.6 0.086 0.094 14.728 16.029 make_images 5014 14.6 1.159 1.350 14.513 15.812 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.723 15.724 sum_up_and_integrate 128 10.3 0.002 0.004 14.039 14.067 integrate_v_rspace 128 11.3 0.003 0.004 13.978 14.010 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.435 13.477 calculate_rho_elec 128 8.7 0.087 0.104 13.434 13.476 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.501 12.538 cp_fm_diag_elpa_base 83 14.4 12.244 12.352 12.495 12.527 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.263 11.766 init_scf_run 11 5.9 0.000 0.001 11.625 11.627 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.625 11.626 multiply_cannon_metrocomm4 27577 15.6 0.099 0.112 3.885 10.830 dbcsr_mm_accdrv_process 62242 16.2 5.381 6.158 10.050 10.623 mp_irecv_dv 69486 16.3 3.684 10.429 3.684 10.429 make_images_data 5014 15.6 0.066 0.078 8.512 10.045 hybrid_alltoall_any 5200 16.5 0.351 1.502 7.213 9.586 pw_transfer 1547 11.6 0.084 0.094 8.643 8.700 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.419 8.481 wfi_extrapolate 11 7.9 0.001 0.001 8.368 8.368 grid_integrate_task_list 128 12.3 7.181 7.605 7.181 7.605 density_rs2pw 128 9.7 0.006 0.006 7.109 7.560 fft_wrap_pw1pw2_140 523 13.2 0.936 0.963 7.433 7.512 cp_fm_cholesky_decompose 22 10.9 7.026 7.093 7.026 7.093 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.267 7.046 calculate_dm_sparse 128 9.5 0.001 0.001 6.468 6.605 mp_sum_l 7950 12.9 4.321 6.452 4.321 6.452 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.104 6.113 fft3d_ps 1291 14.7 2.831 3.026 5.972 6.020 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.390 5.463 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.198 5.423 grid_collocate_task_list 128 9.7 5.003 5.396 5.003 5.396 mp_allgather_i34 2507 14.6 1.526 4.651 1.526 4.651 mp_alltoall_d11v 2415 14.1 4.230 4.651 4.230 4.651 potential_pw2rs 128 12.3 0.015 0.017 4.526 4.541 dbcsr_complete_redistribute 395 12.7 0.761 0.841 3.118 3.963 mp_sum_d 4472 12.1 2.626 3.959 2.626 3.959 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.930000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=804.727273, yerr=6.031597 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 970.739712E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937890820720 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57629. MP_Allreduce 11184 996. MP_Sync 88 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.046 177.923 177.928 qs_mol_dyn_low 1 2.0 0.003 0.003 177.478 177.522 qs_forces 11 3.9 0.004 0.004 177.383 177.385 qs_energies 11 4.9 0.002 0.003 170.847 170.856 scf_env_do_scf 11 5.9 0.001 0.001 154.601 154.603 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 119.513 119.514 velocity_verlet 10 3.0 0.001 0.002 113.840 113.843 dbcsr_multiply_generic 2529 12.6 0.186 0.190 82.803 83.875 qs_scf_new_mos 118 7.6 0.001 0.001 81.986 82.291 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.985 82.291 ot_scf_mini 118 9.6 0.003 0.004 77.698 78.049 multiply_cannon 2529 13.6 0.509 0.526 62.786 67.385 multiply_cannon_loop 2529 14.6 1.142 1.223 59.664 62.476 ot_mini 118 10.6 0.001 0.001 42.853 43.202 init_scf_loop 11 6.9 0.000 0.000 34.986 34.987 mp_waitall_1 172006 16.6 24.645 34.470 24.645 34.470 prepare_preconditioner 11 7.9 0.000 0.000 30.981 31.026 make_preconditioner 11 8.9 0.000 0.000 30.981 31.026 rebuild_ks_matrix 129 8.3 0.001 0.001 30.453 30.885 qs_ks_build_kohn_sham_matrix 129 9.3 0.021 0.039 30.452 30.884 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.661 30.047 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.470 27.865 multiply_cannon_multrec 20232 15.6 13.288 16.578 22.966 26.204 multiply_cannon_metrocomm3 20232 15.6 0.062 0.066 15.000 25.032 qs_ot_get_derivative 118 11.6 0.001 0.002 23.199 23.546 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.677 20.735 apply_single 129 13.6 0.001 0.001 19.677 20.734 qs_ot_get_p 129 10.4 0.001 0.002 20.116 20.545 ot_diis_step 118 11.6 0.018 0.018 19.540 19.540 make_m2s 5058 13.6 0.078 0.084 14.956 16.279 make_images 5058 14.6 1.152 1.237 14.719 16.039 qs_ot_p2m_diag 84 11.4 0.268 0.275 15.723 15.730 multiply_cannon_sync_h2d 20232 15.6 13.767 15.341 13.767 15.341 cp_dbcsr_syevd 84 12.4 0.005 0.006 14.624 14.625 cp_fm_cholesky_invert 11 10.9 14.484 14.493 14.484 14.493 sum_up_and_integrate 129 10.3 0.002 0.004 14.051 14.077 integrate_v_rspace 129 11.3 0.003 0.004 13.990 14.017 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.785 13.822 calculate_rho_elec 129 8.7 0.130 0.144 13.785 13.821 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.326 11.349 cp_fm_diag_elpa_base 84 14.4 10.902 11.067 11.322 11.345 init_scf_run 11 5.9 0.000 0.001 11.228 11.229 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.228 11.228 make_images_data 5058 15.6 0.064 0.074 8.968 10.655 hybrid_alltoall_any 5245 16.5 0.453 2.044 7.681 10.124 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.246 9.506 dbcsr_mm_accdrv_process 41846 16.2 5.931 6.257 9.128 9.283 multiply_cannon_metrocomm4 17703 15.6 0.065 0.073 3.445 9.174 pw_transfer 1559 11.6 0.085 0.103 8.954 9.071 mp_irecv_dv 50659 