=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 03b41477b2e102c540934725bb0f08de7bd8b995 ################# 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 7.3.2, 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 (12.01.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/01 job id: 44538189 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/02 job id: 44538190 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/03 job id: 44538191 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/04 job id: 44538192 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/05 job id: 44538193 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/06 job id: 44538194 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/07 job id: 44538195 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/08 job id: 44538197 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/09 job id: 44538198 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/10 job id: 44538199 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/11 job id: 44538200 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/12 job id: 44538201 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/13 job id: 44538202 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/14 job id: 44538203 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/15 job id: 44538204 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/16 job id: 44538205 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/17 job id: 44538206 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/18 job id: 44538208 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/19 job id: 44538209 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/20 job id: 44538210 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/21 job id: 44538211 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/22 job id: 44538213 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/23 job id: 44538215 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/24 job id: 44538216 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/25 job id: 44538217 --- 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/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/26 job id: 44538218 --- 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 ~~~~~~~ === 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]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 0.047 0.049 137.196 137.197 farming_run 1 2.0 135.544 135.546 137.143 137.146 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458639E+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 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 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 0.012 0.027 116.081 116.081 qs_energies 1 2.0 0.000 0.000 115.846 115.846 mp2_main 1 3.0 0.000 0.000 113.967 113.968 mp2_gpw_main 1 4.0 0.019 0.026 113.143 113.144 mp2_ri_gpw_compute_in 1 5.0 0.172 0.187 93.936 94.387 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.006 55.447 55.900 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.168 41.764 47.152 get_2c_integrals 1 6.0 0.001 0.019 37.425 38.317 integrate_v_rspace 273 8.0 0.436 0.451 25.138 30.328 pw_transfer 6555 10.6 0.368 0.383 27.476 28.240 fft_wrap_pw1pw2 5465 11.4 0.044 0.048 26.115 26.766 grid_integrate_task_list 273 9.0 20.948 26.634 20.948 26.634 fft_wrap_pw1pw2_100 2178 12.4 1.211 1.393 23.635 24.267 compute_2c_integrals 1 7.0 0.003 0.003 19.588 19.588 rpa_ri_compute_en 1 5.0 0.122 0.140 19.102 19.344 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.970 19.320 mp2_eri_2c_integrate_gpw 1 9.0 2.374 2.428 18.967 19.319 cp_fm_cholesky_decompose 12 8.2 17.784 18.688 17.784 18.688 cholesky_decomp 1 7.0 0.000 0.000 16.691 17.584 fft3d_s 5443 13.4 16.182 16.724 16.203 16.746 ao_to_mo_and_store_B_mult_1 272 7.0 10.845 15.557 10.845 15.557 calculate_wavefunction 272 8.0 5.420 5.578 12.541 13.133 rpa_num_int 1 6.0 0.001 0.017 10.623 10.633 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.479 10.485 calc_mat_Q 8 8.0 0.000 0.000 9.338 9.415 contract_S_to_Q 8 9.0 0.000 0.000 8.759 8.836 calc_potential_gpw 544 9.5 0.006 0.006 8.245 8.670 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.259 8.542 potential_pw2rs 545 10.0 0.107 0.108 7.749 8.455 parallel_gemm_fm 14 9.1 0.000 0.000 8.345 8.414 parallel_gemm_fm_cosma 14 10.1 8.345 8.414 8.345 8.414 create_integ_mat 1 6.0 0.014 0.026 7.787 7.796 collocate_single_gaussian 272 10.0 0.039 0.042 7.520 7.763 array2fm 1 7.0 0.000 0.000 6.719 7.204 pw_scatter_s 2720 13.7 4.408 4.628 4.408 4.628 pw_gather_s 2722 13.2 3.901 4.190 3.901 4.190 array2fm_buffer_send 1 8.0 2.883 3.045 2.883 3.045 pw_poisson_solve 545 10.5 1.125 1.174 2.199 2.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.144111, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 0.028 0.040 398.286 398.287 farming_run 1 2.0 397.407 397.428 398.249 398.253 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.228251E+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 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 209.738 209.738 qs_energies 1 2.0 0.000 0.000 209.497 209.506 scf_env_do_scf 1 3.0 0.000 0.000 106.387 106.387 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.514 105.522 rebuild_ks_matrix 4 6.0 0.000 0.000 105.513 105.521 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 105.513 105.520 hfx_ks_matrix 4 8.0 0.001 0.001 105.131 105.134 integrate_four_center 4 9.0 0.145 0.503 105.130 105.134 mp2_main 1 3.0 0.000 0.000 102.816 102.826 mp2_gpw_main 1 4.0 0.031 0.047 101.944 101.955 integrate_four_center_main 4 10.0 0.095 0.503 96.901 99.671 integrate_four_center_bin 264 11.0 96.806 99.661 96.806 99.661 init_scf_loop 1 4.0 0.000 0.000 92.156 92.156 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.117 76.429 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.680 55.992 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.157 42.208 47.207 integrate_v_rspace 95 8.0 0.398 0.563 28.573 33.344 pw_transfer 2240 10.6 0.143 0.164 29.967 30.397 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.959 29.431 ao_to_mo_and_store_B_mult_1 91 7.0 10.786 29.142 10.786 29.142 grid_integrate_task_list 95 9.0 23.863 28.804 23.863 28.804 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 26.678 28.523 fft_wrap_pw1pw2_100 730 12.4 1.293 1.456 26.655 27.093 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.840 1.895 24.719 24.729 get_2c_integrals 1 6.0 0.000 0.000 20.350 20.373 compute_2c_integrals 1 7.0 0.002 0.003 19.340 19.343 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.930 19.218 mp2_eri_2c_integrate_gpw 1 9.0 1.723 1.881 18.929 19.217 fft3d_s 1823 13.4 18.437 18.756 18.450 18.769 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.230 14.230 calculate_wavefunction 91 8.0 2.025 2.051 9.749 9.982 potential_pw2rs 186 10.0 0.034 0.035 8.649 9.204 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.582 8.686 9.187 local_gemm 172 8.0 8.129 8.632 8.129 8.632 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.290 8.521 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.516 7.799 8.266 calc_potential_gpw 182 9.5 0.002 0.002 7.948 8.211 collocate_single_gaussian 91 10.0 0.017 0.022 7.908 8.060 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.443 6.349 6.443 mp_sendrecv_dm3 2068 8.0 5.833 6.318 5.833 6.318 pw_gather_s 912 13.2 4.915 5.422 4.915 5.422 mp_sync 38 10.4 2.822 4.847 2.822 4.847 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.939067, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.927488E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 848525. 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 0.026 0.049 55.193 55.194 qs_mol_dyn_low 1 2.0 0.003 0.004 54.895 54.903 qs_forces 11 3.9 0.002 0.002 54.535 54.536 qs_energies 11 4.9 0.001 0.001 53.052 53.063 scf_env_do_scf 11 5.9 0.000 0.001 46.243 46.243 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 44.158 44.158 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.258 35.779 qs_scf_new_mos 108 7.5 0.000 0.001 33.973 34.294 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.973 34.294 ot_scf_mini 108 9.5 0.002 0.002 32.215 32.450 multiply_cannon 2286 13.5 0.181 0.187 27.167 28.996 multiply_cannon_loop 2286 14.5 1.500 1.588 26.489 28.327 velocity_verlet 10 3.0 0.001 0.001 26.763 26.767 ot_mini 108 10.5 0.001 0.001 20.758 21.006 qs_ot_get_derivative 108 11.5 0.001 0.001 17.717 17.926 mp_waitall_1 245248 16.5 9.558 16.260 9.558 16.260 multiply_cannon_metrocomm3 54864 15.5 0.069 0.076 6.756 14.248 multiply_cannon_multrec 54864 15.5 4.166 6.370 7.629 11.080 rebuild_ks_matrix 119 8.3 0.000 0.000 7.943 8.109 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.943 8.109 mp_sum_l 7207 12.9 6.147 8.027 6.147 8.027 multiply_cannon_sync_h2d 54864 15.5 5.817 7.587 5.817 7.