=== This is the CP2K Performance-Test === Updating e5825b641..2746062f8 Fast-forward src/fist_nonbond_env_types.F | 74 +++++++++++++++++++++----------------- src/qs_diis_types.F | 32 ++++++++++------- tools/conventions/conventions.supp | 51 -------------------------- 3 files changed, 61 insertions(+), 96 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 2746062f82ececc9c9d99636991defc71e6cb498 ################# 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 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # 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. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/01 job id: 45998912 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/02 job id: 45998914 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/03 job id: 45998916 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/04 job id: 45998917 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/05 job id: 45998918 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/06 job id: 45998919 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/07 job id: 45998920 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/08 job id: 45998921 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/09 job id: 45998923 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/10 job id: 45998925 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/11 job id: 45998930 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/12 job id: 45998935 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/13 job id: 45998938 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/14 job id: 45998942 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/15 job id: 45998945 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/16 job id: 45998947 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/17 job id: 45998950 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/18 job id: 45998952 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/19 job id: 45998954 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/20 job id: 45998958 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/21 job id: 45998961 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/22 job id: 45998962 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/23 job id: 45998964 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/24 job id: 45998968 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/25 job id: 45998973 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/26 job id: 45998980 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/27 job id: 45998982 --- 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/2746062f82ececc9c9d99636991defc71e6cb498_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 344 9. 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 0.022 0.034 136.052 136.053 farming_run 1 2.0 135.521 135.522 136.020 136.023 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.466671E+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 1103589. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 116.271 116.271 qs_energies 1 2.0 0.000 0.000 116.072 116.074 mp2_main 1 3.0 0.000 0.000 113.699 113.701 mp2_gpw_main 1 4.0 0.019 0.026 112.530 112.532 mp2_ri_gpw_compute_in 1 5.0 0.176 0.183 93.126 93.572 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.483 55.930 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.164 41.816 47.221 get_2c_integrals 1 6.0 0.000 0.000 36.959 37.469 integrate_v_rspace 273 8.0 0.439 0.454 25.174 30.375 pw_transfer 6555 10.6 0.378 0.394 27.413 28.155 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 26.093 26.771 grid_integrate_task_list 273 9.0 20.965 26.668 20.965 26.668 fft_wrap_pw1pw2_100 2178 12.4 1.174 1.306 23.612 24.284 rpa_ri_compute_en 1 5.0 0.001 0.001 19.298 19.582 compute_2c_integrals 1 7.0 0.003 0.003 19.490 19.495 compute_2c_integrals_loop_lm 1 8.0 0.013 0.019 18.853 19.176 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.425 18.840 19.168 cp_fm_cholesky_decompose 12 8.2 17.572 18.232 17.572 18.232 cholesky_decomp 1 7.0 0.000 0.000 16.312 16.807 fft3d_s 5443 13.4 16.167 16.685 16.189 16.706 ao_to_mo_and_store_B_mult_1 272 7.0 10.845 15.557 10.845 15.557 calculate_wavefunction 272 8.0 5.442 5.601 12.574 13.166 rpa_num_int 1 6.0 0.000 0.001 10.868 10.868 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.667 10.867 calc_mat_Q 8 8.0 0.000 0.000 9.368 9.485 contract_S_to_Q 8 9.0 0.000 0.000 8.788 8.906 calc_potential_gpw 544 9.5 0.005 0.005 8.228 8.597 parallel_gemm_fm 14 9.1 0.000 0.000 8.368 8.476 parallel_gemm_fm_cosma 14 10.1 8.368 8.476 8.368 8.476 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.179 8.455 potential_pw2rs 545 10.0 0.107 0.109 7.692 8.363 create_integ_mat 1 6.0 0.022 0.027 7.887 7.887 collocate_single_gaussian 272 10.0 0.039 0.042 7.459 7.716 array2fm 1 7.0 0.000 0.000 6.819 7.231 pw_scatter_s 2720 13.7 4.434 4.595 4.434 4.595 pw_gather_s 2722 13.2 3.906 4.286 3.906 4.286 array2fm_buffer_send 1 8.0 3.051 3.157 3.051 3.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.532084, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.036 395.781 395.783 farming_run 1 2.0 394.945 394.949 395.748 395.751 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222439E+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 743 386399. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 21132749. 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 0.010 0.030 209.744 209.744 qs_energies 1 2.0 0.000 0.000 209.473 209.486 scf_env_do_scf 1 3.0 0.000 0.000 106.291 106.292 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.400 105.407 rebuild_ks_matrix 4 6.0 0.000 0.000 105.398 105.406 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.398 105.406 hfx_ks_matrix 4 8.0 0.001 0.001 105.016 105.020 integrate_four_center 4 9.0 0.143 0.451 105.015 105.020 mp2_main 1 3.0 0.000 0.000 102.896 102.908 mp2_gpw_main 1 4.0 0.032 0.048 102.056 102.072 integrate_four_center_main 4 10.0 0.102 0.567 96.958 99.222 integrate_four_center_bin 263 11.0 96.856 98.780 96.856 98.780 init_scf_loop 1 4.0 0.000 0.000 92.105 92.105 mp2_ri_gpw_compute_in 1 5.0 0.066 0.094 74.970 75.999 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.555 55.583 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.141 47.333 integrate_v_rspace 95 8.0 0.398 0.567 28.512 33.690 pw_transfer 2240 10.6 0.141 0.161 29.985 30.314 ao_to_mo_and_store_B_mult_1 91 7.0 10.720 29.511 10.720 29.511 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.983 29.285 grid_integrate_task_list 95 9.0 23.793 29.144 23.793 29.144 mp2_ri_gpw_compute_en 1 5.0 0.060 0.075 26.939 28.675 fft_wrap_pw1pw2_100 730 12.4 1.291 1.426 26.704 27.027 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.825 1.888 25.263 25.275 get_2c_integrals 1 6.0 0.000 0.000 20.319 20.350 compute_2c_integrals 1 7.0 0.005 0.021 19.306 19.312 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.919 19.161 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.817 18.917 19.159 fft3d_s 1823 13.4 18.474 18.796 18.488 18.809 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.185 14.185 calculate_wavefunction 91 8.0 2.013 2.042 9.743 9.985 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.589 8.843 9.473 potential_pw2rs 186 10.0 0.034 0.036 8.700 9.241 local_gemm 172 8.0 8.285 8.889 8.285 8.889 mp2_ri_gpw_compute_en_comm 22 7.0 0.502 0.525 8.206 8.724 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.284 8.645 collocate_single_gaussian 91 10.0 0.016 0.021 7.915 8.163 calc_potential_gpw 182 9.5 0.002 0.002 7.900 8.097 mp_sendrecv_dm3 2068 8.0 6.236 6.760 6.236 6.760 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.441 6.345 6.441 pw_gather_s 912 13.2 4.933 5.407 4.933 5.407 mp_sync 37 10.5 2.747 4.881 2.747 4.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.068400, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1509.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 452.628480E+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 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 51.149 51.150 qs_mol_dyn_low 1 2.0 0.003 0.003 50.791 50.799 qs_forces 11 3.9 0.002 0.003 50.730 50.731 qs_energies 11 4.9 0.001 0.001 49.259 49.277 scf_env_do_scf 11 5.9 0.000 0.001 43.334 43.335 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.362 41.362 dbcsr_multiply_generic 2286 12.5 0.093 0.098 31.596 31.948 qs_scf_new_mos 108 7.5 0.000 0.001 31.443 31.721 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.442 31.720 ot_scf_mini 108 9.5 0.002 0.002 29.878 30.059 multiply_cannon 2286 13.5 0.188 0.196 25.135 26.452 multiply_cannon_loop 2286 14.5 1.466 1.557 24.475 25.830 velocity_verlet 10 3.0 0.001 0.002 24.667 24.668 ot_mini 108 10.5 0.001 0.001 17.774 18.026 qs_ot_get_derivative 108 11.5 0.001 0.002 14.952 15.134 mp_waitall_1 245248 16.5 7.482 13.589 7.482 13.589 multiply_cannon_metrocomm3 54864 15.5 0.068 0.075 5.452 11.939 multiply_cannon_multrec 54864 15.5 4.322 6.751 7.561 11.024 qs_ot_get_p 119 10.4 0.001 0.001 7.687 8.000 rebuild_ks_matrix 119 8.3 0.000 0.000 7.782 7.934 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 7.782 7.933 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.854 6.990 multiply_cannon_sync_h2d 54864 15.5 6.044 6.648 6.044 6.648 mp_sum_l 7207 12.9 4.620 6.264 4.620 6.264 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.177 5.643 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.150 5.204 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.939 5.028 init_scf_run 11 5.9 0.000 0.001 4.734 4.735 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.734 4.735 dbcsr_mm_accdrv_process 76910 16.1 1.138 1.826 3.159 4.455 sum_up_and_integrate 119 10.3 0.012 0.014 4.444 4.453 integrate_v_rspace 119 11.3 0.002 0.003 4.432 4.442 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.384 4.384 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.196 4.196 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.097 4.193 calculate_rho_elec 119 8.7 0.012 0.017 4.096 4.192 cp_fm_redistribute_end 50 14.0 2.141 4.171 2.146 4.174 cp_fm_diag_elpa_base 50 14.0 2.021 4.064 2.026 4.071 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.901 3.098 apply_single 119 13.6 0.000 0.000 2.901 3.098 calculate_dm_sparse 119 9.5 0.000 0.000 2.853 2.978 rs_pw_transfer 974 11.9 0.012 0.014 2.654 2.764 calculate_first_density_matrix 1 7.0 0.000 0.001 2.621 2.645 ot_diis_step 108 11.5 0.006 0.006 2.632 2.632 jit_kernel_multiply 13 15.8 1.961 2.567 1.961 2.567 multiply_cannon_metrocomm1 54864 15.5 0.054 0.061 1.425 2.434 acc_transpose_blocks 54864 15.5 0.226 0.256 1.815 2.338 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.317 2.319 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.239 2.314 density_rs2pw 119 9.7 0.004 0.004 2.145 2.295 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.094 2.141 grid_integrate_task_list 119 12.3 2.024 2.111 2.024 2.111 wfi_extrapolate 11 7.9 0.001 0.001 2.036 2.036 init_scf_loop 11 6.9 0.000 0.000 1.957 1.957 potential_pw2rs 119 12.3 0.004 0.004 1.806 1.817 pw_transfer 1439 11.6 0.052 0.057 1.680 1.750 mp_sum_d 4135 12.0 1.134 1.745 1.134 1.745 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.605 1.678 make_m2s 4572 13.5 0.054 0.056 1.542 1.589 make_images 4572 14.5 0.133 0.139 1.460 1.506 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.446 1.481 mp_waitany 12084 13.8 1.238 1.407 1.238 1.407 grid_collocate_task_list 119 9.7 1.292 1.364 1.292 1.364 mp_alltoall_d11v 2130 13.8 1.198 1.344 1.198 1.344 fft3d_ps 1201 14.6 0.373 0.478 1.248 1.319 fft_wrap_pw1pw2_140 487 13.2 0.184 0.200 1.240 1.309 acc_transpose_blocks_kernels 54864 16.5 0.245 0.384 0.861 1.287 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.150000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 488.017920E+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 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 1446641. 