16.2 3.315 8.918 3.315 8.918 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.729 8.851 grid_integrate_task_list 129 12.3 7.379 7.908 7.379 7.908 fft_wrap_pw1pw2_140 527 13.2 1.064 1.097 7.696 7.833 density_rs2pw 129 9.7 0.006 0.006 7.101 7.537 wfi_extrapolate 11 7.9 0.001 0.001 7.477 7.477 cp_fm_cholesky_decompose 22 10.9 7.359 7.392 7.359 7.392 cp_fm_upper_to_full 106 14.8 5.758 7.257 5.758 7.257 dbcsr_complete_redistribute 397 12.7 1.160 1.198 4.848 6.684 calculate_dm_sparse 129 9.5 0.001 0.001 6.460 6.611 fft3d_ps 1301 14.7 2.839 3.087 6.028 6.109 grid_collocate_task_list 129 9.7 5.204 5.754 5.204 5.754 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.554 5.558 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.667 5.497 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.655 5.286 mp_alltoall_d11v 2429 14.1 4.269 5.187 4.269 5.187 mp_allgather_i34 2529 14.6 1.622 4.919 1.622 4.919 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.639 4.795 mp_sum_l 8016 12.9 3.210 4.778 3.210 4.778 potential_pw2rs 129 12.3 0.021 0.023 4.428 4.441 mp_alltoall_i22 720 14.1 2.240 4.172 2.240 4.172 transfer_fm_to_dbcsr 11 9.9 0.018 0.024 2.301 4.018 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.961 3.999 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.716 3.716 calculate_first_density_matrix 1 7.0 0.052 0.342 3.564 3.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.928000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=898.181818, yerr=19.135014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11613065416704 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 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.175286E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922098320 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509591. 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 0.035 0.078 199.358 199.375 qs_mol_dyn_low 1 2.0 0.003 0.003 198.916 198.934 qs_forces 11 3.9 0.004 0.004 198.800 198.805 qs_energies 11 4.9 0.002 0.002 191.516 191.526 scf_env_do_scf 11 5.9 0.001 0.001 173.615 173.626 velocity_verlet 10 3.0 0.002 0.002 131.191 131.194 scf_env_do_scf_inner_loop 118 6.6 0.013 0.267 124.149 124.150 qs_scf_new_mos 118 7.6 0.001 0.001 86.241 86.515 qs_scf_loop_do_ot 118 8.6 0.001 0.001 86.240 86.515 dbcsr_multiply_generic 2534 12.6 0.195 0.201 82.901 83.625 ot_scf_mini 118 9.6 0.003 0.004 81.605 81.911 multiply_cannon 2534 13.6 0.549 0.589 56.951 59.590 multiply_cannon_loop 2534 14.6 1.863 1.952 52.978 54.843 init_scf_loop 11 6.9 0.000 0.000 49.336 49.337 ot_mini 118 10.6 0.001 0.001 44.161 44.457 prepare_preconditioner 11 7.9 0.000 0.000 44.234 44.279 make_preconditioner 11 8.9 0.000 0.000 44.234 44.279 make_full_inverse_cholesky 11 9.9 0.010 0.020 37.746 42.734 multiply_cannon_multrec 30408 15.6 13.693 18.297 26.713 31.314 rebuild_ks_matrix 129 8.3 0.001 0.001 30.283 30.622 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.024 30.283 30.621 mp_waitall_1 149452 16.7 18.191 28.246 18.191 28.246 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.286 27.583 qs_ot_get_derivative 118 11.6 0.001 0.002 24.299 24.599 qs_ot_get_p 129 10.4 0.053 0.099 24.174 24.504 make_m2s 5068 13.6 0.094 0.097 21.105 22.198 make_images 5068 14.6 1.937 2.252 20.791 21.887 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.299 19.842 apply_single 129 13.6 0.001 0.001 19.298 19.842 ot_diis_step 118 11.6 0.018 0.018 19.739 19.740 qs_ot_p2m_diag 84 11.4 0.347 0.395 18.859 18.911 cp_dbcsr_syevd 84 12.4 0.005 0.006 17.387 17.389 cp_fm_cholesky_invert 11 10.9 17.358 17.367 17.358 17.367 cp_fm_upper_to_full 106 14.8 11.462 16.875 11.462 16.875 multiply_cannon_metrocomm3 30408 15.6 0.049 0.052 6.602 15.134 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.621 14.696 calculate_rho_elec 129 8.7 0.171 0.187 14.620 14.695 sum_up_and_integrate 129 10.3 0.002 0.003 14.586 14.614 integrate_v_rspace 129 11.3 0.003 0.004 14.526 14.556 cp_fm_diag_elpa 84 13.4 0.000 0.000 14.026 14.039 cp_fm_diag_elpa_base 84 14.4 13.025 13.354 14.017 14.030 make_images_data 5068 15.6 0.065 0.072 11.437 13.143 dbcsr_complete_redistribute 397 12.7 1.504 1.666 9.302 13.069 dbcsr_mm_accdrv_process 62840 16.2 8.526 9.330 12.585 13.062 hybrid_alltoall_any 5255 16.5 0.537 2.249 10.069 12.344 init_scf_run 11 5.9 0.000 0.001 11.922 11.923 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.922 11.923 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.829 11.617 multiply_cannon_sync_h2d 30408 15.6 10.421 11.386 10.421 11.386 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.219 10.428 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 6.466 10.171 pw_transfer 1559 11.6 0.087 0.103 9.718 9.791 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.491 9.568 mp_alltoall_i22 720 14.1 5.649 9.412 5.649 9.412 fft_wrap_pw1pw2_140 527 13.2 1.263 1.316 8.331 8.419 cp_fm_cholesky_decompose 22 10.9 8.241 8.327 8.241 8.327 grid_integrate_task_list 129 12.3 7.587 7.930 7.587 7.930 wfi_extrapolate 11 7.9 0.001 0.001 7.820 7.820 density_rs2pw 129 9.7 0.006 0.006 7.322 7.693 multiply_cannon_metrocomm4 25340 15.6 0.080 0.090 2.828 7.537 mp_irecv_dv 76866 16.2 2.675 7.258 2.675 7.258 calculate_dm_sparse 129 9.5 0.001 0.001 6.698 6.785 fft3d_ps 1301 14.7 3.022 3.078 6.512 6.569 mp_alltoall_d11v 2429 14.1 5.456 6.259 5.456 6.259 grid_collocate_task_list 129 9.7 5.328 5.716 5.328 5.716 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.648 5.679 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.701 4.790 potential_pw2rs 129 12.3 0.024 0.024 4.714 4.