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.999 7.158 qs_ot_get_p 119 10.4 0.001 0.001 6.689 7.018 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.324 6.798 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.813 5.968 init_scf_run 11 5.9 0.000 0.001 5.569 5.569 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.569 5.569 dbcsr_mm_accdrv_process 76910 16.1 1.191 1.839 3.386 4.734 sum_up_and_integrate 119 10.3 0.012 0.015 4.538 4.544 integrate_v_rspace 119 11.3 0.002 0.002 4.526 4.532 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.135 4.464 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.200 4.308 calculate_rho_elec 119 8.7 0.011 0.017 4.199 4.307 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.764 3.880 calculate_first_density_matrix 1 7.0 0.000 0.000 3.380 3.383 calculate_dm_sparse 119 9.5 0.000 0.000 2.972 3.121 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.871 3.094 apply_single 119 13.6 0.000 0.000 2.871 3.094 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.870 2.871 rs_pw_transfer 974 11.9 0.011 0.012 2.727 2.844 jit_kernel_multiply 13 15.8 2.132 2.807 2.132 2.807 ot_diis_step 108 11.5 0.006 0.006 2.678 2.678 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.666 2.667 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.578 2.661 cp_fm_redistribute_end 50 14.0 2.427 2.645 2.433 2.647 cp_fm_diag_elpa_base 50 14.0 0.212 2.560 0.213 2.572 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.550 2.552 acc_transpose_blocks 54864 15.5 0.231 0.249 1.953 2.522 density_rs2pw 119 9.7 0.004 0.005 2.228 2.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.268 2.323 wfi_extrapolate 11 7.9 0.001 0.001 2.128 2.128 grid_integrate_task_list 119 12.3 2.020 2.112 2.020 2.112 init_scf_loop 11 6.9 0.000 0.000 2.067 2.067 mp_sum_d 4129 12.0 1.443 2.017 1.443 2.017 potential_pw2rs 119 12.3 0.004 0.004 1.884 1.901 pw_transfer 1439 11.6 0.053 0.059 1.809 1.883 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.733 1.809 make_m2s 4572 13.5 0.054 0.056 1.627 1.687 make_images 4572 14.5 0.132 0.138 1.545 1.604 fft3d_ps 1201 14.6 0.370 0.474 1.503 1.571 parallel_gemm_fm 81 9.0 0.000 0.000 1.542 1.548 parallel_gemm_fm_cosma 81 10.0 1.542 1.548 1.542 1.548 mp_alltoall_d11v 2130 13.8 1.269 1.441 1.269 1.441 fft_wrap_pw1pw2_140 487 13.2 0.082 0.096 1.333 1.411 mp_waitany 12084 13.8 1.258 1.404 1.258 1.404 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.378 1.397 acc_transpose_blocks_kernels 54864 16.5 0.254 0.386 0.957 1.377 grid_collocate_task_list 119 9.7 1.290 1.361 1.290 1.361 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.805 1.179 make_basis_sm 11 9.8 0.008 0.019 1.140 1.142 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.194000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 487.911424E+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 104 MP_Alltoall 2060 1030243. 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 0.042 0.132 41.098 41.116 qs_mol_dyn_low 1 2.0 0.004 0.007 40.652 40.659 qs_forces 11 3.9 0.008 0.051 40.533 40.539 qs_energies 11 4.9 0.001 0.005 38.287 38.294 scf_env_do_scf 11 5.9 0.001 0.001 31.770 31.771 scf_env_do_scf_inner_loop 108 6.5 0.003 0.012 29.294 29.295 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.643 23.023 qs_scf_new_mos 108 7.5 0.001 0.001 20.057 20.310 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.057 20.309 ot_scf_mini 108 9.5 0.003 0.003 19.181 19.362 multiply_cannon 2286 13.5 0.210 0.230 17.731 19.330 velocity_verlet 10 3.0 0.002 0.012 18.211 18.212 multiply_cannon_loop 2286 14.5 0.907 0.982 16.656 18.039 ot_mini 108 10.5 0.001 0.002 12.000 12.243 mp_waitall_1 200699 16.5 6.105 11.946 6.105 11.946 multiply_cannon_metrocomm3 27432 15.5 0.069 0.071 4.659 10.659 multiply_cannon_multrec 27432 15.5 1.966 4.597 6.858 10.209 qs_ot_get_derivative 108 11.5 0.014 0.318 9.543 9.729 rebuild_ks_matrix 119 8.3 0.000 0.000 7.748 7.885 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.748 7.884 dbcsr_mm_accdrv_process 47894 16.0 3.736 6.463 4.823 7.170 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.391 6.515 init_scf_run 11 5.9 0.000 0.001 5.092 5.093 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.092 5.093 sum_up_and_integrate 119 10.3 0.024 0.027 4.657 4.728 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.865 4.706 integrate_v_rspace 119 11.3 0.002 0.003 4.633 4.706 qs_ot_get_p 119 10.4 0.001 0.002 4.320 4.553 mp_sum_l 7207 12.9 2.112 4.128 2.112 4.128 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.034 4.119 apply_single 119 13.6 0.000 0.000 3.034 4.119 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.820 3.865 calculate_rho_elec 119 8.7 0.021 0.024 3.820 3.865 calculate_first_density_matrix 1 7.0 0.001 0.002 3.583 3.585 rs_pw_transfer 974 11.9 0.010 0.011 2.702 3.062 multiply_cannon_sync_h2d 27432 15.5 2.202 2.959 2.202 2.959 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.893 2.912 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.736 2.738 make_m2s 4572 13.5 0.052 0.053 2.453 2.679 density_rs2pw 119 9.7 0.004 0.004 2.137 2.641 make_images 4572 14.5 0.200 0.237 2.365 2.588 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.525 2.562 pw_transfer 1439 11.6 0.065 0.072 2.440 2.536 jit_kernel_multiply 9 16.2 1.034 2.470 1.034 2.470 potential_pw2rs 119 12.3 0.006 0.006 2.388 2.460 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.459 2.460 init_scf_loop 11 6.9 0.001 0.004 2.452 2.454 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.348 2.447 calculate_dm_sparse 119 9.5 0.000 0.000 2.369 2.443 ot_diis_step 108 11.5 0.011 0.015 2.407 2.408 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.118 2.213 fft3d_ps 1201 14.6 0.512 0.565 2.049 2.143 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.139 2.139 cp_fm_redistribute_end 50 14.0 1.771 2.116 1.774 2.117 cp_fm_diag_elpa_base 50 14.0 0.327 2.006 0.340 2.063 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.952 2.050 grid_integrate_task_list 119 12.3 1.836 1.945 1.836 1.945 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.597 1.604 prepare_preconditioner 11 7.9 0.000 0.000 1.526 1.552 make_preconditioner 11 8.9 0.000 0.001 1.526 1.552 make_images_data 4572 15.5 0.044 0.051 1.133 1.534 acc_transpose_blocks 27432 15.5 0.109 0.114 1.190 1.494 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.425 1.484 hybrid_alltoall_any 4725 16.4 0.050 0.110 0.979 1.465 wfi_extrapolate 11 7.9 0.001 0.001 1.457 1.457 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.415 1.429 grid_collocate_task_list 119 9.7 1.226 1.367 1.226 1.367 mp_alltoall_z22v 1201 16.6 1.152 1.321 1.152 1.321 mp_allgather_i34 2286 14.5 0.517 1.279 0.517 1.279 mp_alltoall_d11v 2130 13.8 1.174 1.268 1.174 1.268 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.152 1.199 mp_sum_d 4129 12.0 0.695 1.180 0.695 1.180 qs_energies_init_hamiltonians 11 5.9 0.002 0.013 1.114 1.115 yz_to_x 368 15.5 0.112 0.130 0.933 1.046 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.148 0.658 1.017 mp_waitany 5720 13.7 0.642 1.008 0.642 1.008 acc_transpose_blocks_kernels 27432 16.5 0.183 0.270 0.655 0.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.116000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 523.280384E+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 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 2597666. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 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 0.030 0.039 34.072 34.073 qs_mol_dyn_low 1 2.0 0.003 0.011 33.533 33.541 qs_forces 11 3.9 0.002 0.003 33.460 33.461 qs_energies 11 4.9 0.001 0.002 31.900 31.902 scf_env_do_scf 11 5.9 0.001 0.001 25.734 25.735 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.866 22.866 dbcsr_multiply_generic 2286 12.5 0.094 0.096 17.111 17.214 velocity_verlet 10 3.0 0.001 0.001 15.208 15.210 multiply_cannon 2286 13.5 0.202 0.223 13.927 14.732 qs_scf_new_mos 108 7.5 0.001 0.001 14.584 14.605 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.583 14.605 multiply_cannon_loop 2286 14.5 0.637 0.666 13.149 13.946 ot_scf_mini 108 9.5 0.002 0.002 13.866 13.882 ot_mini 108 10.5 0.001 0.001 8.605 8.619 multiply_cannon_multrec 18288 15.5 1.933 2.849 7.557 7.852 qs_ot_get_derivative 108 11.5 0.001 0.001 7.124 7.139 rebuild_ks_matrix 119 8.3 0.000 0.000 6.410 6.423 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.410 6.423 dbcsr_mm_accdrv_process 38222 16.0 4.664 6.292 5.540 6.386 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.665 5.676 init_scf_run 11 5.9 0.000 0.001 4.534 4.534 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.533 4.534 sum_up_and_integrate 119 10.3 0.029 0.030 3.974 3.981 integrate_v_rspace 119 11.3 0.002 0.002 3.944 3.954 mp_waitall_1 158411 16.6 2.536 3.764 2.536 3.764 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.514 3.521 calculate_rho_elec 119 8.7 0.030 0.031 3.514 3.520 qs_ot_get_p 119 10.4 0.001 0.001 3.432 3.457 calculate_first_density_matrix 1 7.0 0.000 0.000 3.327 3.328 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.685 3.306 init_scf_loop 11 6.9 0.000 0.000 2.853 2.853 jit_kernel_multiply 10 16.2 0.825 2.698 0.825 2.698 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.623 2.624 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.365 2.497 rs_pw_transfer 974 11.9 0.009 0.010 2.261 2.485 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.440 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.016 2.325 apply_single 119 13.6 0.000 0.000 2.016 2.325 density_rs2pw 119 9.7 0.004 0.004 2.031 2.251 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.092 2.098 pw_transfer 1439 11.6 0.066 0.071 1.996 2.005 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.902 1.915 grid_integrate_task_list 119 12.3 1.804 1.901 1.804 1.901 make_m2s 4572 13.5 0.044 0.045 1.728 1.882 calculate_dm_sparse 119 9.5 0.000 0.000 1.858 1.872 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.832 1.832 make_images 4572 14.5 0.190 0.203 1.644 1.796 potential_pw2rs 119 12.3 0.007 0.008 1.741 1.748 prepare_preconditioner 11 7.9 0.000 0.000 1.733 1.735 make_preconditioner 11 8.9 0.000 0.000 1.733 1.735 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.582 1.667 multiply_cannon_sync_h2d 18288 15.5 1.395 1.619 1.395 1.619 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.618 1.619 cp_fm_redistribute_end 50 14.0 1.207 1.599 1.208 1.599 fft3d_ps 1201 14.6 0.525 0.545 1.581 1.594 cp_fm_diag_elpa_base 50 14.0 0.375 1.520 0.390 1.