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 0.031 38.797 38.798 qs_mol_dyn_low 1 2.0 0.003 0.005 37.896 37.903 qs_forces 11 3.9 0.002 0.004 37.837 37.838 qs_energies 11 4.9 0.002 0.009 36.164 36.167 scf_env_do_scf 11 5.9 0.001 0.001 30.947 30.948 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.548 28.549 dbcsr_multiply_generic 2286 12.5 0.101 0.103 21.098 21.464 qs_scf_new_mos 108 7.5 0.001 0.001 19.602 19.847 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.601 19.846 ot_scf_mini 108 9.5 0.003 0.004 18.717 18.885 velocity_verlet 10 3.0 0.002 0.005 17.815 17.816 multiply_cannon 2286 13.5 0.207 0.221 16.169 17.530 multiply_cannon_loop 2286 14.5 0.898 0.974 15.078 16.514 ot_mini 108 10.5 0.001 0.002 11.534 11.763 mp_waitall_1 200699 16.5 5.427 10.827 5.427 10.827 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.023 9.380 qs_ot_get_derivative 108 11.5 0.001 0.002 9.092 9.263 multiply_cannon_multrec 27432 15.5 1.984 4.505 5.897 8.643 rebuild_ks_matrix 119 8.3 0.000 0.000 7.066 7.216 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 7.066 7.215 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.235 6.371 dbcsr_mm_accdrv_process 47894 16.0 2.920 5.077 3.843 5.709 qs_ot_get_p 119 10.4 0.001 0.002 4.330 4.555 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.445 4.357 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.096 4.110 apply_single 119 13.6 0.000 0.000 3.096 4.109 sum_up_and_integrate 119 10.3 0.024 0.027 4.086 4.095 integrate_v_rspace 119 11.3 0.002 0.003 4.062 4.073 init_scf_run 11 5.9 0.000 0.001 3.941 3.942 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.941 3.942 mp_sum_l 7207 12.9 2.055 3.928 2.055 3.928 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.672 3.707 calculate_rho_elec 119 8.7 0.021 0.024 3.672 3.707 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.941 2.959 rs_pw_transfer 974 11.9 0.010 0.011 2.428 2.875 multiply_cannon_sync_h2d 27432 15.5 2.188 2.781 2.188 2.781 make_m2s 4572 13.5 0.053 0.055 2.427 2.671 make_images 4572 14.5 0.200 0.236 2.338 2.580 cp_dbcsr_syevd 50 12.0 0.003 0.006 2.516 2.516 density_rs2pw 119 9.7 0.004 0.005 2.005 2.468 calculate_first_density_matrix 1 7.0 0.001 0.002 2.457 2.459 ot_diis_step 108 11.5 0.011 0.013 2.395 2.395 init_scf_loop 11 6.9 0.001 0.005 2.380 2.381 jit_kernel_multiply 12 16.1 0.871 2.301 0.871 2.301 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.211 2.211 cp_fm_redistribute_end 50 14.0 1.120 2.186 1.123 2.188 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.074 2.165 cp_fm_diag_elpa_base 50 14.0 1.034 2.101 1.061 2.140 calculate_dm_sparse 119 9.5 0.000 0.001 2.020 2.098 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.985 1.987 grid_integrate_task_list 119 12.3 1.840 1.951 1.840 1.951 pw_transfer 1439 11.6 0.065 0.070 1.887 1.918 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.803 1.845 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.795 1.828 potential_pw2rs 119 12.3 0.006 0.006 1.817 1.825 make_images_data 4572 15.5 0.045 0.051 1.144 1.571 acc_transpose_blocks 27432 15.5 0.108 0.113 1.203 1.507 prepare_preconditioner 11 7.9 0.000 0.001 1.479 1.506 make_preconditioner 11 8.9 0.000 0.001 1.479 1.506 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.993 1.484 fft_wrap_pw1pw2_140 487 13.2 0.205 0.217 1.435 1.466 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.379 1.435 wfi_extrapolate 11 7.9 0.001 0.001 1.425 1.425 fft3d_ps 1201 14.6 0.523 0.578 1.343 1.369 grid_collocate_task_list 119 9.7 1.231 1.354 1.231 1.354 mp_allgather_i34 2286 14.5 0.532 1.341 0.532 1.341 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.248 mp_alltoall_d11v 2130 13.8 1.122 1.236 1.122 1.236 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.151 1.199 mp_sum_d 4135 12.0 0.616 1.064 0.616 1.064 qs_energies_init_hamiltonians 11 5.9 0.006 0.041 0.971 1.001 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.546 0.992 mp_waitany 5720 13.7 0.527 0.976 0.527 0.976 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.903 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.671 0.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.798000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 523.231232E+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 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1607886. 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 0.043 0.062 33.040 33.041 qs_mol_dyn_low 1 2.0 0.003 0.003 32.628 32.636 qs_forces 11 3.9 0.002 0.003 32.570 32.572 qs_energies 11 4.9 0.001 0.005 30.996 30.999 scf_env_do_scf 11 5.9 0.001 0.001 25.605 25.605 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 23.039 23.040 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.044 17.140 velocity_verlet 10 3.0 0.001 0.002 15.207 15.208 qs_scf_new_mos 108 7.5 0.001 0.001 14.846 14.863 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.845 14.862 multiply_cannon 2286 13.5 0.195 0.201 13.726 14.565 ot_scf_mini 108 9.5 0.009 0.053 14.126 14.140 multiply_cannon_loop 2286 14.5 0.634 0.660 12.942 13.803 ot_mini 108 10.5 0.001 0.001 8.700 8.725 multiply_cannon_multrec 18288 15.5 1.921 2.827 6.998 7.275 qs_ot_get_derivative 108 11.5 0.001 0.001 7.189 7.212 rebuild_ks_matrix 119 8.3 0.000 0.000 6.385 6.405 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.384 6.404 dbcsr_mm_accdrv_process 38222 16.0 4.146 5.564 4.993 5.818 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.654 5.673 init_scf_run 11 5.9 0.000 0.001 4.201 4.201 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.201 4.201 sum_up_and_integrate 119 10.3 0.031 0.032 3.955 3.960 integrate_v_rspace 119 11.3 0.003 0.003 3.923 3.933 mp_waitall_1 158411 16.6 2.527 3.741 2.527 3.741 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.439 3.444 calculate_rho_elec 119 8.7 0.031 0.032 3.438 3.444 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.713 3.342 qs_ot_get_p 119 10.4 0.001 0.002 3.273 3.295 calculate_first_density_matrix 1 7.0 0.000 0.002 2.986 2.987 init_scf_loop 11 6.9 0.002 0.015 2.541 2.545 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.381 2.501 rs_pw_transfer 974 11.9 0.009 0.010 2.253 2.499 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.380 2.381 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.047 2.363 apply_single 119 13.6 0.000 0.000 2.046 2.363 jit_kernel_multiply 11 16.2 0.794 2.276 0.794 2.276 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.216 2.221 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.214 2.221 density_rs2pw 119 9.7 0.004 0.004 1.956 2.201 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.954 1.955 make_m2s 4572 13.5 0.045 0.046 1.780 1.920 pw_transfer 1439 11.6 0.065 0.070 1.906 1.914 grid_integrate_task_list 119 12.3 1.803 1.901 1.803 1.901 calculate_dm_sparse 119 9.5 0.000 0.001 1.886 1.898 make_images 4572 14.5 0.190 0.201 1.694 1.832 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.813 1.824 acc_transpose_blocks 18288 15.5 0.076 0.077 1.774 1.789 potential_pw2rs 119 12.3 0.007 0.008 1.720 1.725 prepare_preconditioner 11 7.9 0.000 0.000 1.715 1.717 make_preconditioner 11 8.9 0.000 0.001 1.715 1.717 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.698 1.707 cp_fm_diag_elpa_base 50 14.0 1.675 1.686 1.696 1.705 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.571 1.652 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.549 1.557 multiply_cannon_sync_h2d 18288 15.5 1.388 1.539 1.388 1.539 mp_sum_l 7207 12.9 1.150 1.505 1.150 1.505 ot_diis_step 108 11.5 0.011 0.011 1.491 1.491 fft_wrap_pw1pw2_140 487 13.2 0.255 0.261 1.455 1.468 grid_collocate_task_list 119 9.7 1.213 1.360 1.213 1.360 acc_transpose_blocks_kernels 18288 16.5 0.211 0.218 1.332 1.341 fft3d_ps 1201 14.6 0.531 0.545 1.297 1.309 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.257 1.261 wfi_extrapolate 11 7.9 0.001 0.001 1.168 1.168 jit_kernel_transpose 5 15.6 1.121 1.123 1.121 1.123 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.962 0.964 make_images_data 4572 15.5 0.045 0.049 0.795 0.958 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.368 0.898 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.872 0.891 hybrid_alltoall_any 4725 16.4 0.056 0.115 0.689 0.879 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.812 0.815 mp_waitany 9880 13.7 0.510 0.769 0.510 0.769 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.120 0.508 0.752 mp_alltoall_d11v 2130 13.8 0.667 0.742 0.667 0.742 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.738 cp_fm_cholesky_invert 11 10.9 0.706 0.709 0.706 0.709 mp_alltoall_z22v 1201 16.6 0.621 0.696 0.621 0.696 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.041000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=2.234219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 545.738752E+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 63496. MP_Allreduce 10074 349. 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 0.056 0.083 35.516 35.517 qs_mol_dyn_low 1 2.0 0.003 0.004 35.175 35.192 qs_forces 11 3.9 0.002 0.002 35.076 35.077 qs_energies 11 4.9 0.001 0.005 33.349 33.355 scf_env_do_scf 11 5.9 0.000 0.001 27.699 27.700 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.339 24.340 dbcsr_multiply_generic 2286 12.5 0.099 0.102 18.562 18.671 velocity_verlet 10 3.0 0.002 0.002 17.569 17.571 qs_scf_new_mos 108 7.5 0.001 0.001 15.978 16.024 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.978 16.023 multiply_cannon 2286 13.5 0.229 0.262 15.041 15.551 ot_scf_mini 108 9.5 0.002 0.004 15.048 15.096 multiply_cannon_loop 2286 14.5 0.937 0.967 14.126 14.554 multiply_cannon_multrec 27432 15.5 2.339 3.026 8.862 9.293 ot_mini 108 10.5 0.001 0.001 9.090 9.151 dbcsr_mm_accdrv_process 47916 15.9 5.470 7.248 6.428 7.643 qs_ot_get_derivative 108 11.5 0.008 0.026 7.285 7.333 rebuild_ks_matrix 119 8.3 0.000 0.000 6.534 6.585 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.534 6.585 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.800 5.846 init_scf_run 11 5.9 0.000 0.001 4.306 4.306 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.306 4.306 sum_up_and_integrate 119 10.3 0.035 0.038 3.769 3.778 integrate_v_rspace 119 11.3 0.003 0.003 3.733 3.744 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.482 3.512 calculate_rho_elec 119 8.7 0.040 0.046 3.481 3.511 qs_ot_get_p 119 10.4 0.001 0.002 3.360 3.432 init_scf_loop 11 6.9 0.002 0.015 3.341 3.342 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.640 3.102 calculate_first_density_matrix 1 7.0 0.000 0.002 2.957 2.959 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.098 2.554 apply_single 119 13.6 0.000 0.000 2.098 2.554 prepare_preconditioner 11 7.9 0.000 0.000 2.498 2.505 make_preconditioner 11 8.9 0.000 0.002 2.498 2.505 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.449 2.451 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.112 2.437 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.290 2.303 make_m2s 4572 13.5 0.055 0.056 2.108 2.227 calculate_dm_sparse 119 9.5 0.000 0.000 2.097 2.148 mp_waitall_1 137007 16.6 1.585 2.145 1.585 2.145 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.132 2.141 rs_pw_transfer 974 11.9 0.009 0.010 1.936 2.135 make_images 4572 14.5 0.270 0.331 2.001 2.117 density_rs2pw 119 9.7 0.004 0.004 1.906 2.102 jit_kernel_multiply 10 15.8 0.895 2.034 0.895 2.034 acc_transpose_blocks 27432 15.5 0.113 0.117 1.980 2.025 pw_transfer 1439 11.6 0.065 0.070 1.986 2.022 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.894 1.934 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.886 1.912 grid_integrate_task_list 119 12.3 1.826 1.911 1.826 1.911 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.822 1.823 ot_diis_step 108 11.5 0.012 0.012 1.765 1.765 fft_wrap_pw1pw2_140 487 13.2 0.290 0.306 1.567 1.607 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.591 1.600 cp_fm_diag_elpa_base 50 14.0 1.557 1.574 1.589 1.598 mp_sum_l 7207 12.9 0.984 1.549 0.984 1.549 potential_pw2rs 119 12.3 0.008 0.010 1.519 1.521 acc_transpose_blocks_kernels 27432 16.5 0.269 0.275 1.351 1.383 fft3d_ps 1201 14.6 0.557 0.605 1.329 1.355 grid_collocate_task_list 119 9.7 1.220 1.336 1.220 1.336 wfi_extrapolate 11 7.9 0.001 0.001 1.299 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.285 1.293 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.666 1.166 cp_fm_upper_to_full 72 14.2 0.815 1.159 0.815 1.159 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.088 1.110 jit_kernel_transpose 5 15.6 1.082 1.109 1.082 1.109 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.097 1.098 multiply_cannon_sync_h2d 27432 15.5 0.999 1.061 0.999 1.061 dbcsr_complete_redistribute 329 12.2 0.120 0.145 0.733 1.004 make_images_data 4572 15.5 0.045 0.048 0.818 0.942 hybrid_alltoall_any 4725 16.4 0.063 0.152 0.693 0.886 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.840 mp_alltoall_d11v 2130 13.8 0.692 0.827 0.692 0.827 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.544 0.810 cp_fm_cholesky_invert 11 10.9 0.723 0.726 0.723 0.726 mp_alltoall_i22 627 13.8 0.418 0.711 0.418 0.711 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.517000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.818182, yerr=3.485507 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 598.777856E+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 63495. MP_Allreduce 10074 348. 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 0.014 0.031 28.595 28.596 qs_mol_dyn_low 1 2.0 0.003 0.004 28.243 28.250 qs_forces 11 3.9 0.002 0.002 28.102 28.104 qs_energies 11 4.9 0.001 0.001 26.364 26.366 scf_env_do_scf 11 5.9 0.000 0.001 21.282 21.282 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.796 18.797 velocity_verlet 10 3.0 0.077 0.090 14.519 14.522 dbcsr_multiply_generic 2286 12.5 0.104 0.116 12.257 12.332 qs_scf_new_mos 108 7.5 0.001 0.001 10.821 10.846 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.821 10.846 multiply_cannon 2286 13.5 0.232 0.237 9.779 10.216 ot_scf_mini 108 9.5 0.002 0.002 10.151 10.177 multiply_cannon_loop 2286 14.5 0.329 0.339 8.889 9.053 multiply_cannon_multrec 9144 15.5 1.662 1.902 6.071 6.329 rebuild_ks_matrix 119 8.3 0.000 0.000 5.997 6.018 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.997 6.017 ot_mini 108 10.5 0.001 0.001 5.665 5.696 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.318 5.337 qs_ot_get_derivative 108 11.5 0.001 0.002 4.370 4.396 dbcsr_mm_accdrv_process 12550 15.8 3.195 4.216 4.309 4.371 sum_up_and_integrate 119 10.3 0.038 0.041 3.673 3.677 integrate_v_rspace 119 11.3 0.003 0.003 3.636 3.641 init_scf_run 11 5.9 0.000 0.001 3.547 3.547 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.546 3.547 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.523 3.531 calculate_rho_elec 119 8.7 0.060 0.061 3.522 3.530 qs_ot_get_p 119 10.4 0.001 0.001 2.733 2.773 init_scf_loop 11 6.9 0.000 0.000 2.467 2.468 calculate_first_density_matrix 1 7.0 0.001 0.001 2.418 2.419 pw_transfer 1439 11.6 0.066 0.069 2.060 2.069 density_rs2pw 119 9.7 0.004 0.004 1.852 1.994 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.966 1.977 grid_integrate_task_list 119 12.3 1.852 1.934 1.852 1.934 make_m2s 4572 13.5 0.035 0.036 1.710 1.862 calculate_dm_sparse 119 9.5 0.000 0.000 1.819 1.838 rs_pw_transfer 974 11.9 0.025 0.025 1.701 1.835 mp_waitall_1 115863 16.7 1.325 1.833 1.325 1.833 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.802 1.805 make_images 4572 14.5 0.267 0.304 1.620 1.770 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.748 1.749 prepare_preconditioner 11 7.9 0.000 0.000 1.724 1.729 make_preconditioner 11 8.9 0.000 0.000 1.724 1.729 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.614 1.640 fft_wrap_pw1pw2_140 487 13.2 0.367 0.376 1.620 1.633 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.605 1.