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.480 4.569 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.433 4.433 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.370 4.423 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=199.375000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1096.727273, yerr=24.643240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/19/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 5.865088E+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 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 1.577751E+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 5.532237E+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 721350092304 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102572. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58346. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838210. 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 0.035 178.620 178.621 qs_mol_dyn_low 1 2.0 0.003 0.003 178.206 178.221 qs_forces 11 3.9 0.004 0.005 178.104 178.107 qs_energies 11 4.9 0.015 0.019 170.590 170.599 scf_env_do_scf 11 5.9 0.001 0.001 149.800 149.813 velocity_verlet 10 3.0 0.004 0.005 114.512 114.516 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 113.207 113.209 qs_scf_new_mos 117 7.6 0.001 0.001 75.176 75.266 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.175 75.265 dbcsr_multiply_generic 2507 12.6 0.184 0.191 73.045 73.446 ot_scf_mini 117 9.6 0.003 0.004 70.731 70.807 multiply_cannon 2507 13.6 0.557 0.591 53.880 57.471 multiply_cannon_loop 2507 14.6 0.805 0.842 50.911 51.642 ot_mini 117 10.6 0.001 0.001 38.001 38.085 init_scf_loop 11 6.9 0.000 0.000 36.434 36.437 prepare_preconditioner 11 7.9 0.000 0.000 32.484 32.508 make_preconditioner 11 8.9 0.000 0.000 32.484 32.508 mp_waitall_1 125778 16.7 25.156 31.380 25.156 31.380 make_full_inverse_cholesky 11 9.9 0.016 0.027 30.265 30.593 rebuild_ks_matrix 128 8.3 0.001 0.001 29.662 29.742 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.662 29.742 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.022 27.091 multiply_cannon_multrec 10028 15.6 10.389 14.508 18.196 21.309 qs_ot_get_derivative 117 11.6 0.002 0.002 21.030 21.101 qs_ot_get_p 128 10.4 0.007 0.008 20.052 20.151 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 12.360 19.261 cp_fm_cholesky_invert 11 10.9 18.494 18.502 18.494 18.502 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.955 17.092 apply_single 128 13.6 0.001 0.001 16.954 17.092 ot_diis_step 117 11.6 0.020 0.020 16.902 16.903 qs_ot_p2m_diag 83 11.4 0.496 0.501 16.036 16.050 make_m2s 5014 13.6 0.065 0.069 15.175 15.998 make_images 5014 14.6 2.147 2.594 14.867 15.691 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.869 14.901 calculate_rho_elec 128 8.7 0.253 0.264 14.869 14.901 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.824 14.825 sum_up_and_integrate 128 10.3 0.002 0.003 14.526 14.576 integrate_v_rspace 128 11.3 0.004 0.004 14.466 14.518 init_scf_run 11 5.9 0.000 0.001 13.801 13.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.801 13.802 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.795 11.806 cp_fm_diag_elpa_base 83 14.4 11.541 11.636 11.790 11.802 multiply_cannon_sync_h2d 10028 15.6 10.886 11.202 10.886 11.202 pw_transfer 1547 11.6 0.085 0.095 10.263 10.302 make_images_data 5014 15.6 0.054 0.064 8.753 10.148 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 10.040 10.086 hybrid_alltoall_any 5200 16.5 0.842 3.774 8.528 10.001 fft_wrap_pw1pw2_140 523 13.2 1.671 1.698 8.804 8.847 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.322 8.388 cp_fm_cholesky_decompose 22 10.9 8.152 8.269 8.152 8.269 dbcsr_mm_accdrv_process 20762 16.1 3.280 4.547 7.453 8.128 grid_integrate_task_list 128 12.3 7.772 8.078 7.772 8.078 density_rs2pw 128 9.7 0.005 0.005 7.363 7.744 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.661 7.529 wfi_extrapolate 11 7.9 0.001 0.001 7.484 7.484 calculate_dm_sparse 128 9.5 0.001 0.001 6.502 6.583 fft3d_ps 1291 14.7 3.131 3.242 6.504 6.534 dbcsr_complete_redistribute 395 12.7 2.381 2.440 5.781 6.173 calculate_first_density_matrix 1 7.0 0.000 0.001 6.059 6.060 grid_collocate_task_list 128 9.7 5.572 5.812 5.572 5.812 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.683 5.702 qs_energies_init_hamiltonians 11 5.9 0.011 0.022 5.378 5.379 mp_alltoall_d11v 2415 14.1 5.087 5.370 5.087 5.370 mp_allgather_i34 2507 14.6 1.173 4.726 1.173 4.726 potential_pw2rs 128 12.3 0.027 0.028 4.615 4.619 parallel_gemm_fm 81 9.0 0.000 0.000 4.572 4.590 parallel_gemm_fm_cosma 81 10.0 4.571 4.590 4.571 4.590 make_basis_sm 11 9.8 0.000 0.000 4.401 4.404 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.230 