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.532 1.540 fft_wrap_pw1pw2_140 487 13.2 0.091 0.094 1.524 1.537 ot_diis_step 108 11.5 0.011 0.011 1.461 1.461 acc_transpose_blocks 18288 15.5 0.075 0.077 1.383 1.406 mp_sum_l 7207 12.9 1.068 1.371 1.068 1.371 qs_energies_init_hamiltonians 11 5.9 0.022 0.026 1.352 1.355 grid_collocate_task_list 119 9.7 1.215 1.352 1.215 1.352 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.231 1.236 wfi_extrapolate 11 7.9 0.001 0.001 1.159 1.159 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.417 1.034 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.942 0.957 make_images_data 4572 15.5 0.044 0.048 0.769 0.946 qs_env_update_s_mstruct 11 6.9 0.365 0.373 0.887 0.930 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.861 0.886 hybrid_alltoall_any 4725 16.4 0.054 0.114 0.664 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.815 0.817 mp_sum_d 4129 12.0 0.413 0.797 0.413 0.797 mp_alltoall_d11v 2130 13.8 0.660 0.760 0.660 0.760 mp_alltoall_z22v 1201 16.6 0.659 0.757 0.659 0.757 jit_kernel_transpose 5 15.6 0.732 0.750 0.732 0.750 mp_waitany 9880 13.7 0.499 0.742 0.499 0.742 arnoldi_extremal 119 11.4 0.001 0.001 0.699 0.731 arnoldi_normal_ev 119 12.4 0.239 0.325 0.698 0.730 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.123 0.493 0.729 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.716 cp_fm_cholesky_invert 11 10.9 0.711 0.714 0.711 0.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.073000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.272727, yerr=2.561637 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 552.988672E+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 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 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 0.146 0.245 36.983 36.985 qs_mol_dyn_low 1 2.0 0.005 0.018 36.224 36.231 qs_forces 11 3.9 0.007 0.044 36.163 36.166 qs_energies 11 4.9 0.001 0.003 34.390 34.399 scf_env_do_scf 11 5.9 0.002 0.017 28.606 28.607 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.145 25.145 dbcsr_multiply_generic 2286 12.5 0.101 0.107 19.188 19.284 velocity_verlet 10 3.0 0.001 0.001 17.840 17.842 qs_scf_new_mos 108 7.5 0.001 0.001 16.531 16.581 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.530 16.581 multiply_cannon 2286 13.5 0.226 0.263 15.682 16.509 ot_scf_mini 108 9.5 0.003 0.004 15.598 15.649 multiply_cannon_loop 2286 14.5 0.936 0.962 14.747 15.377 multiply_cannon_multrec 27432 15.5 2.606 3.416 9.711 10.214 ot_mini 108 10.5 0.001 0.001 9.755 9.826 dbcsr_mm_accdrv_process 47916 15.9 5.770 7.758 7.012 8.558 qs_ot_get_derivative 108 11.5 0.001 0.001 7.964 8.017 rebuild_ks_matrix 119 8.3 0.000 0.000 6.675 6.726 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.675 6.725 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.919 5.962 init_scf_run 11 5.9 0.000 0.001 4.282 4.283 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.282 4.282 sum_up_and_integrate 119 10.3 0.035 0.038 3.871 3.880 integrate_v_rspace 119 11.3 0.002 0.003 3.836 3.845 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.636 3.669 calculate_rho_elec 119 8.7 0.040 0.046 3.636 3.668 init_scf_loop 11 6.9 0.001 0.003 3.426 3.427 qs_ot_get_p 119 10.4 0.001 0.002 3.265 3.351 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.997 3.345 calculate_first_density_matrix 1 7.0 0.000 0.002 2.899 2.900 jit_kernel_multiply 10 15.9 1.181 2.667 1.181 2.667 mp_waitall_1 137007 16.6 1.985 2.646 1.985 2.646 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.981 2.563 apply_single 119 13.6 0.000 0.000 1.980 2.563 prepare_preconditioner 11 7.9 0.000 0.000 2.543 2.552 make_preconditioner 11 8.9 0.000 0.001 2.543 2.552 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.146 2.472 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.360 2.362 rs_pw_transfer 974 11.9 0.009 0.009 2.123 2.321 make_m2s 4572 13.5 0.054 0.055 2.163 2.288 density_rs2pw 119 9.7 0.004 0.004 2.046 2.240 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.178 2.197 calculate_dm_sparse 119 9.5 0.000 0.000 2.131 2.190 pw_transfer 1439 11.6 0.066 0.073 2.144 2.181 make_images 4572 14.5 0.271 0.329 2.056 2.179 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.050 2.093 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.996 2.005 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.891 1.919 grid_integrate_task_list 119 12.3 1.815 1.908 1.815 1.908 fft3d_ps 1201 14.6 0.560 0.611 1.724 1.759 ot_diis_step 108 11.5 0.012 0.012 1.752 1.752 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.694 1.741 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.677 1.678 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.971 1.636 potential_pw2rs 119 12.3 0.008 0.009 1.619 1.622 mp_sum_l 7207 12.9 0.941 1.560 0.941 1.560 acc_transpose_blocks 27432 15.5 0.111 0.114 1.449 1.469 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.442 1.443 cp_fm_redistribute_end 50 14.0 0.955 1.419 0.956 1.420 cp_fm_diag_elpa_base 50 14.0 0.442 1.356 0.461 1.394 grid_collocate_task_list 119 9.7 1.220 1.338 1.220 1.338 wfi_extrapolate 11 7.9 0.001 0.001 1.334 1.334 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.318 1.328 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.189 1.194 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.092 1.110 cp_fm_upper_to_full 72 13.5 0.802 1.109 0.802 1.109 multiply_cannon_sync_h2d 27432 15.5 0.988 1.100 0.988 1.100 dbcsr_complete_redistribute 329 12.2 0.119 0.143 0.774 1.044 make_images_data 4572 15.5 0.045 0.048 0.839 0.968 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.726 0.907 mp_alltoall_d11v 2130 13.8 0.750 0.877 0.750 0.877 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.846 0.851 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 0.823 0.834 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.568 0.834 mp_alltoall_z22v 1201 16.6 0.747 0.791 0.747 0.791 cp_fm_cholesky_invert 11 10.9 0.752 0.755 0.752 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.985000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.363636, yerr=4.269825 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 599.072768E+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 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 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 0.052 0.094 29.658 29.659 qs_mol_dyn_low 1 2.0 0.003 0.004 29.330 29.337 qs_forces 11 3.9 0.002 0.002 29.272 29.273 qs_energies 11 4.9 0.001 0.001 27.540 27.544 scf_env_do_scf 11 5.9 0.000 0.001 21.723 21.723 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.172 19.173 velocity_verlet 10 3.0 0.001 0.001 14.423 14.426 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.804 12.880 qs_scf_new_mos 108 7.5 0.001 0.001 11.103 11.128 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.103 11.127 multiply_cannon 2286 13.5 0.232 0.239 10.326 10.773 ot_scf_mini 108 9.5 0.002 0.002 10.426 10.454 multiply_cannon_loop 2286 14.5 0.331 0.344 9.425 9.608 multiply_cannon_multrec 9144 15.5 1.669 1.920 6.246 6.472 rebuild_ks_matrix 119 8.3 0.000 0.000 6.101 6.121 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.100 6.121 ot_mini 108 10.5 0.001 0.001 5.948 5.981 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.434 5.452 qs_ot_get_derivative 108 11.5 0.001 0.001 4.662 4.689 dbcsr_mm_accdrv_process 12550 15.8 3.201 4.316 4.476 4.553 init_scf_run 11 5.9 0.000 0.001 4.334 4.334 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.333 4.333 sum_up_and_integrate 119 10.3 0.038 0.040 3.708 3.714 integrate_v_rspace 119 11.3 0.003 0.003 3.670 3.676 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.575 3.580 calculate_rho_elec 119 8.7 0.060 0.061 3.574 3.580 calculate_first_density_matrix 1 7.0 0.000 0.000 3.193 3.194 qs_ot_get_p 119 10.4 0.001 0.001 2.692 2.733 init_scf_loop 11 6.9 0.000 0.000 2.530 2.532 pw_transfer 1439 11.6 0.066 0.070 2.179 2.188 jit_kernel_multiply 10 15.6 1.238 2.182 1.238 2.182 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.085 2.097 density_rs2pw 119 9.7 0.004 0.004 1.906 2.050 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.913 1.914 grid_integrate_task_list 119 12.3 1.845 1.912 1.845 1.912 make_m2s 4572 13.5 0.034 0.035 1.720 1.887 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.828 1.843 rs_pw_transfer 974 11.9 0.008 0.008 1.683 1.821 mp_waitall_1 115863 16.7 1.338 1.813 1.338 1.813 calculate_dm_sparse 119 9.5 0.000 0.000 1.786 1.804 make_images 4572 14.5 0.270 0.303 1.631 1.795 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.779 1.781 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.758 1.766 fft3d_ps 1201 14.6 0.563 0.574 1.743 1.752 fft_wrap_pw1pw2_140 487 13.2 0.087 0.092 1.723 1.737 prepare_preconditioner 11 7.9 0.000 0.000 1.725 1.730 make_preconditioner 11 8.9 0.000 0.000 1.725 1.730 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.615 1.642 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.585 1.585 potential_pw2rs 119 12.3 0.010 0.011 1.440 1.443 grid_collocate_task_list 119 9.7 1.270 1.388 1.270 1.388 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.318 1.318 acc_transpose_blocks 9144 15.5 0.038 0.039 1.305 1.310 cp_fm_redistribute_end 50 14.0 0.659 1.300 0.660 1.301 cp_fm_diag_elpa_base 50 14.0 0.596 1.230 0.639 1.287 ot_diis_step 108 11.5 0.013 0.013 1.272 1.273 qs_energies_init_hamiltonians 11 5.9 0.027 0.046 1.249 1.270 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.202 1.207 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.145 1.174 apply_single 119 13.6 0.000 0.000 1.144 1.174 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.108 1.122 wfi_extrapolate 11 7.9 0.001 0.001 1.091 1.091 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 1.064 1.066 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.770 1.025 make_images_data 4572 15.5 0.038 0.042 0.797 0.999 parallel_gemm_fm 81 9.0 0.000 0.000 0.965 0.966 parallel_gemm_fm_cosma 81 10.0 0.965 0.966 0.965 0.966 jit_kernel_transpose 5 15.6 0.946 0.948 0.946 0.948 make_basis_sm 11 9.8 0.001 0.001 0.923 0.924 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.921 cp_fm_cholesky_invert 11 10.9 0.865 0.868 0.865 0.868 mp_alltoall_d11v 2130 13.8 0.752 0.841 0.752 0.