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.546 1.561 jit_kernel_multiply 8 15.9 1.076 1.512 1.076 1.512 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.503 1.509 potential_pw2rs 119 12.3 0.010 0.011 1.398 1.401 grid_collocate_task_list 119 9.7 1.271 1.392 1.271 1.392 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.338 1.346 cp_fm_diag_elpa_base 50 14.0 1.311 1.328 1.336 1.344 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.320 1.321 fft3d_ps 1201 14.6 0.563 0.573 1.295 1.304 ot_diis_step 108 11.5 0.013 0.013 1.282 1.283 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.241 1.246 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.158 1.179 apply_single 119 13.6 0.000 0.000 1.158 1.179 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.128 1.142 wfi_extrapolate 11 7.9 0.001 0.001 1.080 1.080 make_images_data 4572 15.5 0.039 0.042 0.789 0.962 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.754 0.961 acc_transpose_blocks 9144 15.5 0.038 0.039 0.947 0.957 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.919 cp_fm_cholesky_invert 11 10.9 0.880 0.882 0.880 0.882 mp_alltoall_d11v 2130 13.8 0.744 0.834 0.744 0.834 qs_env_update_s_mstruct 11 6.9 0.088 0.104 0.761 0.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.800 0.802 multiply_cannon_sync_h2d 9144 15.5 0.713 0.796 0.713 0.796 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.716 0.725 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.707 0.710 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.315 0.690 mp_alltoall_z22v 1201 16.6 0.606 0.638 0.606 0.638 dbcsr_complete_redistribute 329 12.2 0.211 0.255 0.602 0.627 mp_allgather_i34 2286 14.5 0.221 0.624 0.221 0.624 jit_kernel_transpose 5 15.6 0.590 0.592 0.590 0.592 mp_waitany 5200 13.7 0.450 0.573 0.450 0.573 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.596000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.363636, yerr=3.548949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 757.731328E+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 63729. MP_Allreduce 10074 433. 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 0.016 0.036 40.780 40.781 qs_mol_dyn_low 1 2.0 0.003 0.004 40.517 40.524 qs_forces 11 3.9 0.002 0.003 40.455 40.456 qs_energies 11 4.9 0.001 0.001 38.443 38.446 scf_env_do_scf 11 5.9 0.001 0.001 32.790 32.790 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 25.081 25.082 velocity_verlet 10 3.0 0.002 0.002 22.965 22.970 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.319 17.539 qs_scf_new_mos 108 7.5 0.001 0.001 15.485 15.580 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.484 15.579 ot_scf_mini 108 9.5 0.002 0.002 14.413 14.515 multiply_cannon 2286 13.5 0.299 0.303 13.606 14.509 multiply_cannon_loop 2286 14.5 0.342 0.348 12.360 13.302 ot_mini 108 10.5 0.001 0.001 8.681 8.796 multiply_cannon_multrec 9144 15.5 3.448 4.829 8.624 8.740 init_scf_loop 11 6.9 0.000 0.000 7.683 7.685 rebuild_ks_matrix 119 8.3 0.000 0.001 7.222 7.362 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 7.221 7.361 qs_ot_get_derivative 108 11.5 0.001 0.002 6.671 6.774 prepare_preconditioner 11 7.9 0.000 0.000 6.730 6.742 make_preconditioner 11 8.9 0.000 0.000 6.730 6.742 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.293 6.622 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.488 6.614 dbcsr_mm_accdrv_process 12550 15.8 4.017 5.626 5.053 6.474 cp_fm_upper_to_full 72 14.2 3.179 4.561 3.179 4.561 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.159 4.164 calculate_rho_elec 119 8.7 0.118 0.121 4.158 4.163 sum_up_and_integrate 119 10.3 0.064 0.066 3.925 3.932 integrate_v_rspace 119 11.3 0.003 0.003 3.861 3.868 init_scf_run 11 5.9 0.000 0.001 3.620 3.620 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.620 3.620 qs_ot_get_p 119 10.4 0.001 0.001 3.031 3.164 mp_waitall_1 94719 16.7 2.229 3.161 2.229 3.161 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.414 2.834 dbcsr_complete_redistribute 329 12.2 0.286 0.293 1.945 2.759 pw_transfer 1439 11.6 0.069 0.069 2.599 2.604 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.500 2.506 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.185 2.474 apply_single 119 13.6 0.000 0.000 2.184 2.474 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.651 2.458 make_m2s 4572 13.5 0.038 0.039 2.175 2.351 mp_alltoall_i22 627 13.8 1.409 2.247 1.409 2.247 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.431 2.230 make_images 4572 14.5 0.350 0.380 2.054 2.230 calculate_first_density_matrix 1 7.0 0.000 0.001 2.205 2.207 calculate_dm_sparse 119 9.5 0.000 0.000 2.132 2.152 fft_wrap_pw1pw2_140 487 13.2 0.618 0.622 2.120 2.126 density_rs2pw 119 9.7 0.004 0.004 2.104 2.124 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.263 2.114 grid_integrate_task_list 119 12.3 2.041 2.051 2.041 2.051 ot_diis_step 108 11.5 0.014 0.014 1.987 1.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.922 1.923 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.830 1.832 qs_energies_init_hamiltonians 11 5.9 0.025 0.026 1.772 1.773 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.681 1.714 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.628 1.681 mp_sum_l 7207 12.9 0.993 1.673 0.993 1.673 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.563 1.563 fft3d_ps 1201 14.6 0.593 0.603 1.515 1.519 jit_kernel_multiply 8 15.4 1.008 1.517 1.008 1.517 grid_collocate_task_list 119 9.7 1.476 1.499 1.476 1.499 rs_pw_transfer 974 11.9 0.030 0.050 1.444 1.476 cp_fm_cholesky_invert 11 10.9 1.376 1.379 1.376 1.379 potential_pw2rs 119 12.3 0.014 0.015 1.347 1.349 wfi_extrapolate 11 7.9 0.001 0.001 1.348 1.348 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.069 1.323 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.290 1.290 cp_fm_diag_elpa_base 50 14.0 1.145 1.198 1.288 1.289 make_images_data 4572 15.5 0.043 0.046 1.017 1.222 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.209 1.216 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.114 1.133 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.083 1.099 multiply_cannon_sync_h2d 9144 15.5 1.042 1.049 1.042 1.049 mp_alltoall_d11v 2130 13.8 1.023 1.048 1.023 1.048 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.983 1.015 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.944 0.957 qs_create_task_list 11 7.9 0.001 0.001 0.933 0.945 generate_qs_task_list 11 8.9 0.367 0.386 0.932 0.944 acc_transpose_blocks 9144 15.5 0.038 0.039 0.926 0.929 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.781000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.363636, yerr=16.636612 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 505.200640E+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 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 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 0.012 0.031 81.323 81.324 qs_mol_dyn_low 1 2.0 0.003 0.003 81.075 81.084 qs_forces 11 3.9 0.003 0.003 80.658 80.659 qs_energies 11 4.9 0.001 0.001 77.811 77.822 scf_env_do_scf 11 5.9 0.000 0.001 69.066 69.069 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 63.572 63.572 dbcsr_multiply_generic 2055 12.4 0.105 0.108 50.112 50.338 qs_scf_new_mos 99 7.5 0.000 0.001 46.256 46.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.255 46.393 ot_scf_mini 99 9.5 0.002 0.002 43.869 43.968 multiply_cannon 2055 13.4 0.186 0.191 41.950 42.722 velocity_verlet 10 3.0 0.001 0.002 42.477 42.481 multiply_cannon_loop 2055 14.4 1.504 1.534 41.003 41.798 ot_mini 99 10.5 0.001 0.001 25.718 25.820 qs_ot_get_derivative 99 11.5 0.001 0.002 18.960 19.044 multiply_cannon_multrec 49320 15.4 12.503 13.231 17.357 18.002 rebuild_ks_matrix 110 8.3 0.000 0.000 14.550 14.680 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.550 14.679 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.757 12.867 mp_waitall_1 220248 16.4 10.393 11.230 10.393 11.230 multiply_cannon_sync_h2d 49320 15.4 10.394 10.978 10.394 10.978 qs_ot_get_p 110 10.4 0.001 0.001 9.748 9.862 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.206 7.798 apply_single 110 13.6 0.000 0.000 7.206 7.798 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.060 7.587 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.294 7.482 sum_up_and_integrate 110 10.3 0.036 0.042 7.090 7.105 integrate_v_rspace 110 11.3 0.003 0.004 7.054 7.077 init_scf_run 11 5.9 0.000 0.001 6.741 6.742 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.741 6.741 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.649 6.697 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.542 6.688 calculate_rho_elec 110 8.6 0.021 0.026 6.542 6.687 ot_diis_step 99 11.5 0.006 0.006 6.584 6.584 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.755 5.755 init_scf_loop 11 6.9 0.000 0.000 5.467 5.467 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.370 5.437 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.184 5.214 cp_fm_diag_elpa_base 48 14.0 5.169 5.197 5.183 5.212 dbcsr_mm_accdrv_process 87628 16.1 1.925 1.987 4.729 4.987 mp_sum_l 6514 12.8 3.906 4.698 3.906 4.698 rs_pw_transfer 902 11.9 0.012 0.014 3.690 4.350 wfi_extrapolate 11 7.9 0.001 0.001 4.023 4.024 density_rs2pw 110 9.6 0.004 0.005 3.365 3.955 make_m2s 4110 13.4 0.061 0.066 3.794 3.910 calculate_dm_sparse 110 9.5 0.000 0.001 3.750 3.848 make_images 4110 14.4 0.177 0.191 3.699 3.816 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.291 3.629 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.566 3.570 grid_integrate_task_list 110 12.3 3.226 3.396 3.226 3.396 prepare_preconditioner 11 7.9 0.000 0.000 3.300 3.329 make_preconditioner 11 8.9 0.000 0.000 3.300 3.329 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.216 3.289 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.144 3.222 pw_transfer 1331 11.6 0.055 0.067 3.106 3.188 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.091 3.138 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.017 3.096 fft_wrap_pw1pw2_140 451 13.1 0.453 0.497 2.574 2.663 potential_pw2rs 110 12.3 0.006 0.007 2.636 2.661 calculate_first_density_matrix 1 7.0 0.001 0.001 2.629 2.635 jit_kernel_multiply 13 15.9 2.529 2.548 2.529 2.548 mp_alltoall_d11v 2046 13.8 2.021 2.448 2.021 2.448 grid_collocate_task_list 110 9.6 2.085 2.340 2.085 2.340 mp_waitany 14300 13.8 1.812 2.310 1.812 2.310 fft3d_ps 1111 14.6 0.796 0.884 2.210 2.264 acc_transpose_blocks 49320 15.4 0.209 0.220 2.067 2.136 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.922 1.940 make_images_data 4110 15.4 0.042 0.044 1.719 1.852 mp_sum_d 3889 11.9 1.299 1.833 1.299 1.833 cp_fm_cholesky_invert 11 10.9 1.817 1.821 1.817 1.821 hybrid_alltoall_any 4261 16.3 0.083 0.485 1.487 1.764 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.634 1.661 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.324000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=479.272727, yerr=2.766386 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 587.542528E+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 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 3225140. 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 0.049 0.081 68.254 68.254 qs_mol_dyn_low 1 2.0 0.003 0.004 67.951 67.968 qs_forces 11 3.9 0.003 0.003 67.880 67.881 qs_energies 11 4.9 0.001 0.006 64.568 64.571 scf_env_do_scf 11 5.9 0.001 0.002 55.967 55.970 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 48.553 48.553 dbcsr_multiply_generic 2055 12.4 0.116 0.123 37.295 37.437 velocity_verlet 10 3.0 0.001 0.002 35.669 35.670 qs_scf_new_mos 99 7.5 0.001 0.001 32.371 32.500 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.371 32.499 multiply_cannon 2055 13.4 0.220 0.242 30.919 32.001 ot_scf_mini 99 9.5 0.003 0.004 30.727 30.846 multiply_cannon_loop 2055 14.4 0.922 0.943 29.696 30.627 ot_mini 99 10.5 0.001 0.001 17.920 18.042 multiply_cannon_multrec 24660 15.4 7.646 9.358 13.899 15.510 rebuild_ks_matrix 110 8.3 0.000 0.000 13.679 13.784 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.678 13.783 qs_ot_get_derivative 99 11.5 0.001 0.002 12.156 12.278 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.036 12.132 mp_waitall_1 176588 16.5 7.473 9.829 7.473 9.829 multiply_cannon_sync_h2d 24660 15.4 7.024 8.017 7.024 8.017 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.094 7.655 init_scf_loop 11 6.9 0.001 0.004 7.380 7.381 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.479 7.221 apply_single 110 13.6 0.000 0.001 6.479 7.221 sum_up_and_integrate 110 10.3 0.051 0.059 6.581 6.592 integrate_v_rspace 110 11.3 0.002 0.003 6.530 6.544 dbcsr_mm_accdrv_process 52282 16.1 4.675 5.480 6.092 6.398 qs_ot_get_p 110 10.4 0.001 0.001 6.218 6.351 init_scf_run 11 5.9 0.000 0.001 6.208 6.208 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.207 6.208 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.112 6.123 calculate_rho_elec 110 8.6 0.040 0.048 6.111 6.122 ot_diis_step 99 11.5 0.010 0.010 5.717 5.717 prepare_preconditioner 11 7.9 0.000 0.001 5.376 5.397 make_preconditioner 11 8.9 0.001 0.005 5.376 5.397 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.581 5.323 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.981 5.145 make_m2s 4110 13.4 0.057 0.061 4.036 4.474 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.406 4.425 make_images 4110 14.4 0.397 0.441 3.927 4.361 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.985 3.986 density_rs2pw 110 9.6 0.005 0.008 3.238 3.797 pw_transfer 1331 11.6 0.066 0.075 3.604 3.736 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.497 3.632 rs_pw_transfer 902 11.9 0.013 0.015 2.928 3.538 wfi_extrapolate 11 7.9 0.001 0.001 3.470 3.470 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.434 3.451 cp_fm_diag_elpa_base 48 14.0 3.389 3.408 3.432 3.448 grid_integrate_task_list 110 12.3 3.149 3.315 3.149 3.315 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.271 3.273 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.105 3.177 fft_wrap_pw1pw2_140 451 13.1 0.522 0.539 3.011 3.147 calculate_dm_sparse 110 9.5 0.001 0.001 3.023 3.050 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.903 2.962 make_images_data 4110 15.4 0.046 0.051 2.279 2.708 hybrid_alltoall_any 4261 16.3 0.103 0.449 1.982 2.652 calculate_first_density_matrix 1 7.0 0.001 0.002 2.637 2.639 fft3d_ps 1111 14.6 1.110 1.334 2.467 2.622 cp_fm_cholesky_invert 11 10.9 2.481 2.488 2.481 2.488 grid_collocate_task_list 110 9.6 2.056 2.472 2.056 2.472 potential_pw2rs 110 12.3 0.008 0.009 2.379 2.395 mp_sum_l 6514 12.8 1.733 2.329 1.733 2.329 jit_kernel_multiply 11 16.4 1.065 2.097 1.065 2.097 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.927 1.946 mp_alltoall_d11v 2046 13.8 1.675 1.938 1.675 1.938 mp_waitany 10164 13.8 1.248 1.906 1.248 1.906 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.891 1.892 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.808 multiply_cannon_metrocomm4 22605 15.4 0.074 0.080 0.782 1.766 mp_irecv_dv 57340 16.2 0.657 1.644 0.657 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.608 1.620 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.220 0.991 1.588 acc_transpose_blocks 24660 15.4 0.110 0.113 1.532 1.560 mp_allgather_i34 2055 14.4 0.553 1.509 0.553 1.509 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.471 dbcsr_complete_redistribute 325 12.2 0.251 0.329 1.105 1.