4.281 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.844 4.182 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.586 3.905 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.692 3.761 multiply_cannon_metrocomm4 7521 15.6 0.025 0.028 1.726 3.734 mp_irecv_dv 28860 15.9 1.686 3.658 1.686 3.658 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.562 3.649 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.594 3.621 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.621000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1456.454545, yerr=47.200469 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1971389988864 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.241596E+12 0.0% 0.0% 100.0% flops max/rank 11.787141E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806515136 0.0% 0.0% 100.0% number of processed stacks 1981760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.6 marketing flops 145.659521E+12 ------------------------------------------------------------------------------- # multiplications 2533 max memory usage/rank 3.192783E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101320 MPI messages size (bytes): total size 1.145104E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.301851E+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 45808 35567697920 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4065 58253. MP_Allreduce 11240 1499. MP_Sync 86 MP_Alltoall 1724 36993608. 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 0.030 0.051 299.334 299.340 qs_mol_dyn_low 1 2.0 0.003 0.003 298.368 298.393 qs_forces 11 3.9 0.004 0.005 297.076 297.079 qs_energies 11 4.9 0.002 0.002 288.060 288.064 scf_env_do_scf 11 5.9 0.001 0.001 264.749 264.753 velocity_verlet 10 3.0 0.002 0.002 215.631 215.787 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 139.457 139.460 init_scf_loop 11 6.9 0.000 0.000 125.015 125.020 prepare_preconditioner 11 7.9 0.000 0.000 120.060 120.085 make_preconditioner 11 8.9 0.000 0.000 120.060 120.085 make_full_inverse_cholesky 11 9.9 0.038 0.039 96.249 117.184 qs_scf_new_mos 118 7.6 0.001 0.001 93.618 93.681 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.617 93.680 ot_scf_mini 118 9.6 0.004 0.004 88.611 88.643 dbcsr_multiply_generic 2533 12.6 0.226 0.236 84.531 85.130 cp_fm_upper_to_full 104 14.8 52.665 75.434 52.665 75.434 multiply_cannon 2533 13.6 0.684 0.713 60.841 61.966 multiply_cannon_loop 2533 14.6 1.060 1.089 56.706 58.205 ot_mini 118 10.6 0.001 0.001 45.832 45.867 dbcsr_complete_redistribute 393 12.7 3.983 4.037 29.905 42.556 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.336 38.986 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.769 36.280 rebuild_ks_matrix 129 8.3 0.001 0.001 35.472 35.529 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.019 35.471 35.529 mp_alltoall_i22 716 14.1 21.539 34.333 21.539 34.333 cp_fm_cholesky_invert 11 10.9 33.961 33.968 33.961 33.968 mp_waitall_1 104740 16.8 29.225 33.879 29.225 33.879 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.917 32.972 qs_ot_get_p 129 10.4 0.007 0.008 27.403 27.462 qs_ot_get_derivative 118 11.6 0.002 0.002 25.740 25.775 qs_ot_p2m_diag 82 11.4 0.868 0.873 23.098 23.130 multiply_cannon_metrocomm3 10132 15.6 0.025 0.025 20.001 21.519 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.268 21.270 make_m2s 5066 13.6 0.074 0.077 18.763 20.315 ot_diis_step 118 11.6 0.022 0.024 20.050 20.051 make_images 5066 14.6 3.074 3.261 18.278 19.831 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.376 19.558 apply_single 129 13.6 0.001 0.001 19.375 19.557 multiply_cannon_multrec 10132 15.6 10.351 12.159 18.585 18.664 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.990 17.991 cp_fm_diag_elpa_base 82 14.4 13.661 15.280 17.986 17.986 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.749 17.765 calculate_rho_elec 129 8.7 0.482 0.482 17.749 17.764 sum_up_and_integrate 129 10.3 0.002 0.002 16.863 16.952 integrate_v_rspace 129 11.3 0.004 0.004 16.800 16.889 multiply_cannon_sync_h2d 10132 15.6 14.410 14.440 14.410 14.440 pw_transfer 1559 11.6 0.096 0.097 13.558 13.563 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 13.318 13.324 init_scf_run 11 5.9 0.000 0.001 12.912 12.913 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.911 12.913 hybrid_alltoall_any 5251 16.5 1.329 3.100 10.576 12.884 make_images_data 5066 15.6 0.064 0.070 10.464 12.802 fft_wrap_pw1pw2_140 527 13.2 3.240 3.308 11.800 11.807 dbcsr_mm_accdrv_process 20950 16.1 4.436 6.255 7.981 9.933 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.882 9.913 cp_fm_cholesky_decompose 22 10.9 9.391 9.413 9.391 9.413 wfi_extrapolate 11 7.9 0.001 0.001 9.311 9.311 grid_integrate_task_list 129 12.3 8.622 8.813 8.622 8.813 density_rs2pw 129 9.7 0.005 0.005 8.638 8.690 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 8.209 8.210 fft3d_ps 1301 14.7 3.999 4.013 7.743 7.814 calculate_dm_sparse 129 9.5 0.001 0.001 7.316 7.382 mp_alltoall_d11v 2417 14.1 7.242 7.373 7.242 7.373 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.514 6.553 grid_collocate_task_list 129 9.7 6.470 6.545 6.470 