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.788 0.790 multiply_cannon_sync_h2d 9144 15.5 0.707 0.788 0.707 0.788 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.728 0.737 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.674 0.725 mp_alltoall_z22v 1201 16.6 0.627 0.680 0.627 0.680 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.311 0.648 mp_allgather_i34 2286 14.5 0.234 0.639 0.234 0.639 yz_to_x 606 15.1 0.265 0.276 0.589 0.606 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.659000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.363636, yerr=4.940974 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 749.985792E+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 0.027 0.100 44.407 44.408 qs_mol_dyn_low 1 2.0 0.003 0.004 44.056 44.064 qs_forces 11 3.9 0.001 0.002 43.996 43.997 qs_energies 11 4.9 0.001 0.001 42.000 42.003 scf_env_do_scf 11 5.9 0.001 0.001 34.539 34.539 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.709 26.711 velocity_verlet 10 3.0 0.001 0.001 23.196 23.202 dbcsr_multiply_generic 2286 12.5 0.116 0.120 20.264 20.477 multiply_cannon 2286 13.5 0.314 0.327 16.213 17.379 qs_scf_new_mos 108 7.5 0.001 0.001 16.876 16.969 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.875 16.968 multiply_cannon_loop 2286 14.5 0.344 0.351 14.972 16.181 ot_scf_mini 108 9.5 0.002 0.002 15.802 15.899 multiply_cannon_multrec 9144 15.5 4.010 5.363 10.576 10.944 ot_mini 108 10.5 0.001 0.001 9.956 10.070 qs_ot_get_derivative 108 11.5 0.001 0.001 7.975 8.073 dbcsr_mm_accdrv_process 12550 15.8 4.049 5.595 6.443 7.995 init_scf_loop 11 6.9 0.000 0.000 7.804 7.809 rebuild_ks_matrix 119 8.3 0.000 0.000 7.386 7.590 qs_ks_build_kohn_sham_matrix 119 9.3 0.045 0.114 7.386 7.590 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.685 6.877 prepare_preconditioner 11 7.9 0.000 0.000 6.793 6.837 make_preconditioner 11 8.9 0.000 0.000 6.793 6.837 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.385 6.717 init_scf_run 11 5.9 0.000 0.001 5.404 5.404 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.403 5.403 cp_fm_upper_to_full 72 14.2 3.160 4.509 3.160 4.509 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.273 4.279 calculate_rho_elec 119 8.7 0.118 0.121 4.273 4.278 sum_up_and_integrate 119 10.3 0.065 0.067 4.055 4.060 integrate_v_rspace 119 11.3 0.003 0.003 3.989 3.994 calculate_first_density_matrix 1 7.0 0.000 0.000 3.962 3.964 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.627 3.628 mp_waitall_1 94719 16.7 2.475 3.606 2.475 3.606 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.853 3.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.370 3.392 qs_ot_get_p 119 10.4 0.001 0.001 3.141 3.274 jit_kernel_multiply 10 15.5 2.367 3.146 2.367 3.146 pw_transfer 1439 11.6 0.069 0.069 2.833 2.860 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.932 2.764 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.734 2.761 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.383 2.753 apply_single 119 13.6 0.000 0.000 2.382 2.753 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.422 2.468 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.623 2.442 make_m2s 4572 13.5 0.038 0.038 2.272 2.432 fft3d_ps 1201 14.6 0.599 0.627 2.358 2.384 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.307 2.334 make_images 4572 14.5 0.358 0.388 2.151 2.311 mp_alltoall_i22 627 13.8 1.394 2.231 1.394 2.231 density_rs2pw 119 9.7 0.004 0.004 2.208 2.222 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.402 2.215 calculate_dm_sparse 119 9.5 0.000 0.000 2.143 2.159 grid_integrate_task_list 119 12.3 2.056 2.073 2.056 2.073 mp_sum_l 7207 12.9 1.253 2.069 1.253 2.069 ot_diis_step 108 11.5 0.014 0.014 1.956 1.956 qs_ot_p2m_diag 50 11.0 0.043 0.043 1.917 1.919 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.796 1.796 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.659 1.708 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.646 1.646 acc_transpose_blocks 9144 15.5 0.038 0.038 1.399 1.569 grid_collocate_task_list 119 9.7 1.488 1.500 1.488 1.500 rs_pw_transfer 974 11.9 0.009 0.010 1.448 1.464 potential_pw2rs 119 12.3 0.014 0.014 1.457 1.459 cp_fm_cholesky_invert 11 10.9 1.396 1.399 1.396 1.399 wfi_extrapolate 11 7.9 0.001 0.001 1.373 1.373 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.360 1.360 cp_fm_diag_elpa_base 50 14.0 1.218 1.267 1.359 1.359 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.149 1.320 hybrid_alltoall_any 4725 16.4 0.092 0.189 1.111 1.278 make_images_data 4572 15.5 0.043 0.046 1.051 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.215 1.222 jit_kernel_transpose 5 15.6 1.032 1.204 1.032 1.204 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.131 1.149 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.130 1.148 mp_alltoall_d11v 2130 13.8 1.045 1.063 1.045 1.063 multiply_cannon_sync_h2d 9144 15.5 1.045 1.047 1.045 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.984 1.034 qs_create_task_list 11 7.9 0.001 0.001 0.964 0.980 generate_qs_task_list 11 8.9 0.371 0.390 0.964 0.980 yz_to_x 606 15.1 0.461 0.475 0.942 0.951 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.913 0.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.408000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.727273, yerr=11.848214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.698560E+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 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 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 0.177 0.521 86.324 86.325 qs_mol_dyn_low 1 2.0 0.003 0.004 85.489 85.499 qs_forces 11 3.9 0.002 0.004 85.420 85.421 qs_energies 11 4.9 0.001 0.001 82.583 82.595 scf_env_do_scf 11 5.9 0.000 0.001 71.135 71.137 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.661 65.661 dbcsr_multiply_generic 2055 12.4 0.115 0.125 54.116 54.429 qs_scf_new_mos 99 7.5 0.000 0.001 48.183 48.335 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.182 48.335 ot_scf_mini 99 9.5 0.002 0.002 45.771 45.880 multiply_cannon 2055 13.4 0.194 0.230 44.550 45.334 multiply_cannon_loop 2055 14.4 1.550 1.587 43.603 44.401 velocity_verlet 10 3.0 0.001 0.001 42.105 42.106 ot_mini 99 10.5 0.001 0.001 27.675 27.778 qs_ot_get_derivative 99 11.5 0.001 0.001 20.813 20.938 multiply_cannon_multrec 49320 15.4 12.154 12.888 18.878 19.710 rebuild_ks_matrix 110 8.3 0.000 0.001 14.611 14.746 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.011 14.610 14.745 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.829 12.964 mp_waitall_1 220248 16.4 11.455 12.424 11.455 12.424 multiply_cannon_sync_h2d 49320 15.4 9.988 10.588 9.988 10.588 qs_ot_get_p 110 10.4 0.001 0.001 9.592 9.716 init_scf_run 11 5.9 0.000 0.001 9.420 9.420 scf_env_initial_rho_setup 11 6.9 0.001 0.001 9.420 9.420 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 6.978 8.059 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.217 7.823 apply_single 110 13.6 0.000 0.001 7.217 7.823 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.056 7.573 sum_up_and_integrate 110 10.3 0.036 0.043 7.112 7.125 integrate_v_rspace 110 11.3 0.003 0.003 7.075 7.097 dbcsr_mm_accdrv_process 87628 16.1 2.085 2.171 6.603 6.907 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.639 6.750 calculate_rho_elec 110 8.6 0.021 0.025 6.639 6.749 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.687 6.739 ot_diis_step 99 11.5 0.005 0.006 6.575 6.576 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.336 6.372 mp_sum_l 6514 12.8 5.203 6.151 5.203 6.151 init_scf_loop 11 6.9 0.000 0.000 5.447 5.447 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.365 5.