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.254000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.636364, yerr=6.540402 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 660.279296E+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 66428. MP_Allreduce 9694 567. 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 0.032 0.057 60.437 60.446 qs_mol_dyn_low 1 2.0 0.003 0.004 60.137 60.147 qs_forces 11 3.9 0.003 0.003 60.067 60.068 qs_energies 11 4.9 0.001 0.001 56.807 56.810 scf_env_do_scf 11 5.9 0.000 0.001 48.869 48.869 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.179 40.179 velocity_verlet 10 3.0 0.002 0.002 33.194 33.196 dbcsr_multiply_generic 2055 12.4 0.107 0.112 28.439 28.711 qs_scf_new_mos 99 7.5 0.001 0.001 25.001 25.104 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.001 25.103 ot_scf_mini 99 9.5 0.002 0.003 23.770 23.894 multiply_cannon 2055 13.4 0.211 0.220 22.136 23.330 multiply_cannon_loop 2055 14.4 0.618 0.631 20.981 22.119 ot_mini 99 10.5 0.001 0.001 13.590 13.711 rebuild_ks_matrix 110 8.3 0.000 0.000 12.451 12.594 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.451 12.594 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.954 11.086 multiply_cannon_multrec 16440 15.4 3.818 4.784 9.660 10.575 mp_waitall_1 139946 16.5 7.113 10.395 7.113 10.395 qs_ot_get_derivative 99 11.5 0.001 0.002 9.111 9.231 init_scf_loop 11 6.9 0.000 0.000 8.650 8.651 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.442 7.544 prepare_preconditioner 11 7.9 0.000 0.000 6.877 6.899 make_preconditioner 11 8.9 0.000 0.000 6.877 6.899 sum_up_and_integrate 110 10.3 0.059 0.060 6.623 6.637 integrate_v_rspace 110 11.3 0.003 0.003 6.564 6.578 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.195 6.545 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.068 6.077 calculate_rho_elec 110 8.6 0.059 0.060 6.067 6.076 dbcsr_mm_accdrv_process 34862 16.1 4.826 5.339 5.697 5.844 qs_ot_get_p 110 10.4 0.001 0.001 5.377 5.539 init_scf_run 11 5.9 0.000 0.001 5.536 5.536 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.536 5.536 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.987 5.442 apply_single 110 13.6 0.000 0.000 4.987 5.442 make_m2s 4110 13.4 0.050 0.052 4.185 4.545 ot_diis_step 99 11.5 0.011 0.011 4.448 4.448 make_images 4110 14.4 0.390 0.511 4.069 4.428 density_rs2pw 110 9.6 0.004 0.005 3.177 4.424 multiply_cannon_sync_h2d 16440 15.4 3.655 4.322 3.655 4.322 rs_pw_transfer 902 11.9 0.011 0.012 2.772 4.000 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.155 3.803 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.767 3.770 pw_transfer 1331 11.6 0.065 0.074 3.706 3.717 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.599 3.612 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.429 3.430 grid_integrate_task_list 110 12.3 3.157 3.396 3.157 3.396 fft_wrap_pw1pw2_140 451 13.1 0.638 0.644 3.093 3.110 wfi_extrapolate 11 7.9 0.001 0.001 3.001 3.001 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.894 2.905 cp_fm_diag_elpa_base 48 14.0 2.830 2.858 2.892 2.904 make_images_data 4110 15.4 0.043 0.047 2.420 2.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.846 2.847 hybrid_alltoall_any 4261 16.3 0.106 0.377 2.110 2.765 cp_fm_cholesky_invert 11 10.9 2.749 2.755 2.749 2.755 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.568 calculate_dm_sparse 110 9.5 0.001 0.001 2.508 2.534 grid_collocate_task_list 110 9.6 2.079 2.477 2.079 2.477 mp_waitany 17072 13.8 1.185 2.454 1.185 2.454 calculate_first_density_matrix 1 7.0 0.001 0.001 2.447 2.449 fft3d_ps 1111 14.6 1.094 1.107 2.421 2.438 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.909 2.432 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.308 2.371 potential_pw2rs 110 12.3 0.011 0.011 2.354 2.366 mp_irecv_dv 48980 15.7 0.838 2.307 0.838 2.307 mp_alltoall_d11v 2046 13.8 1.800 2.262 1.800 2.262 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.180 0.895 2.121 mp_sum_l 6514 12.8 1.497 2.098 1.497 2.098 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.954 1.955 dbcsr_complete_redistribute 325 12.2 0.319 0.358 1.501 1.945 cp_fm_upper_to_full 70 14.2 1.385 1.769 1.385 1.769 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.706 cp_fm_cholesky_decompose 22 10.9 1.599 1.618 1.599 1.618 mp_allgather_i34 2055 14.4 0.480 1.571 0.480 1.571 jit_kernel_multiply 8 16.9 0.482 1.547 0.482 1.547 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.479 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.465 1.477 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.985 1.420 rs_gather_matrices 110 12.3 0.233 0.260 0.973 1.378 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.272 1.280 acc_transpose_blocks 16440 15.4 0.072 0.074 1.219 1.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.446000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=626.000000, yerr=7.567875 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 729.051136E+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 66426. MP_Allreduce 9694 608. 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 0.016 0.035 65.020 65.021 qs_mol_dyn_low 1 2.0 0.003 0.003 64.796 64.806 qs_forces 11 3.9 0.003 0.003 64.731 64.731 qs_energies 11 4.9 0.001 0.001 61.292 61.295 scf_env_do_scf 11 5.9 0.000 0.001 53.095 53.098 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 41.274 41.274 velocity_verlet 10 3.0 0.001 0.002 37.267 37.269 dbcsr_multiply_generic 2055 12.4 0.114 0.120 29.204 29.388 qs_scf_new_mos 99 7.5 0.001 0.001 26.199 26.304 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.198 26.303 ot_scf_mini 99 9.5 0.003 0.003 24.579 24.681 multiply_cannon 2055 13.4 0.244 0.271 22.343 23.362 multiply_cannon_loop 2055 14.4 0.884 0.908 20.946 21.423 ot_mini 99 10.5 0.001 0.001 13.924 14.048 multiply_cannon_multrec 24660 15.4 4.215 6.768 12.647 13.811 rebuild_ks_matrix 110 8.3 0.000 0.000 12.226 12.323 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.225 12.323 init_scf_loop 11 6.9 0.000 0.000 11.774 11.775 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.778 10.863 prepare_preconditioner 11 7.9 0.000 0.000 10.036 10.053 make_preconditioner 11 8.9 0.000 0.000 10.036 10.053 qs_ot_get_derivative 99 11.5 0.001 0.002 9.786 9.896 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.231 9.697 dbcsr_mm_accdrv_process 52304 16.0 6.847 8.343 8.287 9.216 sum_up_and_integrate 110 10.3 0.067 0.070 6.554 6.568 integrate_v_rspace 110 11.3 0.003 0.003 6.486 6.500 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.134 6.144 calculate_rho_elec 110 8.6 0.078 0.081 6.134 6.143 mp_waitall_1 121746 16.5 4.200 6.055 4.200 6.055 qs_ot_get_p 110 10.4 0.001 0.001 5.671 5.817 make_m2s 4110 13.4 0.060 0.064 5.295 5.639 make_images 4110 14.4 0.573 0.691 5.155 5.495 init_scf_run 11 5.9 0.000 0.001 5.467 5.467 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.467 5.467 cp_fm_upper_to_full 70 14.2 3.364 4.879 3.364 4.879 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.027 4.110 apply_single 110 13.6 0.000 0.000 4.027 4.110 ot_diis_step 99 11.5 0.011 0.012 4.104 4.104 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.911 3.926 dbcsr_complete_redistribute 325 12.2 0.412 0.456 2.714 3.871 pw_transfer 1331 11.6 0.065 0.074 3.769 3.801 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.663 3.699 density_rs2pw 110 9.6 0.004 0.004 3.064 3.579 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.439 3.440 grid_integrate_task_list 110 12.3 3.236 3.384 3.236 3.384 multiply_cannon_sync_h2d 24660 15.4 3.163 3.329 3.163 3.329 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.275 3.328 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.147 3.292 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.449 3.262 fft_wrap_pw1pw2_140 451 13.1 0.669 0.694 3.134 3.174 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.955 3.014 hybrid_alltoall_any 4261 16.3 0.121 0.460 2.305 3.008 wfi_extrapolate 11 7.9 0.001 0.001 3.004 3.004 make_images_data 4110 15.4 0.046 0.050 2.644 2.999 rs_pw_transfer 902 11.9 0.011 0.013 2.434 2.953 calculate_dm_sparse 110 9.5 0.001 0.001 2.899 2.929 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.796 2.927 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.901 2.911 cp_fm_diag_elpa_base 48 14.0 2.755 2.811 2.899 2.910 mp_alltoall_i22 605 13.7 1.655 2.832 1.655 2.832 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.797 2.799 cp_fm_cholesky_invert 11 10.9 2.750 2.758 2.750 2.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.436 2.468 fft3d_ps 1111 14.6 1.086 1.116 2.423 2.442 grid_collocate_task_list 110 9.6 2.180 2.430 2.180 2.430 calculate_first_density_matrix 1 7.0 0.000 0.001 2.372 2.374 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.238 2.239 potential_pw2rs 110 12.3 0.013 0.013 2.215 2.221 mp_alltoall_d11v 2046 13.8 1.809 2.149 1.809 2.149 jit_kernel_multiply 10 15.6 1.109 1.851 1.109 1.851 cp_fm_cholesky_decompose 22 10.9 1.677 1.740 1.677 1.740 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.699 1.731 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.689 1.703 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.696 mp_waitany 13376 13.8 1.055 1.531 1.055 1.531 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.515 1.531 acc_transpose_blocks 24660 15.4 0.107 0.110 1.512 1.530 multiply_cannon_metrocomm4 20550 15.4 0.062 0.064 0.844 1.518 mp_allgather_i34 2055 14.4 0.497 1.492 0.497 1.492 mp_sum_l 6514 12.8 0.876 1.465 0.876 1.465 mp_irecv_dv 62702 16.1 0.741 1.435 0.741 1.435 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.174 0.800 1.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.021000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.272727, yerr=9.362312 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 829.677568E+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 66424. MP_Allreduce 9694 649. 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 0.032 0.051 54.573 54.577 qs_mol_dyn_low 1 2.0 0.003 0.003 54.329 54.338 qs_forces 11 3.9 0.003 0.003 54.264 54.265 qs_energies 11 4.9 0.001 0.001 50.568 50.573 scf_env_do_scf 11 5.9 0.000 0.001 42.373 42.373 scf_env_do_scf_inner_loop 99 6.5 0.004 0.011 34.780 34.781 velocity_verlet 10 3.0 0.002 0.002 30.819 30.822 dbcsr_multiply_generic 2055 12.4 0.104 0.108 22.655 22.806 qs_scf_new_mos 99 7.5 0.001 0.001 19.823 19.880 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.823 19.879 multiply_cannon 2055 13.4 0.245 0.257 17.365 18.687 ot_scf_mini 99 9.5 0.002 0.002 18.596 18.648 multiply_cannon_loop 2055 14.4 0.320 0.333 16.050 16.310 rebuild_ks_matrix 110 8.3 0.000 0.000 11.870 11.914 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 11.870 11.913 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.534 10.574 ot_mini 99 10.5 0.001 0.001 10.259 10.306 multiply_cannon_multrec 8220 15.4 3.235 4.455 7.542 8.612 mp_waitall_1 103326 16.6 5.930 7.627 5.930 7.627 init_scf_loop 11 6.9 0.000 0.000 7.549 7.549 qs_ot_get_derivative 99 11.5 0.001 0.002 6.445 6.495 sum_up_and_integrate 110 10.3 0.079 0.080 6.467 6.479 integrate_v_rspace 110 11.3 0.003 0.003 6.389 6.403 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.221 6.232 calculate_rho_elec 110 8.6 0.114 0.115 6.221 6.232 prepare_preconditioner 11 7.9 0.000 0.000 5.898 5.903 make_preconditioner 11 8.9 0.000 0.000 5.898 5.903 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.486 5.553 dbcsr_mm_accdrv_process 17442 15.9 2.791 3.751 4.176 5.146 init_scf_run 11 5.9 0.000 0.001 5.088 5.088 scf_env_initial_rho_setup 11 6.9 0.002 0.003 5.088 5.088 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.156 4.528 qs_ot_get_p 110 10.4 0.001 0.001 4.430 4.478 make_m2s 4110 13.4 0.039 0.041 4.074 4.328 make_images 4110 14.4 0.633 0.683 3.943 4.201 pw_transfer 1331 11.6 0.066 0.072 3.905 3.934 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.797 3.830 ot_diis_step 99 11.5 0.012 0.012 3.793 3.793 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.719 3.761 apply_single 110 13.6 0.000 0.000 3.719 3.760 grid_integrate_task_list 110 12.3 3.359 3.533 3.359 3.533 density_rs2pw 110 9.6 0.004 0.005 2.977 3.349 fft_wrap_pw1pw2_140 451 13.1 0.835 0.846 3.296 3.340 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.142 