6.545 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.265 6.329 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.340000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2878.181818, yerr=154.014060 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 1.260782E+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 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 0.018 0.052 87.799 87.800 qs_energies 1 2.0 0.001 0.007 87.278 87.296 ls_scf 1 3.0 0.000 0.001 85.656 85.669 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.580 72.748 multiply_cannon 111 7.7 0.017 0.020 55.879 56.934 multiply_cannon_loop 111 8.7 0.226 0.243 52.435 53.768 ls_scf_main 1 4.0 0.001 0.026 52.360 52.361 density_matrix_trs4 2 5.0 0.002 0.011 46.806 46.898 ls_scf_init_scf 1 4.0 0.000 0.002 30.057 30.058 ls_scf_init_matrix_S 1 5.0 0.000 0.001 28.428 28.496 mp_waitall_1 11031 10.9 22.572 25.710 22.572 25.710 matrix_sqrt_Newton_Schulz 2 6.5 0.179 0.248 25.649 25.670 multiply_cannon_multrec 2664 9.7 8.143 8.997 15.514 17.307 multiply_cannon_sync_h2d 2664 9.7 13.408 15.151 13.408 15.151 make_m2s 222 7.7 0.009 0.011 12.978 13.559 make_images 222 8.7 0.098 0.109 12.955 13.538 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.899 13.510 hybrid_alltoall_any 227 10.6 0.216 1.842 7.014 8.428 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.438 8.294 make_images_data 222 9.7 0.004 0.005 7.570 8.152 dbcsr_mm_accdrv_process 4760 10.4 0.596 0.726 6.987 7.955 dbcsr_mm_accdrv_process_sort 4760 11.4 6.190 7.073 6.190 7.073 calculate_norms 4752 9.8 5.489 6.076 5.489 6.076 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.084 5.224 mp_sum_l 887 5.1 3.200 4.390 3.200 4.390 arnoldi_extremal 4 6.8 0.001 0.005 3.766 3.964 arnoldi_normal_ev 4 7.8 0.048 0.198 3.765 3.964 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.034 3.643 mp_irecv_dv 6231 10.9 2.018 3.609 2.018 3.609 make_images_sizes 222 9.7 0.000 0.000 0.674 3.460 mp_alltoall_i44 222 10.7 0.674 3.460 0.674 3.460 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.387 3.434 ls_scf_post 1 4.0 0.002 0.013 3.239 3.251 build_subspace 16 8.4 0.009 0.013 3.162 3.165 ls_scf_store_result 1 5.0 0.000 0.000 2.856 2.890 dbcsr_special_finalize 555 9.7 0.005 0.006 2.379 2.750 dbcsr_merge_single_wm 555 10.7 0.464 0.598 2.370 2.742 make_images_pack 222 9.7 2.208 2.635 2.210 2.637 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.357 2.601 dbcsr_matrix_vector_mult_local 304 10.0 2.113 2.512 2.115 2.514 dbcsr_sort_data 658 11.4 2.165 2.468 2.165 2.468 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.310 2.391 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.254 2.255 rebuild_ks_matrix 3 7.3 0.000 0.000 2.244 2.246 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.244 2.245 buffer_matrices_ensure_size 222 8.7 1.753 2.086 1.753 2.086 mp_sum_d 1065 5.9 1.272 1.891 1.272 1.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.800000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 2.153558E+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 0.146 0.484 93.564 93.567 qs_energies 1 2.0 0.000 0.000 92.244 92.248 ls_scf 1 3.0 0.000 0.000 90.674 90.679 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.679 75.034 multiply_cannon 111 7.7 0.027 0.040 52.855 57.355 ls_scf_main 1 4.0 0.000 0.000 54.896 54.901 multiply_cannon_loop 111 8.7 0.135 0.145 49.963 53.828 density_matrix_trs4 2 5.0 0.002 0.003 49.240 49.450 ls_scf_init_scf 1 4.0 0.000 0.001 31.709 31.710 mp_waitall_1 9105 10.9 20.822 30.786 20.822 30.786 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.324 30.430 multiply_cannon_multrec 1332 9.7 13.352 17.164 22.609 27.533 matrix_sqrt_Newton_Schulz 2 6.5 0.106 0.184 27.341 27.354 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.591 21.000 make_m2s 222 7.7 0.006 0.007 15.171 15.813 make_images 222 8.7 1.369 1.696 15.140 15.783 dbcsr_mm_accdrv_process 4041 10.4 0.337 0.541 8.849 10.425 hybrid_alltoall_any 227 10.6 0.541 2.543 8.565 10.140 dbcsr_mm_accdrv_process_sort 4041 11.4 8.368 9.884 8.368 9.884 make_images_data 222 9.7 0.004 0.005 8.775 9.682 mp_sum_l 887 5.1 5.170 8.557 5.170 8.557 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.192 7.564 mp_irecv_dv 3311 11.0 3.172 7.510 3.172 7.510 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.942 6.823 multiply_cannon_sync_h2d 1332 9.7 4.897 6.712 4.897 6.712 calculate_norms 2376 9.8 6.056 6.661 6.056 6.661 arnoldi_extremal 4 6.8 0.000 0.000 5.316 5.422 arnoldi_normal_ev 4 7.8 0.132 0.198 5.316 5.422 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.913 5.147 build_subspace 16 8.4 0.014 0.021 4.466 4.469 ls_scf_post 1 4.0 0.000 0.000 4.069 4.074 ls_scf_store_result 1 5.0 0.000 0.000 3.421 3.516 dbcsr_matrix_vector_mult 304 9.0 0.010 0.023 3.156 3.376 dbcsr_matrix_vector_mult_local 304 10.0 2.748 3.219 2.750 3.221 mp_allgather_i34 111 8.7 0.860 3.116 0.860 3.116 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.612 2.704 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.099 2.595 dbcsr_data_new 4174 10.1 2.117 2.415 2.117 2.415 make_images_pack 222 9.7 1.826 2.122 1.828 2.124 dbcsr_sort_data 436 11.2 1.868 2.121 1.868 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.891 1.893 rebuild_ks_matrix 3 7.3 0.000 0.000 1.878 1.880 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.878 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.567000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1775.