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.338 5.342 calculate_first_density_matrix 1 7.0 0.000 0.000 5.303 5.316 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.955 5.000 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.908 4.909 cp_fm_redistribute_end 48 14.0 4.286 4.886 4.289 4.887 cp_fm_diag_elpa_base 48 14.0 0.591 4.753 0.595 4.776 rs_pw_transfer 902 11.9 0.011 0.013 3.655 4.405 jit_kernel_multiply 13 15.9 4.235 4.293 4.235 4.293 calculate_dm_sparse 110 9.5 0.000 0.001 4.103 4.204 density_rs2pw 110 9.6 0.004 0.005 3.436 4.189 wfi_extrapolate 11 7.9 0.001 0.001 4.020 4.020 make_m2s 4110 13.4 0.060 0.065 3.891 4.008 make_images 4110 14.4 0.177 0.189 3.797 3.918 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.633 3.541 grid_integrate_task_list 110 12.3 3.239 3.389 3.239 3.389 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.260 3.334 pw_transfer 1331 11.6 0.054 0.065 3.244 3.308 prepare_preconditioner 11 7.9 0.000 0.000 3.282 3.299 make_preconditioner 11 8.9 0.000 0.000 3.282 3.299 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.156 3.222 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.052 3.092 fft_wrap_pw1pw2_140 451 13.1 0.171 0.191 2.699 2.770 fft3d_ps 1111 14.6 0.779 0.881 2.683 2.738 potential_pw2rs 110 12.3 0.006 0.007 2.629 2.647 acc_transpose_blocks 49320 15.4 0.224 0.234 2.509 2.587 mp_waitany 14300 13.8 1.815 2.586 1.815 2.586 mp_alltoall_d11v 2046 13.8 2.110 2.407 2.110 2.407 grid_collocate_task_list 110 9.6 2.080 2.312 2.080 2.312 mp_sum_d 3883 11.9 1.479 2.029 1.479 2.029 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.963 1.989 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.941 make_images_data 4110 15.4 0.042 0.044 1.785 1.907 parallel_gemm_fm 81 9.0 0.000 0.000 1.828 1.835 parallel_gemm_fm_cosma 81 10.0 1.828 1.835 1.828 1.835 hybrid_alltoall_any 4261 16.3 0.081 0.480 1.549 1.813 cp_fm_cholesky_invert 11 10.9 1.771 1.775 1.771 1.775 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.325000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.816384E+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 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 1388012. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 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 0.098 0.566 72.646 72.648 qs_mol_dyn_low 1 2.0 0.004 0.005 71.724 71.734 qs_forces 11 3.9 0.082 0.324 71.657 71.658 qs_energies 11 4.9 0.008 0.023 67.918 67.923 scf_env_do_scf 11 5.9 0.001 0.001 56.788 56.791 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.220 49.222 dbcsr_multiply_generic 2055 12.4 0.115 0.122 39.543 39.746 velocity_verlet 10 3.0 0.001 0.001 36.085 36.086 multiply_cannon 2055 13.4 0.223 0.247 32.860 34.228 qs_scf_new_mos 99 7.5 0.001 0.001 32.847 32.976 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.846 32.975 multiply_cannon_loop 2055 14.4 0.935 0.960 31.587 32.632 ot_scf_mini 99 9.5 0.003 0.004 31.199 31.316 ot_mini 99 10.5 0.001 0.001 18.367 18.496 multiply_cannon_multrec 24660 15.4 7.610 9.444 15.530 17.548 rebuild_ks_matrix 110 8.3 0.000 0.001 13.902 14.044 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.902 14.043 qs_ot_get_derivative 99 11.5 0.001 0.001 12.591 12.710 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.186 12.306 mp_waitall_1 176588 16.5 7.718 10.194 7.718 10.194 init_scf_run 11 5.9 0.000 0.001 8.632 8.633 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.632 8.633 dbcsr_mm_accdrv_process 52282 16.1 6.088 7.242 7.760 8.122 multiply_cannon_sync_h2d 24660 15.4 7.032 8.032 7.032 8.032 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.314 7.729 init_scf_loop 11 6.9 0.001 0.004 7.533 7.537 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.625 7.412 apply_single 110 13.6 0.000 0.001 6.625 7.412 sum_up_and_integrate 110 10.3 0.051 0.057 6.695 6.715 integrate_v_rspace 110 11.3 0.002 0.002 6.643 6.665 qs_ot_get_p 110 10.4 0.001 0.001 6.235 6.390 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.198 6.214 calculate_rho_elec 110 8.6 0.039 0.047 6.197 6.213 ot_diis_step 99 11.5 0.010 0.010 5.731 5.731 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.859 5.651 prepare_preconditioner 11 7.9 0.000 0.001 5.470 5.488 make_preconditioner 11 8.9 0.000 0.002 5.470 5.488 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.064 5.225 calculate_first_density_matrix 1 7.0 0.000 0.001 5.021 5.023 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 4.733 4.735 make_m2s 4110 13.4 0.057 0.059 4.138 4.612 make_images 4110 14.4 0.402 0.445 4.030 4.500 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.352 4.372 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.320 4.368 pw_transfer 1331 11.6 0.066 0.076 3.791 3.938 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.927 3.927 density_rs2pw 110 9.6 0.004 0.004 3.330 3.905 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.684 3.834 jit_kernel_multiply 11 16.5 1.323 3.812 1.323 3.812 rs_pw_transfer 902 11.9 0.012 0.013 2.990 3.619 wfi_extrapolate 11 7.9 0.001 0.001 3.521 3.521 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.406 3.408 cp_fm_redistribute_end 48 14.0 2.549 3.383 2.551 3.383 grid_integrate_task_list 110 12.3 3.157 3.358 3.157 3.358 cp_fm_diag_elpa_base 48 14.0 0.798 3.240 0.829 3.330 calculate_dm_sparse 110 9.5 0.001 0.001 3.302 3.328 fft_wrap_pw1pw2_140 451 13.1 0.203 0.222 3.164 3.319 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.135 3.198 fft3d_ps 1111 14.6 1.101 1.318 3.033 3.175 mp_sum_l 6514 12.8 2.308 2.966 2.308 2.966 make_images_data 4110 15.4 0.047 0.053 2.314 2.810 hybrid_alltoall_any 4261 16.3 0.101 0.443 2.031 2.774 cp_fm_cholesky_invert 11 10.9 2.517 2.523 2.517 2.523 potential_pw2rs 110 12.3 0.008 0.009 2.480 2.495 grid_collocate_task_list 110 9.6 2.053 2.474 2.053 2.474 mp_alltoall_d11v 2046 13.8 1.708 1.984 1.708 1.984 qs_energies_init_hamiltonians 11 5.9 0.025 0.030 1.964 1.969 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.931 1.952 mp_waitany 10164 13.8 1.259 1.917 1.259 1.917 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.881 1.898 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.885 1.897 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.995 1.620 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.784 1.614 acc_transpose_blocks 24660 15.4 0.111 0.114 1.563 1.606 mp_allgather_i34 2055 14.4 0.602 1.597 0.602 1.597 make_basis_sm 11 9.8 0.000 0.001 1.554 1.556 parallel_gemm_fm 81 9.0 0.000 0.000 1.519 1.527 parallel_gemm_fm_cosma 81 10.0 1.519 1.527 1.519 1.527 mp_irecv_dv 57340 16.2 0.662 1.483 0.662 1.483 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.473 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.648000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.363636, yerr=7.819239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 660.361216E+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 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 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 0.079 0.187 63.297 63.310 qs_mol_dyn_low 1 2.0 0.042 0.314 62.059 62.072 qs_forces 11 3.9 0.005 0.030 61.385 61.398 qs_energies 11 4.9 0.002 0.005 58.080 58.084 scf_env_do_scf 11 5.9 0.001 0.001 48.449 48.449 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 39.816 39.818 velocity_verlet 10 3.0 0.009 0.068 32.647 32.654 dbcsr_multiply_generic 2055 12.4 0.119 0.148 29.664 29.932 qs_scf_new_mos 99 7.5 0.001 0.001 24.585 24.675 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.584 24.675 multiply_cannon 2055 13.4 0.211 0.224 23.476 24.638 ot_scf_mini 99 9.5 0.003 0.004 23.370 23.475 multiply_cannon_loop 2055 14.4 0.617 0.633 22.284 23.351 ot_mini 99 10.5 0.001 0.001 13.409 13.517 rebuild_ks_matrix 110 8.3 0.000 0.000 12.525 12.643 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.024 12.524 12.642 multiply_cannon_multrec 16440 15.4 3.957 4.893 10.926 11.904 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.998 11.106 mp_waitall_1 139946 16.5 6.981 10.168 6.981 10.168 qs_ot_get_derivative 99 11.5 0.001 0.001 8.999 9.105 init_scf_loop 11 6.9 0.037 0.106 8.597 8.604 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.284 7.484 dbcsr_mm_accdrv_process 34862 16.1 5.671 6.679 6.824 7.122 init_scf_run 11 5.9 0.000 0.001 7.081 7.081 scf_env_initial_rho_setup 11 6.9 0.001 0.003 7.081 7.081 prepare_preconditioner 11 7.9 0.000 0.000 6.640 6.658 make_preconditioner 11 8.9 0.001 0.002 6.640 6.658 sum_up_and_integrate 110 10.3 0.060 0.062 6.633 6.655 integrate_v_rspace 110 11.3 0.003 0.003 6.573 6.595 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.979 6.344 qs_rho_update_rho_low 110 7.6 0.001 0.002 6.090 6.100 calculate_rho_elec 110 8.6 0.058 0.059 6.089 6.100 qs_ot_get_p 110 10.4 0.007 0.055 5.263 5.392 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.928 5.374 apply_single 110 13.6 0.000 0.000 4.927 5.373 make_m2s 4110 13.4 0.050 0.052 4.126 4.531 make_images 4110 14.4 0.393 0.513 4.012 4.413 density_rs2pw 110 9.6 0.004 0.004 3.188 4.389 ot_diis_step 99 11.5 0.011 0.014 4.383 4.383 multiply_cannon_sync_h2d 16440 15.4 3.693 4.383 3.693 4.383 calculate_first_density_matrix 1 7.0 0.001 0.002 4.001 4.002 rs_pw_transfer 902 11.9 0.010 0.011 2.747 3.946 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.839 3.840 pw_transfer 1331 11.6 0.066 0.073 3.809 3.819 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.702 3.715 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.042 3.669 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.618 3.622 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.465 3.508 grid_integrate_task_list 110 12.3 3.165 3.389 3.165 3.389 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.283 3.283 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.205 3.221 fft3d_ps 1111 14.6 1.088 1.097 2.999 3.012 wfi_extrapolate 11 7.9 0.002 0.010 2.990 2.990 calculate_dm_sparse 110 9.5 0.001 0.001 2.948 2.980 make_images_data 