3.145 multiply_cannon_sync_h2d 8220 15.4 2.910 3.023 2.910 3.023 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.847 2.848 cp_fm_cholesky_invert 11 10.9 2.795 2.798 2.795 2.798 make_images_data 4110 15.4 0.038 0.043 2.328 2.742 wfi_extrapolate 11 7.9 0.001 0.001 2.702 2.702 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.666 2.667 hybrid_alltoall_any 4261 16.3 0.200 0.862 2.289 2.665 calculate_dm_sparse 110 9.5 0.001 0.001 2.488 2.538 grid_collocate_task_list 110 9.6 2.278 2.537 2.278 2.537 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.519 2.521 rs_pw_transfer 902 11.9 0.011 0.011 2.111 2.516 fft3d_ps 1111 14.6 1.134 1.196 2.341 2.359 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.349 2.357 cp_fm_diag_elpa_base 48 14.0 2.295 2.316 2.348 2.355 calculate_first_density_matrix 1 7.0 0.000 0.000 2.291 2.293 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.186 2.202 potential_pw2rs 110 12.3 0.016 0.016 2.021 2.024 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 2.004 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.849 1.871 mp_alltoall_d11v 2046 13.8 1.591 1.868 1.591 1.868 jit_kernel_multiply 10 15.7 1.077 1.813 1.077 1.813 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.730 1.757 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.663 1.668 cp_fm_cholesky_decompose 22 10.9 1.638 1.651 1.638 1.651 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.504 1.624 mp_allgather_i34 2055 14.4 0.501 1.571 0.501 1.571 dbcsr_complete_redistribute 325 12.2 0.563 0.578 1.408 1.501 mp_waitany 9240 13.8 1.043 1.484 1.043 1.484 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.473 qs_create_task_list 11 7.9 0.000 0.001 1.219 1.322 generate_qs_task_list 11 8.9 0.374 0.443 1.218 1.322 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.267 mp_irecv_dv 24056 15.7 0.457 1.222 0.457 1.222 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.165 0.793 1.201 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.134 1.167 rs_gather_matrices 110 12.3 0.326 0.366 0.920 1.156 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.577000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=779.818182, yerr=11.945468 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 1.334641E+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 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.037 86.064 86.065 qs_mol_dyn_low 1 2.0 0.003 0.003 85.764 85.775 qs_forces 11 3.9 0.003 0.003 85.695 85.696 qs_energies 11 4.9 0.001 0.001 81.541 81.542 scf_env_do_scf 11 5.9 0.001 0.001 71.620 71.620 velocity_verlet 10 3.0 0.002 0.002 55.170 55.176 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 43.620 43.622 dbcsr_multiply_generic 2055 12.4 0.118 0.123 28.273 28.409 init_scf_loop 11 6.9 0.000 0.000 27.931 27.932 prepare_preconditioner 11 7.9 0.000 0.000 25.919 25.930 make_preconditioner 11 8.9 0.000 0.000 25.919 25.930 qs_scf_new_mos 99 7.5 0.001 0.001 25.582 25.664 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.581 25.663 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.235 25.390 ot_scf_mini 99 9.5 0.002 0.002 23.842 23.916 multiply_cannon 2055 13.4 0.345 0.373 21.396 22.159 multiply_cannon_loop 2055 14.4 0.339 0.342 19.622 19.973 cp_fm_upper_to_full 70 14.2 12.675 18.072 12.675 18.072 rebuild_ks_matrix 110 8.3 0.001 0.001 13.974 14.055 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.973 14.055 ot_mini 99 10.5 0.001 0.001 13.373 13.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.653 12.726 dbcsr_complete_redistribute 325 12.2 1.023 1.042 7.217 10.361 multiply_cannon_multrec 8220 15.4 4.374 4.571 9.629 9.719 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.207 9.358 qs_ot_get_derivative 99 11.5 0.002 0.002 8.768 8.842 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.669 8.778 mp_alltoall_i22 605 13.7 5.299 8.479 5.299 8.479 mp_waitall_1 84994 16.7 7.305 8.244 7.305 8.244 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.604 7.642 calculate_rho_elec 110 8.6 0.225 0.225 7.604 7.642 sum_up_and_integrate 110 10.3 0.150 0.151 7.226 7.241 integrate_v_rspace 110 11.3 0.004 0.004 7.076 7.091 init_scf_run 11 5.9 0.000 0.001 5.683 5.683 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.682 5.683 make_m2s 4110 13.4 0.043 0.044 5.123 5.646 make_images 4110 14.4 0.875 0.928 4.936 5.457 qs_ot_get_p 110 10.4 0.001 0.001 5.338 5.421 dbcsr_mm_accdrv_process 11614 15.7 3.186 3.668 5.111 5.369 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.732 5.252 apply_single 110 13.6 0.000 0.000 4.732 5.251 cp_fm_cholesky_invert 11 10.9 5.092 5.096 5.092 5.096 pw_transfer 1331 11.6 0.075 0.076 4.959 4.966 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.646 4.894 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.843 4.849 ot_diis_step 99 11.5 0.015 0.016 4.590 4.590 fft_wrap_pw1pw2_140 451 13.1 1.337 1.343 4.239 4.244 multiply_cannon_sync_h2d 8220 15.4 3.949 3.954 3.949 3.954 qs_ot_p2m_diag 48 11.0 0.151 0.157 3.819 3.827 grid_integrate_task_list 110 12.3 3.661 3.711 3.661 3.711 density_rs2pw 110 9.6 0.004 0.004 3.649 3.679 qs_energies_init_hamiltonians 11 5.9 0.012 0.013 3.656 3.657 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.806 3.535 make_images_data 4110 15.4 0.042 0.045 2.799 3.422 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.387 3.387 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.793 3.289 wfi_extrapolate 11 7.9 0.001 0.001 3.265 3.265 calculate_dm_sparse 110 9.5 0.001 0.001 3.177 3.196 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.816 2.822 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.810 2.810 cp_fm_diag_elpa_base 48 14.0 2.279 2.473 2.808 2.809 fft3d_ps 1111 14.6 1.301 1.311 2.773 2.779 grid_collocate_task_list 110 9.6 2.632 2.651 2.632 2.651 potential_pw2rs 110 12.3 0.021 0.022 2.366 2.371 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.330 2.369 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.333 2.362 calculate_first_density_matrix 1 7.0 0.001 0.001 2.300 2.301 rs_pw_transfer 902 11.9 0.011 0.012 2.217 2.252 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.177 2.225 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.190 mp_alltoall_d11v 2046 13.8 1.874 1.964 1.874 1.964 cp_fm_cholesky_decompose 22 10.9 1.931 1.948 1.931 1.948 qs_create_task_list 11 7.9 0.003 0.003 1.889 1.934 generate_qs_task_list 11 8.9 0.732 0.783 1.887 1.931 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.920 1.928 jit_kernel_multiply 10 15.5 1.726 1.810 1.726 1.810 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.741 1.791 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.728 1.742 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.065000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1193.181818, yerr=58.471156 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/15/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 1.094965E+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 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 631.930880E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2550202. 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 0.013 0.029 204.483 204.484 qs_mol_dyn_low 1 2.0 0.003 0.004 204.128 204.141 qs_forces 11 3.9 0.005 0.005 204.045 204.047 qs_energies 11 4.9 0.001 0.001 198.496 198.508 scf_env_do_scf 11 5.9 0.001 0.001 182.147 182.151 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.838 161.840 dbcsr_multiply_generic 2507 12.6 0.178 0.182 123.694 124.545 qs_scf_new_mos 117 7.6 0.001 0.001 122.088 122.434 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.087 122.433 velocity_verlet 10 3.0 0.001 0.002 122.345 122.346 ot_scf_mini 117 9.6 0.003 0.003 115.499 115.763 multiply_cannon 2507 13.6 0.239 0.248 100.729 102.957 multiply_cannon_loop 2507 14.6 2.057 2.108 98.652 100.718 ot_mini 117 10.6 0.001 0.001 65.244 65.542 multiply_cannon_multrec 60168 15.6 33.479 35.308 41.623 43.738 qs_ot_get_derivative 117 11.6 0.001 0.002 40.341 40.629 rebuild_ks_matrix 128 8.3 0.001 0.001 33.564 33.947 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.563 33.946 mp_waitall_1 267128 16.5 27.817 30.869 27.817 30.869 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.147 30.459 multiply_cannon_sync_h2d 60168 15.6 27.695 29.813 27.695 29.813 qs_ot_get_p 128 10.4 0.001 0.001 28.430 28.724 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.411 25.096 apply_single 128 13.6 0.001 0.001 24.411 25.095 ot_diis_step 117 11.6 0.008 0.008 24.614 24.616 qs_ot_p2m_diag 83 11.4 0.079 0.092 21.931 21.979 init_scf_loop 11 6.9 0.000 0.001 20.236 20.237 cp_dbcsr_syevd 83 12.4 0.004 0.005 19.389 19.390 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.119 19.304 multiply_cannon_metrocomm3 60168 15.6 0.113 0.118 15.574 17.623 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.477 16.508 cp_fm_diag_elpa_base 83 14.4 16.412 16.453 16.474 16.506 prepare_preconditioner 11 7.9 0.000 0.000 15.639 15.676 make_preconditioner 11 8.9 0.000 0.000 15.639 15.676 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.877 15.032 make_m2s 5014 13.6 0.106 0.114 13.720 14.170 sum_up_and_integrate 128 10.3 0.089 0.108 14.066 14.085 integrate_v_rspace 128 11.3 0.004 0.005 13.976 13.998 make_images 5014 14.6 0.399 0.419 13.536 13.993 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.522 13.665 calculate_rho_elec 128 8.7 0.046 0.065 13.521 13.665 init_scf_run 11 5.9 0.000 0.001 12.238 12.239 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.238 12.239 density_rs2pw 128 9.7 0.006 0.007 6.881 10.069 mp_sum_l 7870 13.0 7.967 9.428 7.967 9.428 wfi_extrapolate 11 7.9 0.001 0.001 8.985 8.985 cp_fm_cholesky_invert 11 10.9 8.902 8.910 8.902 8.910 rs_pw_transfer 1046 11.9 0.017 0.019 5.482 8.683 calculate_dm_sparse 128 9.5 0.001 0.001 8.387 8.521 dbcsr_mm_accdrv_process 124484 16.2 3.116 3.263 7.705 8.241 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.824 7.995 pw_transfer 1547 11.6 0.074 0.090 7.709 7.905 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 5.865 7.898 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.750 7.804 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.505 7.696 make_images_data 5014 15.6 0.064 0.070 6.747 7.656 grid_integrate_task_list 128 12.3 7.019 7.487 7.019 7.487 hybrid_alltoall_any 5200 16.5 0.294 2.252 5.935 7.175 fft_wrap_pw1pw2_140 523 13.2 1.281 1.339 6.629 6.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.645 6.653 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.785 5.872 mp_waitany 16020 13.9 2.601 5.799 2.601 5.799 grid_collocate_task_list 128 9.7 4.567 5.797 4.567 5.797 mp_alltoall_d11v 2415 14.1 4.225 5.672 4.225 5.672 fft3d_ps 1291 14.7 2.160 2.855 5.261 5.589 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.303 2.079 5.287 cp_fm_cholesky_decompose 22 10.9 4.566 4.580 4.566 4.580 potential_pw2rs 128 12.3 0.009 0.011 4.464 4.482 mp_sum_d 4465 12.1 3.468 4.134 3.468 4.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.484000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.272727, yerr=6.836944 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.435328E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58147. MP_Allreduce 11016 966. MP_Sync 86 MP_Alltoall 1955 5616330. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.093 187.075 187.077 qs_mol_dyn_low 1 2.0 0.003 0.004 186.518 186.545 qs_forces 11 3.9 0.005 0.006 186.435 186.436 qs_energies 11 4.9 0.002 0.006 179.665 179.676 scf_env_do_scf 11 5.9 0.001 0.001 163.198 163.208 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 130.739 130.741 velocity_verlet 10 3.0 0.001 0.002 117.128 117.129 dbcsr_multiply_generic 2485 12.5 0.186 0.190 95.637 96.769 qs_scf_new_mos 116 7.6 0.001 0.001 91.414 92.032 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.413 92.031 ot_scf_mini 116 9.6 0.004 0.005 86.669 87.313 multiply_cannon 2485 13.5 0.472 0.522 76.219 80.064 multiply_cannon_loop 2485 14.5 1.242 1.281 73.414 75.829 ot_mini 116 10.6 0.001 0.001 48.826 49.432 mp_waitall_1 212858 16.6 24.051 37.374 24.051 37.374 multiply_cannon_multrec 29820 15.5 21.732 25.520 31.277 35.449 rebuild_ks_matrix 127 8.3 0.001 0.001 32.781 33.233 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 32.781 33.233 init_scf_loop 11 6.9 0.001 0.004 32.374 32.375 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.452 29.858 multiply_cannon_metrocomm3 29820 15.5 0.090 0.096 15.581 28.992 prepare_preconditioner 11 7.9 0.000 0.000 28.037 28.128 make_preconditioner 11 8.9 0.000 0.001 28.037 28.128 qs_ot_get_derivative 116 11.6 0.001 0.003 27.125 27.785 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.725 27.299 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.791 22.921 apply_single 127 13.6 0.001 0.001 21.791 22.921 ot_diis_step 116 11.6 0.014 0.015 21.531 21.534 multiply_cannon_sync_h2d 29820 15.5 19.081 21.242 19.081 21.242 qs_ot_get_p 127 10.4 0.001 0.002 19.955 20.609 cp_fm_cholesky_invert 11 10.9 16.375 16.387 16.375 16.387 qs_ot_p2m_diag 82 11.4 0.185 0.213 15.332 15.363 make_m2s 4970 13.5 0.090 0.097 13.909 15.020 make_images 4970 14.5 1.137 1.320 13.699 14.806 sum_up_and_integrate 127 10.3 0.114 0.130 14.472 14.505 integrate_v_rspace 127 11.3 0.003 0.004 14.357 14.396 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.220 14.221 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.800 13.836 calculate_rho_elec 127 8.7 0.087 0.103 13.799 13.836 init_scf_run 11 5.9 0.000 0.001 11.614 11.616 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.614 11.616 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.176 11.203 cp_fm_diag_elpa_base 82 14.4 10.928 10.993 11.170 11.193 qs_ot_get_derivative_diag 76 12.4 0.002 0.004 10.587 11.070 multiply_cannon_metrocomm4 27335 15.5 0.097 0.110 3.779 10.363 mp_irecv_dv 68888 16.3 3.585 9.982 3.585 9.982 make_images_data 4970 15.5 0.063 0.069 8.493 9.892 density_rs2pw 127 9.7 0.006 0.007 7.251 9.701 dbcsr_mm_accdrv_process 61726 16.2 4.541 5.431 9.002 9.614 hybrid_alltoall_any 5155 16.4 0.341 1.508 7.248 9.087 pw_transfer 1535 11.6 0.085 0.102 8.634 8.702 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.411 8.478 wfi_extrapolate 11 7.9 0.001 0.001 8.268 8.268 rs_pw_transfer 1038 11.9 0.015 0.017 5.528 8.012 fft_wrap_pw1pw2_140 519 13.2 1.331 1.351 7.468 7.572 grid_integrate_task_list 127 12.3 7.112 7.531 7.112 7.531 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.104 6.828 cp_fm_cholesky_decompose 22 10.9 6.734 6.821 6.734 6.821 calculate_dm_sparse 127 9.5 0.001 0.001 6.419 6.559 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.131 6.140 grid_collocate_task_list 127 9.7 4.667 5.815 4.667 5.815 fft3d_ps 1281 14.7 2.793 2.971 5.737 5.800 mp_sum_l 7804 13.0 3.794 5.724 3.794 5.724 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.238 5.372 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.269 5.340 mp_waitany 11660 13.9 2.467 5.021 2.467 5.021 mp_alltoall_d11v 2401 14.1 4.018 4.894 4.018 4.894 potential_pw2rs 127 12.3 0.015 0.017 4.806 4.827 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.379 2.083 4.546 mp_allgather_i34 2485 14.5 1.377 4.250 1.377 4.250 mp_sum_d 4454 12.1 2.677 4.130 2.677 4.130 dbcsr_complete_redistribute 393 12.7 0.827 0.941 3.175 3.999 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.714 3.