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 2.882703E+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 0.025 0.055 93.848 93.849 qs_energies 1 2.0 0.000 0.000 93.231 93.234 ls_scf 1 3.0 0.000 0.000 91.196 91.200 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.052 75.320 ls_scf_main 1 4.0 0.000 0.000 56.260 56.264 multiply_cannon 111 7.7 0.036 0.099 51.922 55.606 multiply_cannon_loop 111 8.7 0.116 0.129 49.197 53.751 density_matrix_trs4 2 5.0 0.002 0.003 50.335 50.542 mp_waitall_1 7281 11.0 23.448 33.448 23.448 33.448 ls_scf_init_scf 1 4.0 0.000 0.001 31.225 31.228 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.037 30.112 matrix_sqrt_Newton_Schulz 2 6.5 0.189 0.222 27.566 27.577 multiply_cannon_multrec 888 9.7 12.639 15.573 21.197 24.730 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.051 22.874 make_m2s 222 7.7 0.006 0.007 16.133 16.851 make_images 222 8.7 1.584 1.863 16.095 16.811 make_images_data 222 9.7 0.004 0.004 9.461 10.451 hybrid_alltoall_any 227 10.6 0.642 2.946 9.122 10.232 dbcsr_mm_accdrv_process 3754 10.4 0.322 0.540 8.083 9.386 dbcsr_mm_accdrv_process_sort 3754 11.4 7.637 8.845 7.637 8.845 mp_sum_l 887 5.1 5.299 8.814 5.299 8.814 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.468 7.588 multiply_cannon_sync_h2d 888 9.7 6.097 7.569 6.097 7.569 mp_irecv_dv 2335 11.1 2.451 7.524 2.451 7.524 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.069 6.866 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.581 6.779 arnoldi_extremal 4 6.8 0.000 0.000 5.661 5.850 arnoldi_normal_ev 4 7.8 0.166 0.247 5.660 5.850 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.764 4.990 build_subspace 16 8.4 0.014 0.020 4.763 4.770 calculate_norms 1584 9.8 4.338 4.753 4.338 4.753 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.448 3.765 ls_scf_post 1 4.0 0.000 0.000 3.711 3.715 mp_allgather_i34 111 8.7 0.837 3.648 0.837 3.648 dbcsr_matrix_vector_mult_local 304 10.0 3.030 3.603 3.032 3.605 ls_scf_store_result 1 5.0 0.000 0.000 3.442 3.534 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.847 2.957 dbcsr_data_new 4116 9.9 2.106 2.437 2.106 2.437 make_images_sizes 222 9.7 0.000 0.000 0.882 2.278 mp_alltoall_i44 222 10.7 0.882 2.277 0.882 2.277 dbcsr_sort_data 325 11.1 1.840 2.096 1.840 2.096 qs_energies_init_hamiltonians 1 3.0 0.020 0.051 2.027 2.027 dbcsr_finalize 304 7.8 0.026 0.032 1.604 1.899 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.892 1.894 make_images_pack 222 9.7 1.622 1.877 1.625 1.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.849000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2252.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 3.320074E+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 0.030 0.047 98.463 98.465 qs_energies 1 2.0 0.000 0.000 97.640 97.645 ls_scf 1 3.0 0.000 0.000 95.467 95.471 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.068 78.322 ls_scf_main 1 4.0 0.000 0.000 58.531 58.533 multiply_cannon 111 7.7 0.053 0.109 51.543 55.998 density_matrix_trs4 2 5.0 0.002 0.003 52.410 52.552 multiply_cannon_loop 111 8.7 0.153 0.169 46.516 48.976 ls_scf_init_scf 1 4.0 0.000 0.001 33.666 33.669 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.466 32.519 matrix_sqrt_Newton_Schulz 2 6.5 0.206 0.543 29.981 29.992 mp_waitall_1 6369 11.0 22.555 28.810 22.555 28.810 multiply_cannon_multrec 1332 9.7 14.118 17.526 22.019 25.093 make_m2s 222 7.7 0.006 0.007 21.011 22.434 make_images 222 8.7 3.137 3.599 20.961 22.387 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.213 16.993 make_images_data 222 9.7 0.004 0.005 11.703 13.208 hybrid_alltoall_any 227 10.6 0.796 3.754 11.077 12.928 dbcsr_mm_accdrv_process 3641 10.4 0.311 0.553 7.547 9.041 dbcsr_mm_accdrv_process_sort 3641 11.4 7.083 8.546 7.083 8.546 mp_sum_l 887 5.1 3.924 7.255 3.924 7.255 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.095 6.247 mp_irecv_dv 3229 10.9 2.072 6.166 2.072 6.166 arnoldi_extremal 4 6.8 0.000 0.000 5.835 6.046 arnoldi_normal_ev 4 7.8 0.132 0.394 5.835 6.046 multiply_cannon_sync_h2d 1332 9.7 5.459 6.011 5.459 6.011 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.025 5.723 build_subspace 16 8.4 0.014 0.021 4.956 4.965 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.752 mp_allgather_i34 111 8.7 2.145 4.490 2.145 4.490 calculate_norms 2376 9.8 4.159 4.488 4.159 4.488 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.467 4.388 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.629 3.936 dbcsr_matrix_vector_mult_local 304 10.0 3.211 3.704 3.214 3.706 dbcsr_sort_data 658 11.4 3.108 3.404 3.108 3.404 ls_scf_post 1 4.0 0.000 0.000 3.270 3.274 dbcsr_special_finalize 555 9.7 0.006 0.007 2.854 3.150 dbcsr_merge_single_wm 555 10.7 0.541 0.658 2.846 3.142 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.018 3.074 ls_scf_store_result 1 5.0 