4110 15.4 0.043 0.048 2.364 2.851 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.056 2.824 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.781 2.782 cp_fm_redistribute_end 48 14.0 1.736 2.757 1.738 2.757 cp_fm_diag_elpa_base 48 14.0 0.959 2.613 1.016 2.718 cp_fm_cholesky_invert 11 10.9 2.616 2.622 2.616 2.622 mp_waitany 17072 13.8 1.238 2.482 1.238 2.482 grid_collocate_task_list 110 9.6 2.082 2.478 2.082 2.478 potential_pw2rs 110 12.3 0.011 0.011 2.353 2.364 multiply_cannon_metrocomm4 14385 15.4 0.044 0.047 0.866 2.348 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.292 2.347 mp_irecv_dv 48980 15.7 0.797 2.230 0.797 2.230 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.179 0.952 2.148 mp_alltoall_d11v 2046 13.8 1.744 2.126 1.744 2.126 jit_kernel_multiply 8 16.6 0.766 2.103 0.766 2.103 mp_sum_l 6514 12.8 1.487 2.080 1.487 2.080 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.032 2.034 dbcsr_complete_redistribute 325 12.2 0.330 0.386 1.507 2.008 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.912 1.925 cp_fm_upper_to_full 70 13.6 1.396 1.855 1.396 1.855 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.726 1.740 cp_fm_cholesky_decompose 22 10.9 1.547 1.564 1.547 1.564 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.062 1.551 mp_sum_d 3881 11.9 1.143 1.528 1.143 1.528 mp_allgather_i34 2055 14.4 0.510 1.527 0.510 1.527 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.503 multiply_cannon_metrocomm1 16440 15.4 0.029 0.030 0.431 1.406 acc_transpose_blocks 16440 15.4 0.072 0.074 1.291 1.302 rs_gather_matrices 110 12.3 0.234 0.262 0.977 1.272 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.310000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=8.288197 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 735.150080E+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 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 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 0.034 68.254 68.256 qs_mol_dyn_low 1 2.0 0.003 0.004 67.860 67.869 qs_forces 11 3.9 0.002 0.002 67.781 67.782 qs_energies 11 4.9 0.002 0.006 64.335 64.339 scf_env_do_scf 11 5.9 0.001 0.001 53.532 53.535 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.439 41.440 velocity_verlet 10 3.0 0.001 0.001 37.144 37.146 dbcsr_multiply_generic 2055 12.4 0.115 0.117 31.936 32.198 multiply_cannon 2055 13.4 0.243 0.257 24.942 26.360 qs_scf_new_mos 99 7.5 0.001 0.001 26.093 26.185 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.093 26.185 ot_scf_mini 99 9.5 0.003 0.004 24.479 24.570 multiply_cannon_loop 2055 14.4 0.885 0.904 23.534 24.146 multiply_cannon_multrec 24660 15.4 4.260 6.788 15.125 16.231 ot_mini 99 10.5 0.001 0.001 14.144 14.259 rebuild_ks_matrix 110 8.3 0.000 0.000 12.877 13.000 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.877 13.000 init_scf_loop 11 6.9 0.001 0.005 12.053 12.054 dbcsr_mm_accdrv_process 52304 16.0 7.382 11.113 10.719 11.641 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.427 11.539 qs_ot_get_derivative 99 11.5 0.001 0.001 10.026 10.125 prepare_preconditioner 11 7.9 0.000 0.000 9.832 9.848 make_preconditioner 11 8.9 0.000 0.002 9.832 9.848 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.066 9.525 init_scf_run 11 5.9 0.000 0.001 7.953 7.953 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.952 7.953 sum_up_and_integrate 110 10.3 0.066 0.069 6.636 6.649 integrate_v_rspace 110 11.3 0.003 0.003 6.569 6.585 mp_waitall_1 121746 16.5 4.463 6.478 4.463 6.478 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.247 6.256 calculate_rho_elec 110 8.6 0.077 0.081 6.246 6.255 make_m2s 4110 13.4 0.059 0.062 5.428 5.801 make_images 4110 14.4 0.576 0.699 5.288 5.657 qs_ot_get_p 110 10.4 0.001 0.001 5.318 5.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.247 5.249 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.891 4.930 calculate_first_density_matrix 1 7.0 0.000 0.002 4.858 4.861 cp_fm_upper_to_full 70 13.8 3.293 4.669 3.293 4.669 jit_kernel_multiply 10 15.6 3.007 4.295 3.007 4.295 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.005 4.107 apply_single 110 13.6 0.000 0.000 4.005 4.107 ot_diis_step 99 11.5 0.011 0.011 4.084 4.084 pw_transfer 1331 11.6 0.065 0.074 3.971 4.001 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.865 3.898 dbcsr_complete_redistribute 325 12.2 0.419 0.461 2.631 3.766 density_rs2pw 110 9.6 0.004 0.004 3.144 3.655 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.575 3.589 grid_integrate_task_list 110 12.3 3.253 3.438 3.253 3.438 multiply_cannon_sync_h2d 24660 15.4 3.176 3.368 3.176 3.368 fft_wrap_pw1pw2_140 451 13.1 0.203 0.213 3.321 3.355 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.254 3.299 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.103 3.232 hybrid_alltoall_any 4261 16.3 0.119 0.455 2.336 3.204 fft3d_ps 1111 14.6 1.095 1.131 3.154 3.178 make_images_data 4110 15.4 0.046 0.051 2.791 3.171 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.528 3.108 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.100 3.100 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.989 3.042 wfi_extrapolate 11 7.9 0.001 0.001 3.002 3.003 rs_pw_transfer 902 11.9 0.010 0.011 2.423 2.946 calculate_dm_sparse 110 9.5 0.001 0.001 2.906 2.940 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.756 2.868 mp_alltoall_i22 605 13.7 1.642 2.792 1.642 2.792 cp_fm_cholesky_invert 11 10.9 2.627 2.636 2.627 2.636 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.580 2.582 cp_fm_redistribute_end 48 14.0 1.291 2.555 1.293 2.556 cp_fm_diag_elpa_base 48 14.0 1.182 2.420 1.260 2.531 grid_collocate_task_list 110 9.6 2.179 2.459 2.179 2.459 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.305 2.305 potential_pw2rs 110 12.3 0.012 0.013 2.268 2.275 mp_alltoall_d11v 2046 13.8 1.815 2.055 1.815 2.055 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.680 1.714 cp_fm_cholesky_decompose 22 10.9 1.661 1.706 1.661 1.706 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.696 1.706 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.695 mp_allgather_i34 2055 14.4 0.506 1.640 0.506 1.640 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.827 1.577 mp_waitany 13376 13.8 1.067 1.554 1.067 1.554 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.527 1.542 acc_transpose_blocks 24660 15.4 0.106 0.108 1.500 1.525 mp_irecv_dv 62702 16.1 0.729 1.499 0.729 1.499 mp_sum_l 6514 12.8 0.883 1.428 0.883 1.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.256000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.454545, yerr=10.369217 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 839.872512E+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 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 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 0.016 0.032 59.793 59.794 qs_mol_dyn_low 1 2.0 0.004 0.008 59.492 59.500 qs_forces 11 3.9 0.002 0.003 58.923 58.926 qs_energies 11 4.9 0.001 0.001 55.229 55.235 scf_env_do_scf 11 5.9 0.000 0.001 43.530 43.531 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.557 35.558 velocity_verlet 10 3.0 0.001 0.001 31.719 31.753 dbcsr_multiply_generic 2055 12.4 0.105 0.106 26.248 26.399 multiply_cannon 2055 13.4 0.246 0.256 20.825 22.292 qs_scf_new_mos 99 7.5 0.001 0.001 20.255 20.291 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.254 20.291 multiply_cannon_loop 2055 14.4 0.323 0.335 19.475 19.835 ot_scf_mini 99 9.5 0.002 0.002 19.016 19.038 rebuild_ks_matrix 110 8.3 0.000 0.001 12.087 12.126 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.086 12.126 multiply_cannon_multrec 8220 15.4 3.264 4.577 10.734 11.638 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.736 10.770 ot_mini 99 10.5 0.001 0.001 10.313 10.328 init_scf_run 11 5.9 0.000 0.001 8.559 8.560 scf_env_initial_rho_setup 11 6.9 0.001 0.002 8.559 8.559 dbcsr_mm_accdrv_process 17442 15.9 3.144 4.344 7.341 8.328 init_scf_loop 11 6.9 0.000 0.000 7.928 7.930 mp_waitall_1 103326 16.6 5.929 7.394 5.929 7.394 sum_up_and_integrate 110 10.3 0.079 0.080 6.608 6.623 qs_ot_get_derivative 99 11.5 0.001 0.001 6.531 6.551 integrate_v_rspace 110 11.3 0.003 0.003 6.529 6.542 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.415 6.430 calculate_rho_elec 110 8.6 0.115 0.116 6.415 6.429 prepare_preconditioner 11 7.9 0.000 0.000 5.998 6.001 make_preconditioner 11 8.9 0.000 0.000 5.998 6.001 calculate_first_density_matrix 1 7.0 0.000 0.000 5.721 5.722 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.582 5.658 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.463 5.464 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.126 5.142 qs_ot_get_p 110 10.4 0.001 0.001 5.018 5.040 jit_kernel_multiply 10 15.7 3.891 4.973 3.891 4.973 make_m2s 4110 13.4 0.038 0.039 4.195 4.475 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.925 4.362 make_images 4110 14.4 0.639 0.697 4.065 4.348 pw_transfer 1331 11.6 0.066 0.071 4.190 4.200 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.083 4.096 ot_diis_step 99 11.5 0.012 0.012 3.762 3.762 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.685 3.720 apply_single 110 13.6 0.000 0.000 3.685 3.720 grid_integrate_task_list 110 12.3 3.350 3.580 3.350 3.580 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.547 3.561 density_rs2pw 110 9.6 0.004 0.004 3.148 3.522 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.464 3.468 fft3d_ps 1111 14.6 1.147 1.171 