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.077000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.272727, yerr=5.462116 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 951.271424E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58169. MP_Allreduce 11010 1007. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.091 173.976 173.990 qs_mol_dyn_low 1 2.0 0.003 0.004 173.150 173.181 qs_forces 11 3.9 0.012 0.053 172.998 173.005 qs_energies 11 4.9 0.021 0.081 166.334 166.344 scf_env_do_scf 11 5.9 0.001 0.002 150.243 150.244 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.668 115.669 velocity_verlet 10 3.0 0.002 0.002 111.009 111.012 dbcsr_multiply_generic 2485 12.5 0.178 0.185 81.146 82.288 qs_scf_new_mos 116 7.6 0.001 0.001 77.852 78.151 qs_scf_loop_do_ot 116 8.6 0.001 0.001 77.851 78.150 ot_scf_mini 116 9.6 0.004 0.004 73.750 74.119 multiply_cannon 2485 13.5 0.525 0.542 61.595 66.428 multiply_cannon_loop 2485 14.5 0.850 0.878 58.324 60.969 ot_mini 116 10.6 0.001 0.001 42.088 42.483 init_scf_loop 11 6.9 0.002 0.006 34.479 34.480 mp_waitall_1 169034 16.6 24.520 33.844 24.520 33.844 rebuild_ks_matrix 127 8.3 0.001 0.001 30.536 31.033 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.536 31.033 prepare_preconditioner 11 7.9 0.000 0.000 30.435 30.485 make_preconditioner 11 8.9 0.000 0.001 30.435 30.485 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.094 29.503 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.481 27.935 multiply_cannon_multrec 19880 15.5 13.477 16.218 22.290 25.074 multiply_cannon_metrocomm3 19880 15.5 0.057 0.061 14.871 24.490 qs_ot_get_derivative 116 11.6 0.002 0.002 22.584 22.952 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.626 20.609 apply_single 127 13.6 0.001 0.001 19.626 20.608 ot_diis_step 116 11.6 0.018 0.018 19.399 19.399 qs_ot_get_p 127 10.4 0.001 0.002 17.173 17.630 multiply_cannon_sync_h2d 19880 15.5 14.118 15.801 14.118 15.801 make_m2s 4970 13.5 0.082 0.087 14.505 15.691 make_images 4970 14.5 1.168 1.246 14.272 15.456 sum_up_and_integrate 127 10.3 0.132 0.143 14.431 14.459 integrate_v_rspace 127 11.3 0.004 0.005 14.298 14.326 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.118 14.147 calculate_rho_elec 127 8.7 0.129 0.144 14.118 14.146 cp_fm_cholesky_invert 11 10.9 14.100 14.109 14.100 14.109 qs_ot_p2m_diag 82 11.4 0.263 0.269 12.959 12.967 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.986 11.988 init_scf_run 11 5.9 0.000 0.001 10.871 10.897 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.871 10.897 make_images_data 4970 15.5 0.059 0.067 8.955 10.572 hybrid_alltoall_any 5155 16.4 0.428 1.956 7.860 9.858 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.062 9.082 cp_fm_diag_elpa_base 82 14.4 8.667 8.818 9.060 9.079 multiply_cannon_metrocomm4 17395 15.5 0.062 0.071 3.409 9.077 qs_ot_get_derivative_diag 76 12.4 0.003 0.004 8.795 9.071 density_rs2pw 127 9.7 0.006 0.007 7.009 8.966 pw_transfer 1535 11.6 0.085 0.102 8.804 8.902 mp_irecv_dv 49801 16.2 3.288 8.838 3.288 8.838 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.582 8.684 dbcsr_mm_accdrv_process 41158 16.2 4.622 5.383 8.262 8.397 fft_wrap_pw1pw2_140 519 13.2 1.413 1.462 7.651 7.763 grid_integrate_task_list 127 12.3 7.226 7.692 7.226 7.692 wfi_extrapolate 11 7.9 0.001 0.001 7.380 7.380 cp_fm_cholesky_decompose 22 10.9 7.216 7.245 7.216 7.245 cp_fm_upper_to_full 104 14.8 5.718 7.226 5.718 7.226 rs_pw_transfer 1038 11.9 0.017 0.019 4.990 7.018 dbcsr_complete_redistribute 393 12.7 1.170 1.202 4.396 6.169 calculate_dm_sparse 127 9.5 0.001 0.001 5.868 5.982 fft3d_ps 1281 14.7 2.728 2.962 5.715 5.807 grid_collocate_task_list 127 9.7 4.866 5.606 4.866 5.606 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.543 5.551 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.567 5.234 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.254 5.027 mp_alltoall_d11v 2401 14.1 4.191 4.946 4.191 4.946 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.655 4.801 mp_sum_l 7804 13.0 3.311 4.773 3.311 4.773 mp_allgather_i34 2485 14.5 1.781 4.740 1.781 4.740 potential_pw2rs 127 12.3 0.021 0.023 4.564 4.574 mp_waitany 11660 13.9 2.347 4.386 2.347 4.386 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.320 4.043 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.933 3.977 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.349 1.946 3.953 qs_energies_init_hamiltonians 11 5.9 0.006 0.011 3.917 3.918 mp_alltoall_i22 712 14.1 1.954 3.837 1.954 3.837 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.552 3.597 mp_sum_d 4451 12.1 2.355 3.576 2.355 3.576 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.990000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=897.000000, yerr=14.051852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 4.387845E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 6033792 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.1 marketing flops 145.664755E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.130693E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155960 MPI messages size (bytes): total size 2.039892E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.764674E+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 323584 36561747968 131072 < size <= 4194304 723224 792350162944 4194304 < size <= 16777216 70800 669922351760 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57660. MP_Allreduce 11099 1087. MP_Sync 88 MP_Alltoall 1724 12509527. 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 0.029 0.062 187.451 187.452 qs_mol_dyn_low 1 2.0 0.003 0.012 186.955 186.968 qs_forces 11 3.9 0.004 0.005 186.845 186.853 qs_energies 11 4.9 0.003 0.014 179.689 179.697 scf_env_do_scf 11 5.9 0.001 0.001 163.077 163.085 velocity_verlet 10 3.0 0.002 0.002 125.068 125.071 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.033 117.034 dbcsr_multiply_generic 2535 12.6 0.189 0.194 79.731 80.574 qs_scf_new_mos 118 7.6 0.001 0.001 79.260 79.566 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.259 79.566 ot_scf_mini 118 9.6 0.003 0.005 74.764 75.083 multiply_cannon 2535 13.6 0.560 0.607 54.947 58.472 multiply_cannon_loop 2535 14.6 1.197 1.227 51.393 53.538 init_scf_loop 11 6.9 0.001 0.006 45.926 45.927 ot_mini 118 10.6 0.001 0.001 42.505 42.827 prepare_preconditioner 11 7.9 0.000 0.001 41.806 41.837 make_preconditioner 11 8.9 0.000 0.001 41.806 41.837 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.457 40.483 multiply_cannon_multrec 30420 15.6 13.963 19.601 25.985 31.238 rebuild_ks_matrix 129 8.3 0.001 0.001 30.104 30.386 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.021 30.104 30.385 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.154 27.397 mp_waitall_1 149508 16.7 17.619 26.703 17.619 26.703 qs_ot_get_derivative 118 11.6 0.002 0.002 22.766 23.081 make_m2s 5070 13.6 0.098 0.102 20.384 21.362 make_images 5070 14.6 1.953 2.258 20.081 21.058 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.145 19.689 apply_single 129 13.6 0.001 0.001 19.145 19.689 ot_diis_step 118 11.6 0.018 0.019 19.621 19.622 qs_ot_get_p 129 10.4 0.001 0.001 18.460 18.821 cp_fm_upper_to_full 106 14.8 11.509 16.929 11.509 16.929 cp_fm_cholesky_invert 11 10.9 15.814 15.822 15.814 15.822 multiply_cannon_metrocomm3 30420 15.6 0.048 0.050 6.605 15.541 sum_up_and_integrate 129 10.3 0.141 0.151 14.774 14.796 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.640 14.671 calculate_rho_elec 129 8.7 0.174 0.190 14.640 14.670 integrate_v_rspace 129 11.3 0.004 0.004 14.633 14.658 qs_ot_p2m_diag 84 11.4 0.347 0.394 14.313 14.366 cp_dbcsr_syevd 84 12.4 0.005 0.005 13.039 13.041 make_images_data 5070 15.6 0.063 0.068 11.093 12.764 dbcsr_complete_redistribute 397 12.7 1.500 1.619 8.978 12.752 multiply_cannon_sync_h2d 30420 15.6 11.818 12.549 11.818 12.549 dbcsr_mm_accdrv_process 62852 16.2 7.492 8.291 11.588 12.146 hybrid_alltoall_any 5257 16.5 0.528 2.218 9.795 11.740 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.650 11.431 init_scf_run 11 5.9 0.000 0.001 10.727 10.728 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.726 10.728 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.327 10.007 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.855 9.873 cp_fm_diag_elpa_base 84 14.4 8.889 9.227 9.849 9.865 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.446 9.679 mp_alltoall_i22 720 14.1 5.598 9.499 5.598 9.499 pw_transfer 1559 11.6 0.087 0.101 9.193 9.274 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.966 9.050 density_rs2pw 129 9.7 0.006 0.006 6.996 8.376 fft_wrap_pw1pw2_140 527 13.2 1.573 1.607 7.997 8.086 grid_integrate_task_list 129 12.3 7.557 7.997 7.557 7.997 wfi_extrapolate 11 7.9 0.001 0.001 7.672 7.672 cp_fm_cholesky_decompose 22 10.9 7.410 7.513 7.410 7.513 multiply_cannon_metrocomm4 25350 15.6 0.084 0.096 2.922 7.445 mp_irecv_dv 76889 16.2 2.767 7.168 2.767 7.168 calculate_dm_sparse 129 9.5 0.001 0.001 6.170 6.280 rs_pw_transfer 1054 12.0 0.014 0.015 4.604 6.065 fft3d_ps 1301 14.7 2.862 2.968 5.870 5.935 grid_collocate_task_list 129 9.7 5.112 5.778 5.112 5.778 mp_alltoall_d11v 2429 14.1 4.941 5.614 4.941 5.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.411 5.458 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.391 4.488 potential_pw2rs 129 12.3 0.024 0.024 4.477 4.488 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.460 4.462 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.375 4.457 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.157 4.198 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.452000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.000000, yerr=15.159755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 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 11528896499712 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 93.514757E+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 6755939872 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 1.511436E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350154512 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58209. MP_Allreduce 11002 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.070 171.386 171.402 qs_mol_dyn_low 1 2.0 0.003 0.003 170.901 170.924 qs_forces 11 3.9 0.004 0.004 170.761 170.766 qs_energies 11 4.9 0.001 0.001 163.197 163.205 scf_env_do_scf 11 5.9 0.001 0.001 146.073 146.078 velocity_verlet 10 3.0 0.002 0.002 112.411 112.415 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 111.034 111.035 dbcsr_multiply_generic 2507 12.6 0.181 0.191 73.798 74.222 qs_scf_new_mos 117 7.6 0.001 0.001 72.725 72.824 qs_scf_loop_do_ot 117 8.6 0.001 0.001 72.724 72.823 ot_scf_mini 117 9.6 0.003 0.004 68.295 68.385 multiply_cannon 2507 13.6 0.581 0.611 54.271 59.214 multiply_cannon_loop 2507 14.6 0.444 0.459 49.712 50.514 ot_mini 117 10.6 0.001 0.001 39.284 39.359 init_scf_loop 11 6.9 0.000 0.000 34.895 34.897 mp_waitall_1 125778 16.7 25.864 33.105 25.864 33.105 prepare_preconditioner 11 7.9 0.000 0.000 30.950 30.980 make_preconditioner 11 8.9 0.000 0.000 30.950 30.980 rebuild_ks_matrix 128 8.3 0.001 0.001 29.715 29.770 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.019 29.714 29.