0.000 0.000 3.012 3.073 dbcsr_data_release 10477 10.7 1.577 2.442 1.577 2.442 qs_energies_init_hamiltonians 1 3.0 0.004 0.009 2.164 2.164 dbcsr_finalize 304 7.8 0.049 0.061 1.805 1.983 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.465000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 4.745802E+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 0.036 0.057 93.189 93.190 qs_energies 1 2.0 0.000 0.000 92.402 92.407 ls_scf 1 3.0 0.000 0.000 90.417 90.420 dbcsr_multiply_generic 111 6.7 0.017 0.019 71.903 72.161 ls_scf_main 1 4.0 0.000 0.000 57.130 57.130 multiply_cannon 111 7.7 0.083 0.174 53.104 56.229 multiply_cannon_loop 111 8.7 0.088 0.093 50.481 52.219 density_matrix_trs4 2 5.0 0.002 0.003 50.167 50.249 ls_scf_init_scf 1 4.0 0.000 0.001 29.919 29.923 mp_waitall_1 5436 11.0 25.156 29.525 25.156 29.525 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.660 28.703 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.463 26.474 multiply_cannon_multrec 444 9.7 13.660 16.295 20.818 22.237 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.118 16.012 make_m2s 222 7.7 0.005 0.005 13.824 14.815 make_images 222 8.7 2.035 2.483 13.756 14.745 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.293 14.638 hybrid_alltoall_any 227 10.6 0.800 3.811 8.343 10.119 make_images_data 222 9.7 0.003 0.004 8.563 9.889 multiply_cannon_sync_h2d 444 9.7 6.758 8.046 6.758 8.046 dbcsr_mm_accdrv_process 3003 10.4 0.421 0.764 6.839 7.999 dbcsr_mm_accdrv_process_sort 3003 11.4 6.384 7.458 6.384 7.458 arnoldi_extremal 4 6.8 0.000 0.000 5.812 5.826 arnoldi_normal_ev 4 7.8 0.002 0.005 5.812 5.825 build_subspace 16 8.4 0.015 0.020 5.413 5.421 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.384 4.567 mp_sum_l 887 5.1 2.963 4.552 2.963 4.552 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.163 4.309 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.089 3.700 4.091 mp_allgather_i34 111 8.7 1.209 3.838 1.209 3.838 calculate_norms 792 9.8 3.607 3.731 3.607 3.731 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.547 3.679 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.164 3.629 mp_irecv_dv 1241 11.2 1.530 3.619 1.530 3.619 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.527 3.596 make_images_sizes 222 9.7 0.000 0.000 0.888 3.464 mp_alltoall_i44 222 10.7 0.888 3.464 0.888 3.464 ls_scf_post 1 4.0 0.000 0.000 3.368 3.373 ls_scf_store_result 1 5.0 0.000 0.000 3.157 3.191 dbcsr_data_new 4608 9.7 1.787 2.315 1.787 2.315 dbcsr_finalize 304 7.8 0.062 0.077 2.208 2.299 dbcsr_merge_all 275 8.9 0.479 0.527 2.066 2.139 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.032 2.032 rebuild_ks_matrix 3 7.3 0.000 0.000 2.000 2.000 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.000 2.000 qs_energies_init_hamiltonians 1 3.0 0.004 0.017 1.970 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.190000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3685.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 8.850829E+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 0.084 0.125 109.332 109.333 qs_energies 1 2.0 0.000 0.000 107.953 107.956 ls_scf 1 3.0 0.000 0.000 104.653 104.656 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.909 78.059 ls_scf_main 1 4.0 0.000 0.000 63.619 63.764 density_matrix_trs4 2 5.0 0.002 0.003 54.485 54.547 multiply_cannon 111 7.7 0.145 0.290 51.912 53.922 multiply_cannon_loop 111 8.7 0.098 0.099 48.917 49.518 ls_scf_init_scf 1 4.0 0.001 0.001 37.246 37.267 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.534 35.559 matrix_sqrt_Newton_Schulz 2 6.5 0.006 0.024 32.834 32.844 mp_waitall_1 4527 11.1 22.157 25.549 22.157 25.549 make_m2s 222 7.7 0.005 0.005 22.365 23.473 make_images 222 8.7 3.568 3.868 22.257 23.363 multiply_cannon_multrec 444 9.7 17.790 18.429 22.423 23.084 hybrid_alltoall_any 227 10.6 1.652 3.604 12.613 15.408 make_images_data 222 9.7 0.003 0.004 12.778 14.933 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.769 14.132 multiply_cannon_sync_h2d 444 9.7 8.788 8.836 8.788 8.836 arnoldi_extremal 4 6.8 0.000 0.000 7.516 7.539 arnoldi_normal_ev 4 7.8 0.006 0.032 7.516 7.539 build_subspace 16 8.4 0.026 0.036 6.941 6.953 dbcsr_matrix_vector_mult 304 9.0 0.017 0.035 5.507 5.676 dbcsr_matrix_vector_mult_local 304 10.0 5.059 5.382 5.061 5.384 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.217 5.311 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.888 5.134 dbcsr_mm_accdrv_process 1814 10.4 0.273 0.384 4.451 4.600 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.238 4.109 4.238 compute_matrix_preconditioner 1 6.0 0.002 0.002 4.169 4.175 ls_scf_post 1 4.0 0.000 0.000 3.789 3.792 acc_transpose_blocks 444 9.7 0.003 0.003 3.202 3.630 make_images_sizes 222 9.7 0.000 0.000 1.464 3.562 mp_alltoall_i44 222 10.7 1.464 3.562 1.464 3.562 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 3.119 3.547 jit_kernel_transpose 1 13.0 3.113 3.541 3.113 3.541 ls_scf_store_result 1 5.0 0.000 0.000 3.497 3.534 mp_allgather_i34 111 8.7 1.120 3.533 1.120 3.533 calculate_norms 792 9.8 3.234 3.272 3.234 3.272 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 3.269 3.269 