3.324 3.338 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.159 3.159 multiply_cannon_sync_h2d 8220 15.4 2.918 3.046 2.918 3.046 calculate_dm_sparse 110 9.5 0.001 0.001 3.000 3.043 cp_fm_cholesky_invert 11 10.9 2.893 2.897 2.893 2.897 make_images_data 4110 15.4 0.038 0.043 2.408 2.819 hybrid_alltoall_any 4261 16.3 0.200 0.863 2.343 2.787 wfi_extrapolate 11 7.9 0.001 0.001 2.742 2.742 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.693 2.696 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.661 2.662 cp_fm_redistribute_end 48 14.0 0.676 2.635 0.679 2.636 cp_fm_diag_elpa_base 48 14.0 1.771 2.448 1.947 2.609 rs_pw_transfer 902 11.9 0.010 0.011 2.183 2.580 grid_collocate_task_list 110 9.6 2.277 2.538 2.277 2.538 potential_pw2rs 110 12.3 0.015 0.016 2.143 2.151 mp_alltoall_d11v 2046 13.8 1.638 2.049 1.638 2.049 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.980 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.964 1.976 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.943 1.957 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.752 1.764 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.686 1.691 cp_fm_cholesky_decompose 22 10.9 1.674 1.688 1.674 1.688 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.529 1.656 mp_allgather_i34 2055 14.4 0.535 1.651 0.535 1.651 mp_waitany 9240 13.8 1.070 1.499 1.070 1.499 dbcsr_complete_redistribute 325 12.2 0.556 0.578 1.401 1.487 rs_gather_matrices 110 12.3 0.327 0.372 0.941 1.348 qs_create_task_list 11 7.9 0.001 0.001 1.229 1.329 generate_qs_task_list 11 8.9 0.379 0.445 1.228 1.329 acc_transpose_blocks 8220 15.4 0.036 0.037 1.237 1.252 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.801 1.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.794000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=786.454545, yerr=14.304250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 1.384006E+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 0.018 0.037 90.385 90.387 qs_mol_dyn_low 1 2.0 0.003 0.004 90.084 90.092 qs_forces 11 3.9 0.002 0.003 90.018 90.020 qs_energies 11 4.9 0.001 0.001 85.864 85.866 scf_env_do_scf 11 5.9 0.001 0.001 72.752 72.752 velocity_verlet 10 3.0 0.001 0.001 55.261 55.267 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.785 44.787 dbcsr_multiply_generic 2055 12.4 0.119 0.120 32.234 32.309 init_scf_loop 11 6.9 0.000 0.000 27.897 27.899 qs_scf_new_mos 99 7.5 0.001 0.001 26.425 26.465 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.424 26.465 multiply_cannon 2055 13.4 0.342 0.367 25.266 26.150 prepare_preconditioner 11 7.9 0.000 0.000 25.866 25.872 make_preconditioner 11 8.9 0.000 0.000 25.866 25.872 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.192 25.351 ot_scf_mini 99 9.5 0.002 0.002 24.685 24.704 multiply_cannon_loop 2055 14.4 0.342 0.344 23.507 24.009 cp_fm_upper_to_full 70 14.2 12.592 17.992 12.592 17.992 ot_mini 99 10.5 0.001 0.001 14.255 14.276 rebuild_ks_matrix 110 8.3 0.001 0.001 14.186 14.213 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.185 14.212 multiply_cannon_multrec 8220 15.4 4.911 5.108 13.333 13.630 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.855 12.881 dbcsr_complete_redistribute 325 12.2 1.017 1.038 7.218 10.308 qs_ot_get_derivative 99 11.5 0.001 0.001 9.799 9.820 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.199 9.287 init_scf_run 11 5.9 0.000 0.001 8.843 8.843 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.842 8.842 dbcsr_mm_accdrv_process 11614 15.7 3.444 3.943 8.279 8.729 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.658 8.725 mp_alltoall_i22 605 13.7 5.277 8.355 5.277 8.355 mp_waitall_1 84994 16.7 7.400 8.251 7.400 8.251 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.772 7.807 calculate_rho_elec 110 8.6 0.227 0.227 7.771 7.806 sum_up_and_integrate 110 10.3 0.150 0.151 7.372 7.386 integrate_v_rspace 110 11.3 0.004 0.004 7.223 7.236 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.638 5.639 make_m2s 4110 13.4 0.043 0.043 5.139 5.630 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.786 5.555 apply_single 110 13.6 0.000 0.000 4.786 5.554 calculate_first_density_matrix 1 7.0 0.000 0.000 5.445 5.445 make_images 4110 14.4 0.885 0.935 4.951 5.441 qs_ot_get_p 110 10.4 0.001 0.001 5.328 5.351 pw_transfer 1331 11.6 0.075 0.076 5.313 5.320 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.724 5.252 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.195 5.204 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.163 5.171 jit_kernel_multiply 10 15.2 4.635 5.155 4.635 5.155 cp_fm_cholesky_invert 11 10.9 5.104 5.108 5.104 5.108 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.531 4.541 ot_diis_step 99 11.5 0.015 0.016 4.439 4.439 fft3d_ps 1111 14.6 1.293 1.302 4.388 4.397 multiply_cannon_sync_h2d 8220 15.4 3.947 3.951 3.947 3.951 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.867 3.874 density_rs2pw 110 9.6 0.004 0.004 3.814 3.852 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.090 3.730 grid_integrate_task_list 110 12.3 3.658 3.714 3.658 3.714 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.683 3.685 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.820 3.547 calculate_dm_sparse 110 9.5 0.001 0.001 3.486 3.506 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.433 3.433 make_images_data 4110 15.4 0.041 0.044 2.794 3.403 wfi_extrapolate 11 7.9 0.001 0.001 3.276 3.276 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.857 2.857 cp_fm_diag_elpa_base 48 14.0 2.328 2.517 2.855 2.855 grid_collocate_task_list 110 9.6 2.630 2.655 2.630 2.655 potential_pw2rs 110 12.3 0.021 0.021 2.497 2.499 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.322 2.335 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.217 2.296 rs_pw_transfer 902 11.9 0.011 0.011 2.249 2.283 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.083 2.157 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.037 2.052 mp_alltoall_d11v 2046 13.8 1.896 1.991 1.896 1.991 cp_fm_cholesky_decompose 22 10.9 1.950 1.964 1.950 1.964 qs_create_task_list 11 7.9 0.001 0.001 1.900 1.948 generate_qs_task_list 11 8.9 0.735 0.789 1.900 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.935 1.937 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.387000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.818182, yerr=73.662397 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.239808E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56760. MP_Allreduce 11253 785. MP_Sync 170 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 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 0.023 0.079 208.994 208.995 qs_mol_dyn_low 1 2.0 0.004 0.005 208.509 208.523 qs_forces 11 3.9 0.003 0.003 206.925 206.926 qs_energies 11 4.9 0.001 0.002 201.305 201.322 scf_env_do_scf 11 5.9 0.001 0.001 182.416 182.420 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.011 162.013 dbcsr_multiply_generic 2507 12.6 0.179 0.183 126.102 126.713 velocity_verlet 10 3.0 0.001 0.002 125.510 125.527 qs_scf_new_mos 117 7.6 0.001 0.001 120.423 120.671 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.423 120.670 ot_scf_mini 117 9.6 0.003 0.003 113.807 114.052 multiply_cannon 2507 13.6 0.237 0.243 102.940 104.573 multiply_cannon_loop 2507 14.6 2.114 2.178 100.817 102.621 ot_mini 117 10.6 0.001 0.001 65.291 65.582 multiply_cannon_multrec 60168 15.6 33.142 35.627 43.932 46.126 qs_ot_get_derivative 117 11.6 0.001 0.001 40.454 40.694 rebuild_ks_matrix 128 8.3 0.001 0.001 33.903 34.332 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.903 34.331 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.455 30.815 mp_waitall_1 267128 16.5 28.155 30.633 28.155 30.633 multiply_cannon_sync_h2d 60168 15.6 27.206 29.082 27.206 29.082 qs_ot_get_p 128 10.4 0.001 0.001 26.702 27.006 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.255 25.006 apply_single 128 13.6 0.001 0.001 24.254 25.005 ot_diis_step 117 11.6 0.008 0.008 24.457 24.458 init_scf_loop 11 6.9 0.000 0.000 20.329 20.329 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.945 20.054 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.831 18.994 multiply_cannon_metrocomm3 60168 15.6 0.114 0.120 15.504 17.521 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.355 17.356 prepare_preconditioner 11 7.9 0.000 0.000 15.730 15.790 make_preconditioner 11 8.9 0.000 0.000 15.730 15.790 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.967 15.137 init_scf_run 11 5.9 0.000 0.001 14.749 14.749 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.748 14.749 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.559 14.564 cp_fm_redistribute_end 83 14.4 11.517 14.496 11.527 14.497 sum_up_and_integrate 128 10.3 0.089 0.108 14.415 14.429 make_m2s 5014 13.6 0.104 0.112 13.888 14.379 integrate_v_rspace 128 11.3 0.003 0.004 14.326 14.345 cp_fm_diag_elpa_base 83 14.4 2.927 14.120 2.958 14.243 make_images 5014 14.6 0.399 0.419 13.707 14.204 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.046 14.144 calculate_rho_elec 128 8.7 0.045 0.064 14.045 14.143 dbcsr_mm_accdrv_process 124484 16.2 3.316 3.454 10.355 10.889 density_rs2pw 128 9.7 0.006 0.007 7.364 10.557 mp_sum_l 7870 13.0 8.037 9.467 8.037 9.467 rs_pw_transfer 1046 11.9 0.016 0.018 5.902 9.149 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 9.068 9.077 wfi_extrapolate 11 7.9 0.001 0.001 8.993 8.993 cp_fm_cholesky_invert 11 10.9 8.901 8.909 8.901 8.909 calculate_dm_sparse 128 9.5 0.001 0.001 8.507 8.595 pw_transfer 1547 11.6 0.076 0.097 8.138 8.