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.953 29.231 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.978 27.028 multiply_cannon_multrec 10028 15.6 10.389 14.570 17.667 21.201 ot_diis_step 117 11.6 0.020 0.022 19.876 19.876 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.454 19.766 apply_single 128 13.6 0.001 0.001 19.454 19.765 multiply_cannon_metrocomm3 10028 15.6 0.023 0.024 12.362 19.448 qs_ot_get_derivative 117 11.6 0.002 0.002 19.339 19.434 make_m2s 5014 13.6 0.068 0.071 16.043 18.415 make_images 5014 14.6 2.267 2.636 15.735 18.103 cp_fm_cholesky_invert 11 10.9 17.676 17.682 17.676 17.682 qs_ot_get_p 128 10.4 0.001 0.001 16.150 16.233 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.184 15.226 calculate_rho_elec 128 8.7 0.256 0.266 15.183 15.226 sum_up_and_integrate 128 10.3 0.179 0.189 14.875 14.922 integrate_v_rspace 128 11.3 0.004 0.004 14.695 14.753 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.530 12.547 multiply_cannon_sync_h2d 10028 15.6 11.626 12.376 11.626 12.376 make_images_data 5014 15.6 0.052 0.060 9.854 12.246 hybrid_alltoall_any 5200 16.5 0.819 3.647 9.719 12.150 cp_dbcsr_syevd 83 12.4 0.005 0.005 11.415 11.417 init_scf_run 11 5.9 0.000 0.001 10.336 10.336 scf_env_initial_rho_setup 11 6.9 0.006 0.008 10.335 10.336 pw_transfer 1547 11.6 0.086 0.096 9.370 9.394 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.147 9.178 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.482 8.494 cp_fm_diag_elpa_base 83 14.4 8.245 8.314 8.478 8.491 fft_wrap_pw1pw2_140 523 13.2 1.911 1.946 8.128 8.163 grid_integrate_task_list 128 12.3 7.745 8.161 7.745 8.161 cp_fm_cholesky_decompose 22 10.9 7.746 7.845 7.746 7.845 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.771 7.837 mp_allgather_i34 2507 14.6 2.763 7.786 2.763 7.786 density_rs2pw 128 9.7 0.006 0.006 6.758 7.630 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.434 7.571 dbcsr_mm_accdrv_process 20762 16.1 2.594 3.274 6.903 7.548 wfi_extrapolate 11 7.9 0.001 0.001 7.334 7.334 calculate_dm_sparse 128 9.5 0.001 0.001 6.090 6.182 grid_collocate_task_list 128 9.7 5.395 6.059 5.395 6.059 fft3d_ps 1291 14.7 2.759 2.829 5.627 5.663 mp_alltoall_d11v 2415 14.1 4.884 5.600 4.884 5.600 dbcsr_complete_redistribute 395 12.7 2.099 2.154 4.958 5.355 qs_energies_init_hamiltonians 11 5.9 0.024 0.027 5.330 5.331 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.158 5.164 rs_pw_transfer 1046 11.9 0.014 0.014 4.024 4.909 multiply_cannon_metrocomm4 7521 15.6 0.024 0.028 1.851 4.367 potential_pw2rs 128 12.3 0.027 0.027 4.315 4.327 mp_irecv_dv 28860 15.9 1.816 4.294 1.816 4.294 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.126 4.149 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.884 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.678 3.716 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.639 3.645 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.227 3.556 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.443 3.507 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.402000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.818182, yerr=52.226462 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963540389888 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019185291264 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019185291264 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820203E+12 0.0% 0.0% 100.0% flops max/rank 11.609359E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705859104 0.0% 0.0% 100.0% number of processed stacks 1951856 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.6 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 2.909381E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99840 MPI messages size (bytes): total size 1.127791E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.295984E+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 45208 35114713088 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108641296 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 59878. MP_Allreduce 10875 1518. MP_Sync 86 MP_Alltoall 1700 36954490. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.070 0.083 293.369 293.379 qs_mol_dyn_low 1 2.0 0.003 0.004 292.784 292.801 qs_forces 11 3.9 0.005 0.005 292.677 292.680 qs_energies 11 4.9 0.001 0.001 283.379 283.394 scf_env_do_scf 11 5.9 0.001 0.001 261.410 261.424 velocity_verlet 10 3.0 0.002 0.002 211.162 211.169 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 137.627 137.629 init_scf_loop 11 6.9 0.000 0.000 123.532 123.535 prepare_preconditioner 11 7.9 0.000 0.000 118.556 118.582 make_preconditioner 11 8.9 0.000 0.000 118.556 118.582 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.607 115.733 qs_scf_new_mos 116 7.6 0.001 0.001 87.806 87.906 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.805 87.905 ot_scf_mini 116 9.6 0.004 0.004 83.121 83.162 dbcsr_multiply_generic 2496 12.6 0.209 0.221 81.807 82.345 cp_fm_upper_to_full 104 14.8 52.632 75.580 52.632 75.580 multiply_cannon 2496 13.6 0.707 0.758 57.832 58.306 multiply_cannon_loop 2496 14.6 0.468 0.477 54.284 55.620 ot_mini 116 10.6 0.001 0.001 43.484 43.524 dbcsr_complete_redistribute 393 12.7 3.984 4.031 29.507 42.245 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.165 38.910 rebuild_ks_matrix 127 8.3 0.001 0.001 37.704 37.753 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 37.704 37.752 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.902 36.503 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.710 34.759 mp_alltoall_i22 712 14.1 21.728 34.622 21.728 34.622 cp_fm_cholesky_invert 11 10.9 32.962 32.968 32.962 32.968 mp_waitall_1 103208 16.8 27.625 31.579 27.625 31.579 qs_ot_get_p 127 10.4 0.001 0.001 24.385 24.389 qs_ot_get_derivative 116 11.6 0.002 0.003 23.665 23.707 qs_ot_p2m_diag 82 11.4 0.867 0.873 20.462 20.491 make_m2s 4992 13.6 0.077 0.080 19.540 20.464 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.138 20.159 calculate_rho_elec 127 8.7 0.477 0.478 20.138 20.159 make_images 4992 14.6 3.691 3.872 19.065 19.987 ot_diis_step 116 11.6 0.022 0.023 19.795 19.796 multiply_cannon_metrocomm3 9984 15.6 0.023 0.023 18.430 19.512 sum_up_and_integrate 127 10.3 0.319 0.323 19.322 19.418 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.081 19.273 apply_single 127 13.6 0.001 0.001 19.081 19.273 integrate_v_rspace 127 11.3 0.004 0.004 19.002 19.097 cp_dbcsr_syevd 82 12.4 0.005 0.006 18.781 18.783 multiply_cannon_multrec 9984 15.6 10.388 12.121 17.811 17.897 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.700 15.700 cp_fm_diag_elpa_base 82 14.4 11.354 12.929 15.696 15.697 multiply_cannon_sync_h2d 9984 15.6 15.534 15.542 15.534 15.542 hybrid_alltoall_any 5177 16.4 1.297 3.040 10.670 12.473 make_images_data 4992 15.6 0.060 0.064 10.476 12.153 init_scf_run 11 5.9 0.000 0.001 12.034 12.034 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.034 12.034 pw_transfer 1535 11.6 0.092 0.094 11.127 11.141 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 10.894 10.905 fft_wrap_pw1pw2_140 519 13.2 3.052 3.087 9.700 9.712 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.173 9.211 mp_alltoall_d11v 2401 14.1 8.289 9.125 8.289 9.125 dbcsr_mm_accdrv_process 20634 16.0 3.756 5.601 7.185 9.023 wfi_extrapolate 11 7.9 0.001 0.001 8.886 8.886 cp_fm_cholesky_decompose 22 10.9 8.770 8.792 8.770 8.792 grid_integrate_task_list 127 12.3 8.512 8.677 8.512 8.677 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.967 7.968 density_rs2pw 127 9.7 0.005 0.005 7.748 7.872 calculate_dm_sparse 127 9.5 0.001 0.001 6.565 6.640 grid_collocate_task_list 127 9.7 6.311 6.340 6.311 6.340 rs_scatter_matrices 138 9.7 3.612 4.552 6.046 6.267 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.195 6.252 fft3d_ps 1281 14.7 2.820 2.830 5.999 6.015 copy_dbcsr_to_fm 185 11.7 0.004 0.004 5.931 6.003 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.379000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2587.545455, yerr=165.007087 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 1.262428E+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 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 18358086. 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 0.016 0.034 84.639 84.641 qs_energies 1 2.0 0.000 0.000 84.153 84.166 ls_scf 1 3.0 0.000 0.000 83.247 83.260 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.473 72.624 multiply_cannon 111 7.7 0.018 0.021 55.887 57.482 multiply_cannon_loop 111 8.7 0.206 0.219 52.432 54.018 ls_scf_main 1 4.0 0.000 0.000 52.123 52.135 density_matrix_trs4 2 5.0 0.002 0.003 46.654 46.756 ls_scf_init_scf 1 4.0 0.000 0.000 28.087 28.088 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.002 27.058 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.897 24.920 mp_waitall_1 11031 10.9 22.447 24.207 22.447 24.207 multiply_cannon_multrec 2664 9.7 8.142 8.847 15.441 17.084 multiply_cannon_sync_h2d 2664 9.7 13.578 15.710 13.578 15.710 make_m2s 222 7.7 0.008 0.011 12.951 13.556 make_images 222 8.7 0.100 0.109 12.929 13.536 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.733 12.373 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.508 8.785 make_images_data 222 9.7 0.004 0.005 7.538 8.074 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.616 6.919 7.933 hybrid_alltoall_any 227 10.6 0.215 1.830 6.461 7.704 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.137 6.209 7.137 calculate_norms 4752 9.8 5.496 6.056 5.496 6.056 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.045 5.186 mp_sum_l 807 5.4 3.097 4.911 3.097 4.911 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.335 3.988 make_images_sizes 222 9.7 0.000 0.000 0.664 3.482 mp_alltoall_i44 222 10.7 0.664 3.482 0.664 3.482 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.051 3.236 mp_irecv_dv 6231 10.9 2.034 3.214 2.034 3.214 arnoldi_extremal 4 6.8 0.000 0.000 3.159 3.185 arnoldi_normal_ev 4 7.8 0.001 0.003 3.159 3.185 build_subspace 16 8.4 0.009 0.012 3.068 3.070 ls_scf_post 1 4.0 0.000 0.000 3.037 3.050 ls_scf_store_result 1 5.0 0.000 0.000 2.859 2.901 dbcsr_special_finalize 555 9.7 0.005 0.006 2.396 2.823 dbcsr_merge_single_wm 555 10.7 0.464 0.614 2.387 2.815 make_images_pack 222 9.7 2.206 2.620 2.207 2.622 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.306 2.547 dbcsr_sort_data 658 11.4 2.179 2.501 2.179 2.501 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.461 2.071 2.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.242 2.316 buffer_matrices_ensure_size 222 8.7 1.756 2.100 1.756 2.100 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.763 1.764 rebuild_ks_matrix 3 7.3 0.000 0.000 1.754 1.755 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.006 1.754 1.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.641000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 2.100511E+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 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.056 91.292 91.293 qs_energies 1 2.0 0.000 0.000 90.448 90.453 ls_scf 1 3.0 0.005 0.022 89.137 89.141 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.076 75.450 multiply_cannon 111 7.7 0.029 0.038 53.274 56.942 ls_scf_main 1 4.0 0.001 0.023 54.743 54.759 multiply_cannon_loop 111 8.7 0.116 0.123 50.033 53.312 density_matrix_trs4 2 5.0 0.003 0.015 48.962 49.166 ls_scf_init_scf 1 4.0 0.000 0.002 30.810 30.812 mp_waitall_1 9105 10.9 20.992 30.646 20.992 30.646 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.599 29.706 multiply_cannon_multrec 1332 9.7 13.271 17.138 22.478 27.472 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.203 27.217 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.628 21.278 make_m2s 222 7.7 0.006 0.008 15.254 15.892 make_images 222 8.7 1.578 1.956 15.224 15.863 dbcsr_mm_accdrv_process 4041 10.4 0.327 0.452 8.809 10.375 dbcsr_mm_accdrv_process_sort 4041 11.4 8.399 9.923 8.399 9.923 make_images_data 222 9.7 0.004 0.004 8.742 9.672 hybrid_alltoall_any 227 10.6 0.522 2.476 8.200 9.074 mp_sum_l 807 5.4 5.229 8.283 5.229 8.283 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.239 7.662 mp_irecv_dv 3311 11.0 3.220 7.608 3.220 7.608 calculate_norms 2376 9.8 6.009 6.843 6.009 6.843 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.074 6.729 multiply_cannon_sync_h2d 1332 9.7 4.824 6.169 4.824 6.169 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.086 5.320 arnoldi_extremal 4 6.8 0.000 0.000 4.635 4.655 arnoldi_normal_ev 4 7.8 0.003 0.007 4.635 4.655 build_subspace 16 8.4 0.014 0.021 4.380 4.383 ls_scf_post 1 4.0 0.000 0.000 3.579 3.584 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.145 3.373 ls_scf_store_result 1 5.0 0.000 0.000 3.241 3.362 dbcsr_matrix_vector_mult_local 304 10.0 2.749 3.223 2.751 3.225 ls_scf_dm_to_ks 2 5.0 0.001 0.008 2.607 2.737 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.255 2.712 make_images_pack 222 9.7 2.033 2.426 2.035 2.428 mp_allgather_i34 111 8.7 0.941 2.277 0.941 2.277 dbcsr_sort_data 436 11.2 1.829 2.077 1.829 2.077 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.883 1.886 rebuild_ks_matrix 3 7.3 0.000 0.000 1.871 1.873 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.012 1.871 1.873 dbcsr_data_new 4174 10.1 1.610 1.836 1.610 1.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.293000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1746.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 2.771096E+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 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.052 93.413 93.414 qs_energies 1 2.0 0.000 0.000 92.653 92.658 ls_scf 1 3.0 0.000 0.001 91.253 91.259 dbcsr_multiply_generic 111 6.7 0.016 0.019 76.028 76.322 ls_scf_main 1 4.0 0.000 0.002 56.961 56.964 multiply_cannon 111 7.7 0.038 0.078 52.626 56.067 multiply_cannon_loop 111 8.7 0.100 0.106 49.015 52.623 density_matrix_trs4 2 5.0 0.002 0.004 50.983 51.181 mp_waitall_1 7281 11.0 23.568 33.639 23.568 33.639 ls_scf_init_scf 1 4.0 0.000 0.003 30.705 30.708 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.534 29.603 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.102 27.114 multiply_cannon_multrec 888 9.7 12.540 15.375 21.069 24.495 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.043 22.876 make_m2s 222 7.7 0.006 0.006 16.759 18.028 make_images 222 8.7 1.965 2.297 16.721 17.988 make_images_data 222 9.7 0.004 0.004 9.604 10.743 hybrid_alltoall_any 227 10.6 0.618 2.859 9.313 10.547 dbcsr_mm_accdrv_process 3754 10.4 0.242 0.421 8.048 9.282 dbcsr_mm_accdrv_process_sort 3754 11.4 7.676 8.861 7.676 8.861 mp_sum_l 807 5.4 5.104 8.416 5.104 8.416 multiply_cannon_sync_h2d 888 9.7 5.971 7.314 5.971 7.314 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.445 6.912 mp_irecv_dv 2335 11.1 2.431 6.872 2.431 6.872 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.836 6.468 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.539 6.458 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.972 5.235 arnoldi_extremal 4 6.8 0.000 0.000 5.077 5.090 arnoldi_normal_ev 4 7.8 0.001 0.005 5.076 5.090 build_subspace 16 8.4 0.014 0.020 4.759 4.765 calculate_norms 1584 9.8 4.318 4.705 4.318 4.705 mp_allgather_i34 111 8.7 1.457 3.790 1.457 3.790 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.434 3.755 ls_scf_post 1 4.0 0.000 0.001 3.588 3.595 dbcsr_matrix_vector_mult_local 304 10.0 3.019 3.588 3.021 3.590 ls_scf_store_result 1 5.0 0.000 0.000 3.321 3.416 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.826 2.941 make_images_pack 222 9.7 1.829 2.123 1.832 2.126 make_images_sizes 222 9.7 0.000 0.000 0.835 2.070 mp_alltoall_i44 222 10.7 0.835 2.069 0.835 2.069 dbcsr_sort_data 325 11.1 1.885 2.054 1.885 2.054 dbcsr_data_release 9322 10.9 1.298 1.955 1.298 1.955 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.908 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.890 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.888 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.414000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2165.