dbcsr_finalize 304 7.8 0.082 0.089 3.079 3.170 dbcsr_merge_all 275 8.9 0.885 0.921 2.864 2.950 dbcsr_complete_redistribute 5 7.6 1.449 1.481 2.762 2.866 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.621 2.643 rebuild_ks_matrix 3 7.3 0.000 0.000 2.555 2.577 qs_ks_build_kohn_sham_matrix 3 8.3 0.145 0.166 2.555 2.577 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.526 dbcsr_sort_data 325 11.1 2.441 2.499 2.441 2.499 dbcsr_data_new 6591 9.6 1.827 2.326 1.827 2.326 dbcsr_new_transposed 4 7.5 0.242 0.253 2.272 2.281 dbcsr_frobenius_norm 74 6.6 2.053 2.137 2.201 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.124 2.206 dbcsr_add_anytype 103 7.2 0.858 0.891 2.124 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.333000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6978.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b263019182a81f50cd70653ab7c48fb92fe52198_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 594.259968E+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 1331638041. 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 0.059 0.120 293.649 293.652 qs_mol_dyn_low 1 2.0 0.004 0.032 292.313 292.406 qs_forces 5 3.8 0.005 0.025 291.310 291.314 qs_energies 5 4.8 0.003 0.030 287.841 287.872 scf_env_do_scf 5 5.8 0.020 0.425 271.144 271.147 scf_env_do_scf_inner_loop 105 6.6 0.005 0.068 236.121 236.123 qs_scf_new_mos 105 7.6 0.000 0.001 185.211 185.411 qs_scf_loop_do_ot 105 8.6 0.001 0.001 185.211 185.410 ot_scf_mini 105 9.6 0.003 0.024 175.011 175.159 dbcsr_multiply_generic 1445 12.2 0.130 0.138 138.676 139.136 velocity_verlet 4 3.0 0.009 0.385 120.334 120.337 multiply_cannon 1445 13.2 0.275 0.284 117.116 119.473 multiply_cannon_loop 1445 14.2 2.832 2.981 114.708 116.147 qs_ot_get_p 112 10.4 0.001 0.002 81.324 81.564 qs_ot_p2m_diag 40 11.0 0.020 0.037 68.958 69.057 cp_dbcsr_syevd 40 12.0 0.002 0.002 65.299 65.301 ot_mini 105 10.6 0.001 0.012 63.125 63.247 cp_fm_syevd 40 13.0 0.000 0.000 59.033 59.172 cp_fm_redistribute_end 40 14.0 26.693 53.248 26.704 53.254 cp_fm_syevd_base 40 14.0 26.533 53.089 26.533 53.089 mp_waitall_1 488190 16.1 39.653 46.907 39.653 46.907 multiply_cannon_multrec 69360 15.2 29.667 35.421 40.248 45.715 qs_ot_get_derivative 55 11.6 0.001 0.010 41.159 41.297 multiply_cannon_metrocomm3 69360 15.2 0.201 0.213 26.768 35.112 init_scf_loop 7 6.6 0.000 0.004 34.947 34.970 rebuild_ks_matrix 110 8.4 0.000 0.000 33.609 34.024 qs_ks_build_kohn_sham_matrix 110 9.4 0.138 0.455 33.609 34.023 multiply_cannon_sync_h2d 69360 15.2 28.851 32.723 28.851 32.723 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.894 31.279 prepare_preconditioner 7 7.6 0.000 0.001 28.792 28.947 make_preconditioner 7 8.6 0.002 0.098 28.792 28.947 qs_rho_update_rho_low 110 7.6 0.001 0.003 24.367 24.773 calculate_rho_elec 110 8.6 0.030 0.032 24.367 24.772 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.226 23.469 apply_single 62 13.6 0.000 0.000 23.226 23.469 ot_new_cg_direction 55 11.6 0.001 0.003 21.215 21.217 make_full_inverse_cholesky 7 9.6 0.000 0.000 20.057 20.118 density_rs2pw 110 9.6 0.005 0.006 18.113 18.534 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.170 15.321 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.751 14.886 transfer_rs2pw 445 10.6 0.007 0.008 14.248 14.708 mp_sum_l 4764 12.2 12.892 13.807 12.892 13.807 init_scf_run 5 5.8 0.000 0.001 13.803 13.804 scf_env_initial_rho_setup 5 6.8 0.001 0.002 13.803 13.804 cp_fm_cholesky_invert 7 10.6 13.646 13.664 13.646 13.664 pw_transfer 1645 12.4 0.079 0.100 13.148 13.491 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 13.009 13.355 calculate_dm_sparse 110 9.5 0.000 0.001 12.391 12.613 sum_up_and_integrate 60 10.3 0.001 0.004 12.161 12.179 integrate_v_rspace 60 11.3 0.002 0.030 12.144 12.161 fft_wrap_pw1pw2_240 915 15.0 0.819 0.907 11.206 11.519 dbcsr_mm_accdrv_process 154766 15.8 6.165 6.351 10.449 11.406 check_diag 80 13.5 8.683 8.935 11.217 11.360 qs_vxc_create 110 10.4 0.002 0.009 11.094 11.156 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 11.074 11.134 fft3d_pb 915 16.0 2.384 2.687 9.421 9.860 make_m2s 2890 13.2 0.077 0.084 8.240 8.841 make_images 2890 14.2 0.240 0.260 8.133 8.734 transfer_rs2pw_30 110 11.6 1.248 1.317 7.610 8.712 make_full_single_inverse 7 9.6 0.001 0.002 8.363 8.520 mp_sendrecv_dv 168740 12.6 8.234 8.456 8.234 8.456 acc_transpose_blocks 69360 15.2 0.358 0.382 8.008 8.413 calculate_first_density_matrix 1 7.0 0.000 0.004 8.386 8.406 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.226 8.242 potential_pw2rs 60 12.3 0.002 0.003 8.095 8.177 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 4.891 7.974 xc_rho_set_and_dset_create 110 12.4 0.075 0.095 7.716 7.967 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.677 7.726 xc_vxc_pw_create 60 11.3 0.039 0.050 7.460 7.522 mp_alltoall_z22v 2340 17.7 6.575 7.054 6.575 7.054 transfer_pw2rs 245 13.2 0.003 0.004 6.868 6.940 xc_pw_derive 510 13.4 0.006 0.007 6.773 6.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=293.652000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b263019182a81f50cd70653ab7c48fb92fe52198 Summary: empty Status: OK