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 3.317838E+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 266696. MP_Allreduce 3058 13371. 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 0.037 0.051 97.512 97.514 qs_energies 1 2.0 0.000 0.000 96.771 96.777 ls_scf 1 3.0 0.000 0.001 95.090 95.098 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.849 79.134 ls_scf_main 1 4.0 0.000 0.002 59.035 59.036 multiply_cannon 111 7.7 0.058 0.135 51.866 56.398 density_matrix_trs4 2 5.0 0.002 0.004 52.918 53.069 multiply_cannon_loop 111 8.7 0.115 0.125 46.841 49.853 ls_scf_init_scf 1 4.0 0.000 0.002 32.811 32.812 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.493 31.573 mp_waitall_1 6369 11.0 23.140 29.620 23.140 29.620 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.020 29.029 multiply_cannon_multrec 1332 9.7 14.123 17.127 21.965 24.950 make_m2s 222 7.7 0.006 0.008 21.149 22.537 make_images 222 8.7 3.136 3.586 21.099 22.489 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.621 17.558 make_images_data 222 9.7 0.004 0.004 11.848 13.375 hybrid_alltoall_any 227 10.6 0.795 3.742 11.259 13.030 dbcsr_mm_accdrv_process 3641 10.4 0.219 0.406 7.480 9.007 dbcsr_mm_accdrv_process_sort 3641 11.4 7.107 8.589 7.107 8.589 mp_sum_l 807 5.4 4.261 8.338 4.261 8.338 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.291 6.639 multiply_cannon_sync_h2d 1332 9.7 5.508 6.182 5.508 6.182 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.099 6.134 mp_irecv_dv 3229 10.9 2.075 6.051 2.075 6.051 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.499 5.682 arnoldi_extremal 4 6.8 0.000 0.000 5.194 5.215 arnoldi_normal_ev 4 7.8 0.001 0.005 5.194 5.215 build_subspace 16 8.4 0.014 0.021 4.853 4.860 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.553 4.692 calculate_norms 2376 9.8 4.201 4.554 4.201 4.554 mp_allgather_i34 111 8.7 2.157 4.485 2.157 4.485 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.557 3.842 dbcsr_matrix_vector_mult_local 304 10.0 3.182 3.652 3.185 3.654 dbcsr_sort_data 658 11.4 3.087 3.367 3.087 3.367 ls_scf_post 1 4.0 0.000 0.001 3.245 3.251 dbcsr_special_finalize 555 9.7 0.006 0.007 2.830 3.136 dbcsr_merge_single_wm 555 10.7 0.532 0.655 2.821 3.128 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.989 3.061 ls_scf_store_result 1 5.0 0.000 0.000 2.993 3.049 dbcsr_data_release 10477 10.7 1.579 2.415 1.579 2.415 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.041 2.042 rebuild_ks_matrix 3 7.3 0.000 0.000 2.018 2.019 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.018 2.019 dbcsr_finalize 304 7.8 0.050 0.061 1.795 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.514000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 4.712940E+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 265558. MP_Allreduce 3049 15663. 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 0.055 0.085 98.879 98.892 qs_energies 1 2.0 0.000 0.000 98.035 98.039 ls_scf 1 3.0 0.000 0.000 96.050 96.059 dbcsr_multiply_generic 111 6.7 0.017 0.019 77.557 77.781 ls_scf_main 1 4.0 0.000 0.000 61.534 61.535 multiply_cannon 111 7.7 0.115 0.207 54.861 59.977 density_matrix_trs4 2 5.0 0.002 0.003 54.396 54.491 multiply_cannon_loop 111 8.7 0.069 0.078 50.311 51.683 mp_waitall_1 5436 11.0 26.093 31.798 26.093 31.798 ls_scf_init_scf 1 4.0 0.000 0.000 30.898 30.903 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.647 29.683 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.366 27.373 multiply_cannon_multrec 444 9.7 14.052 16.409 21.081 24.287 make_m2s 222 7.7 0.004 0.005 17.777 20.271 make_images 222 8.7 3.721 4.409 17.716 20.211 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.878 15.492 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.934 14.308 make_images_data 222 9.7 0.003 0.004 10.082 12.431 hybrid_alltoall_any 227 10.6 0.790 3.792 9.815 12.325 multiply_cannon_sync_h2d 444 9.7 6.546 9.017 6.546 9.017 dbcsr_mm_accdrv_process 3003 10.4 0.179 0.356 6.734 7.842 dbcsr_mm_accdrv_process_sort 3003 11.4 6.414 7.497 6.414 7.497 mp_allgather_i34 111 8.7 2.755 6.948 2.755 6.948 mp_sum_l 807 5.4 2.945 5.713 2.945 5.713 arnoldi_extremal 4 6.8 0.000 0.000 5.697 5.701 arnoldi_normal_ev 4 7.8 0.002 0.005 5.697 5.700 build_subspace 16 8.4 0.015 0.020 5.314 5.324 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.580 4.796 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.643 4.475 mp_irecv_dv 1241 11.2 1.625 4.450 1.625 4.450 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.046 4.356 dbcsr_matrix_vector_mult 304 9.0 0.011 0.022 4.075 4.280 dbcsr_matrix_vector_mult_local 304 10.0 3.636 4.087 3.638 4.089 calculate_norms 792 9.8 3.537 3.671 3.537 3.671 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.546 3.658 ls_scf_post 1 4.0 0.000 0.000 3.617 3.621 make_images_sizes 222 9.7 0.000 0.000 1.044 3.619 mp_alltoall_i44 222 10.7 1.044 3.619 1.044 3.619 ls_scf_store_result 1 5.0 0.000 0.000 3.392 3.445 dbcsr_finalize 304 7.8 0.062 0.078 2.202 2.285 dbcsr_merge_all 275 8.9 0.472 0.525 2.051 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.072 2.073 rebuild_ks_matrix 3 7.3 0.000 0.000 2.039 2.040 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.004 2.039 2.040 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.892000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3635.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 8.825467E+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 284111. MP_Allreduce 3043 21950. 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 0.094 0.140 107.894 107.901 qs_energies 1 2.0 0.000 0.000 106.463 106.476 ls_scf 1 3.0 0.000 0.000 103.532 103.543 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.401 77.505 ls_scf_main 1 4.0 0.000 0.000 65.662 65.664 density_matrix_trs4 2 5.0 0.002 0.003 56.501 56.562 multiply_cannon 111 7.7 0.202 0.289 49.907 51.827 multiply_cannon_loop 111 8.7 0.068 0.070 46.372 47.416 ls_scf_init_scf 1 4.0 0.000 0.000 34.160 34.161 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.569 32.579 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.769 29.780 mp_waitall_1 4527 11.1 22.059 25.906 22.059 25.906 make_m2s 222 7.7 0.005 0.005 23.811 24.814 make_images 222 8.7 4.604 5.008 23.705 24.706 multiply_cannon_multrec 444 9.7 17.894 18.607 22.489 23.189 make_images_data 222 9.7 0.003 0.004 13.063 15.464 hybrid_alltoall_any 227 10.6 1.665 3.639 12.880 15.448 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.421 11.033 multiply_cannon_sync_h2d 444 9.7 8.848 8.889 8.848 8.889 arnoldi_extremal 4 6.8 0.000 0.000 7.102 7.111 arnoldi_normal_ev 4 7.8 0.004 0.010 7.102 7.111 build_subspace 16 8.4 0.026 0.036 6.557 6.573 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.243 5.392 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.000 5.264 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.133 5.222 dbcsr_matrix_vector_mult_local 304 10.0 4.860 5.140 4.863 5.142 dbcsr_mm_accdrv_process 1814 10.4 0.257 0.324 4.432 4.566 dbcsr_mm_accdrv_process_sort 1814 11.4 4.129 4.269 4.129 4.269 ls_scf_post 1 4.0 0.000 0.000 3.709 3.721 mp_allgather_i34 111 8.7 1.138 3.575 1.138 3.575 ls_scf_store_result 1 5.0 0.000 0.000 3.414 3.425 make_images_sizes 222 9.7 0.000 0.000 1.403 3.422 mp_alltoall_i44 222 10.7 1.402 3.422 1.402 3.422 calculate_norms 792 9.8 3.238 3.281 3.238 3.281 dbcsr_finalize 304 7.8 0.082 0.089 3.095 3.169 dbcsr_merge_all 275 8.9 0.890 0.915 2.881 2.950 qs_energies_init_hamiltonians 1 3.0 0.007 0.016 2.901 2.902 dbcsr_data_release 12724 10.6 2.332 2.883 2.332 2.883 dbcsr_complete_redistribute 5 7.6 1.424 1.468 2.751 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.410 2.543 dbcsr_sort_data 325 11.1 2.450 2.508 2.450 2.508 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.483 2.484 rebuild_ks_matrix 3 7.3 0.000 0.000 2.417 2.419 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.006 2.417 2.418 dbcsr_new_transposed 4 7.5 0.254 0.273 2.282 2.296 dbcsr_frobenius_norm 74 6.6 2.056 2.140 2.203 2.242 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.203 dbcsr_add_anytype 103 7.2 0.860 0.893 2.131 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.901000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6810.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/2746062f82ececc9c9d99636991defc71e6cb498_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 590.667776E+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 76559. MP_Allreduce 13152 2094. MP_Sync 1064 MP_Alltoall 2588 926725822. 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 0.029 0.107 222.152 222.156 qs_mol_dyn_low 1 2.0 0.004 0.027 221.207 221.228 qs_forces 5 3.8 0.004 0.015 221.075 221.080 qs_energies 5 4.8 0.001 0.042 218.059 218.082 scf_env_do_scf 5 5.8 0.001 0.038 204.042 204.047 scf_env_do_scf_inner_loop 105 6.6 0.016 0.050 178.099 178.128 qs_scf_new_mos 105 7.6 0.001 0.001 137.800 137.997 qs_scf_loop_do_ot 105 8.6 0.001 0.009 137.800 137.996 dbcsr_multiply_generic 1445 12.2 0.126 0.136 130.399 130.801 ot_scf_mini 105 9.6 0.003 0.005 128.160 128.352 multiply_cannon 1445 13.2 0.275 0.286 111.961 114.307 multiply_cannon_loop 1445 14.2 2.377 2.511 110.245 111.593 velocity_verlet 4 3.0 0.007 0.019 105.380 105.382 ot_mini 105 10.6 0.001 0.001 59.129 59.262 multiply_cannon_multrec 69360 15.2 31.808 37.221 39.580 45.185 mp_waitall_1 488190 16.1 32.899 40.004 32.899 40.004 qs_ot_get_p 112 10.4 0.001 0.001 39.608 39.951 qs_ot_get_derivative 55 11.6 0.001 0.010 37.611 37.746 multiply_cannon_sync_h2d 69360 15.2 30.797 34.608 30.797 34.608 multiply_cannon_metrocomm3 69360 15.2 0.189 0.202 24.323 31.442 rebuild_ks_matrix 110 8.4 0.000 0.000 28.958 29.167 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.015 28.958 29.167 qs_ot_p2m_diag 40 11.0 0.020 0.030 29.032 29.119 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.561 26.754 cp_dbcsr_syevd 40 12.0 0.002 0.002 25.951 25.952 init_scf_loop 7 6.6 0.000 0.016 25.911 25.912 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.911 23.146 apply_single 62 13.6 0.000 0.000 22.911 23.146 prepare_preconditioner 7 7.6 0.000 0.001 20.979 21.008 make_preconditioner 7 8.6 0.000 0.002 20.979 21.008 cp_fm_syevd 40 13.0 0.000 0.000 20.855 20.996 ot_new_cg_direction 55 11.6 0.001 0.001 20.864 20.865 qs_rho_update_rho_low 110 7.6 0.001 0.003 17.196 17.489 calculate_rho_elec 110 8.6 0.030 0.034 17.196 17.488 cp_fm_redistribute_end 40 14.0 8.120 16.199 8.125 16.201 cp_fm_syevd_base 40 14.0 8.069 16.152 8.069 16.152 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.332 14.526 make_full_inverse_cholesky 7 9.6 0.000 0.000 13.994 14.059 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.841 13.919 rs_pw_transfer 690 11.5 0.020 0.037 11.428 12.855 density_rs2pw 110 9.6 0.006 0.007 11.307 12.743 mp_sum_l 4684 12.4 11.754 12.686 11.754 12.686 pw_transfer 1645 12.4 0.083 0.103 12.004 12.256 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 11.861 12.117 calculate_dm_sparse 110 9.5 0.000 0.001 11.250 11.390 init_scf_run 5 5.8 0.000 0.001 11.212 11.213 scf_env_initial_rho_setup 5 6.8 0.001 0.004 11.212 11.213 qs_vxc_create 110 10.4 0.002 0.006 10.381 10.425 fft_wrap_pw1pw2_240 915 15.0 1.197 1.302 10.050 10.297 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.922 9.981 cp_fm_cholesky_invert 7 10.6 9.832 9.841 9.832 9.841 check_diag 80 13.5 8.586 8.862 9.380 9.521 sum_up_and_integrate 60 10.3 0.028 0.031 8.622 8.633 integrate_v_rspace 60 11.3 0.002 0.003 8.594 8.605 dbcsr_mm_accdrv_process 154766 15.8 4.039 4.207 7.639 8.389 fft3d_pb 915 16.0 2.395 2.677 8.106 8.364 xc_rho_set_and_dset_create 110 12.4 0.076 0.102 7.247 7.501 xc_vxc_pw_create 60 11.3 0.039 0.050 6.975 7.019 make_m2s 2890 13.2 0.078 0.088 6.250 6.810 make_full_single_inverse 7 9.6 0.001 0.001 6.720 6.749 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 6.719 6.734 make_images 2890 14.2 0.241 0.260 6.144 6.703 acc_transpose_blocks 69360 15.2 0.360 0.381 6.242 6.581 multiply_cannon_metrocomm1 69360 15.2 0.093 0.099 4.130 6.436 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.346 6.415 calculate_first_density_matrix 1 7.0 0.001 0.002 6.376 6.388 xc_pw_derive 510 13.4 0.005 0.007 6.168 6.262 mp_waitany 7680 13.5 4.434 6.007 4.434 6.007 mp_alltoall_z22v 2340 17.7 5.158 5.414 5.158 5.414 multiply_cannon_metrocomm4 67915 15.2 0.181 0.194 2.067 4.946 potential_pw2rs 60 12.3 0.003 0.003 4.807 4.836 acc_transpose_blocks_kernels 69360 16.2 0.855 0.913 4.354 4.570 grid_collocate_task_list 110 9.6 4.083 4.558 4.083 4.558 wfi_extrapolate 5 7.8 0.000 0.001 4.509 4.509 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=222.156000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=560.400000, yerr=3.382307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 2746062f82ececc9c9d99636991defc71e6cb498 Summary: empty Status: OK