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.236 8.324 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.933 8.195 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.800 7.933 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.765 7.861 make_images_data 5014 15.6 0.065 0.071 6.790 7.730 grid_integrate_task_list 128 12.3 7.064 7.614 7.064 7.614 hybrid_alltoall_any 5200 16.5 0.288 2.256 5.949 7.523 multiply_cannon_metrocomm1 60168 15.6 0.088 0.096 6.081 7.392 fft_wrap_pw1pw2_140 523 13.2 0.444 0.510 6.894 7.202 fft3d_ps 1291 14.7 2.106 2.891 6.721 6.893 grid_collocate_task_list 128 9.7 4.554 5.851 4.554 5.851 mp_waitany 16020 13.9 2.663 5.841 2.663 5.841 calculate_first_density_matrix 1 7.0 0.000 0.000 5.587 5.594 mp_alltoall_d11v 2415 14.1 4.302 5.560 4.302 5.560 rs_pw_transfer_RS2PW_140 139 11.5 0.277 0.290 2.132 5.355 jit_kernel_multiply 13 15.9 4.975 5.055 4.975 5.055 potential_pw2rs 128 12.3 0.009 0.011 4.778 4.804 cp_fm_cholesky_decompose 22 10.9 4.647 4.665 4.647 4.665 mp_sum_d 4464 12.1 3.597 4.301 3.597 4.301 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.995000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.181818, yerr=7.132744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 826.793984E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5618999. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 211.868 211.869 qs_mol_dyn_low 1 2.0 0.004 0.004 211.513 211.526 qs_forces 11 3.9 0.003 0.007 194.022 194.023 qs_energies 11 4.9 0.001 0.002 187.198 187.211 scf_env_do_scf 11 5.9 0.001 0.001 167.827 167.837 velocity_verlet 10 3.0 0.001 0.001 137.148 137.513 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.789 134.792 dbcsr_multiply_generic 2507 12.6 0.186 0.191 101.377 102.568 qs_scf_new_mos 117 7.6 0.001 0.001 94.383 94.940 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.383 94.939 ot_scf_mini 117 9.6 0.004 0.004 89.597 90.091 multiply_cannon 2507 13.6 0.479 0.524 80.835 85.180 multiply_cannon_loop 2507 14.6 1.268 1.310 77.525 80.585 ot_mini 117 10.6 0.001 0.001 50.579 51.155 multiply_cannon_multrec 30084 15.6 22.105 26.953 34.822 39.887 mp_waitall_1 214728 16.6 24.629 39.668 24.629 39.668 rebuild_ks_matrix 128 8.3 0.001 0.001 33.647 34.072 qs_ks_build_kohn_sham_matrix 128 9.3 0.089 0.173 33.646 34.071 init_scf_loop 11 6.9 0.000 0.000 32.946 32.947 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.286 30.692 qs_ot_get_derivative 117 11.6 0.001 0.002 28.672 29.180 multiply_cannon_metrocomm3 30084 15.6 0.095 0.103 15.857 29.071 prepare_preconditioner 11 7.9 0.000 0.000 28.421 28.556 make_preconditioner 11 8.9 0.000 0.000 28.421 28.556 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.115 27.664 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.644 23.760 apply_single 128 13.6 0.001 0.001 22.644 23.760 ot_diis_step 117 11.6 0.014 0.015 21.738 21.740 multiply_cannon_sync_h2d 30084 15.6 19.420 21.726 19.420 21.726 qs_ot_get_p 128 10.4 0.001 0.001 20.717 21.270 mp_sum_dm 438 4.9 17.198 17.625 17.198 17.625 md_output 10 3.0 0.000 0.000 0.369 17.478 md_write_output 11 3.9 0.008 0.342 0.365 17.475 update_particle_set 20 4.0 0.000 0.000 17.106 17.470 write_trajectory 44 4.9 0.001 0.006 0.357 17.133 write_particle_coordinates 11 5.9 0.357 17.126 0.357 17.126 cp_fm_cholesky_invert 11 10.9 16.639 16.652 16.639 16.652 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.949 15.981 make_m2s 5014 13.6 0.089 0.095 14.252 15.900 make_images 5014 14.6 1.154 1.344 14.042 15.690 sum_up_and_integrate 128 10.3 0.114 0.131 14.858 14.890 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.791 14.793 integrate_v_rspace 128 11.3 0.004 0.009 14.744 14.780 init_scf_run 11 5.9 0.000 0.001 14.599 14.601 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.599 14.600 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.218 14.253 calculate_rho_elec 128 8.7 0.088 0.105 14.217 14.252 dbcsr_mm_accdrv_process 62242 16.2 5.225 8.128 12.168 12.966 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.748 11.755 cp_fm_redistribute_end 83 14.4 6.864 11.690 6.878 11.694 cp_fm_diag_elpa_base 83 14.4 4.583 11.187 4.800 11.558 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.890 11.260 multiply_cannon_metrocomm4 27577 15.6 0.094 0.106 3.700 10.688 make_images_data 5014 15.6 0.063 0.072 8.711 10.646 mp_irecv_dv 69486 16.3 3.510 10.297 3.510 10.297 hybrid_alltoall_any 5200 16.5 0.338 1.491 7.404 9.945 density_rs2pw 128 9.7 0.006 0.007 7.585 9.927 pw_transfer 1547 11.6 0.088 0.113 9.070 9.143 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.997 9.005 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 8.843 8.917 wfi_extrapolate 11 7.9 0.001 0.001 8.451 8.451 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 8.131 8.250 rs_pw_transfer 1046 11.9 0.014 0.016 5.719 8.078 fft_wrap_pw1pw2_140 523 13.2 0.473 0.526 7.812 7.898 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.491 7.816 grid_integrate_task_list 128 12.3 7.198 7.623 7.198 7.623 fft3d_ps 1291 14.7 2.746 2.909 7.208 7.256 cp_fm_cholesky_decompose 22 10.9 6.800 6.891 6.800 6.891 calculate_dm_sparse 128 9.5 0.001 0.001 6.665 6.828 mp_sum_l 7870 13.0 4.506 6.465 4.506 6.465 calculate_first_density_matrix 1 7.0 0.000 0.002 5.970 5.972 grid_collocate_task_list 128 9.7 4.703 5.855 4.703 5.855 jit_kernel_multiply 12 15.9 3.949 5.670 3.949 5.670 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.297 5.355 potential_pw2rs 128 12.3 0.015 0.018 5.086 5.102 mp_allgather_i34 2507 14.6 1.883 4.969 1.883 4.969 mp_waitany 11748 13.9 2.472 4.821 2.472 4.821 mp_alltoall_d11v 2415 14.1 4.122 4.677 4.122 4.677 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.374 2.071 4.407 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=211.869000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.000000, yerr=1.954017 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.951022E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 948.527104E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56602. MP_Allreduce 11361 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_comm_split 84 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.127 0.260 183.230 183.231 qs_mol_dyn_low 1 2.0 0.004 0.005 182.593 182.606 qs_forces 11 3.9 0.007 0.037 180.988 180.995 qs_energies 11 4.9 0.002 0.010 174.322 174.331 scf_env_do_scf 11 5.9 0.048 0.381 156.273 156.274 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 120.442 120.443 velocity_verlet 10 3.0 0.003 0.021 116.315 116.364 dbcsr_multiply_generic 2529 12.6 0.183 0.188 84.770 85.991 qs_scf_new_mos 118 7.6 0.001 0.001 81.326 81.617 qs_scf_loop_do_ot 118 8.6 0.001 0.002 81.326 81.616 ot_scf_mini 118 9.6 0.004 0.005 77.156 77.505 multiply_cannon 2529 13.6 0.508 0.526 64.890 69.641 multiply_cannon_loop 2529 14.6 0.866 0.894 61.637 64.432 ot_mini 118 10.6 0.001 0.001 42.815 43.153 init_scf_loop 11 6.9 0.049 0.392 35.352 35.353 mp_waitall_1 172006 16.6 25.177 34.949 25.177 34.949 rebuild_ks_matrix 129 8.3 0.001 0.001 31.373 31.836 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.019 31.372 31.835 prepare_preconditioner 11 7.9 0.000 0.001 30.874 30.922 make_preconditioner 11 8.9 0.001 0.005 30.874 30.922 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.522 29.932 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.287 28.713 multiply_cannon_multrec 20232 15.6 13.577 16.799 24.795 28.005 multiply_cannon_metrocomm3 20232 15.6 0.061 0.064 15.366 25.674 qs_ot_get_derivative 118 11.6 0.001 0.002 23.035 23.384 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.907 20.925 apply_single 129 13.6 0.001 0.001 19.907 20.925 qs_ot_get_p 129 10.4 0.001 0.001 19.585 20.002 ot_diis_step 118 11.6 0.018 0.018 19.674 19.674 multiply_cannon_sync_h2d 20232 15.6 14.400 16.090 14.400 16.090 make_m2s 5058 13.6 0.081 0.086 14.776 16.029 make_images 5058 14.6 1.188 1.288 14.541 15.792 qs_ot_p2m_diag 84 11.4 0.269 0.275 15.278 15.284 sum_up_and_integrate 129 10.3 0.132 0.144 14.910 14.939 integrate_v_rspace 129 11.3 0.004 0.005 14.778 14.805 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.670 14.695 calculate_rho_elec 129 8.7 0.133 0.148 14.669 14.694 cp_fm_cholesky_invert 11 10.9 14.446 14.455 14.446 14.455 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.272 14.274 init_scf_run 11 5.9 0.000 0.001 13.023 13.023 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.022 13.023 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.092 11.093 cp_fm_redistribute_end 84 14.4 4.193 11.034 4.208 11.036 cp_fm_diag_elpa_base 84 14.4 6.398 10.441 6.807 10.917 dbcsr_mm_accdrv_process 41846 16.2 4.929 7.751 10.665 10.792 make_images_data 5058 15.6 0.061 0.068 9.056 10.745 hybrid_alltoall_any 5245 16.5 0.436 1.993 7.965 9.956 density_rs2pw 129 9.7 0.006 0.006 7.416 9.540 pw_transfer 1559 11.6 0.087 0.109 9.334 9.450 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.060 9.318 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.107 9.228 multiply_cannon_metrocomm4 17703 15.6 0.061 0.070 3.446 9.221 mp_irecv_dv 50659 16.2 3.324 8.975 3.324 8.975 fft_wrap_pw1pw2_140 527 13.2 0.481 0.524 8.057 8.184 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.942 7.948 grid_integrate_task_list 129 12.3 7.333 7.881 7.333 7.881 cp_fm_upper_to_full 106 14.5 5.790 7.512 5.790 7.512 wfi_extrapolate 11 7.9 0.001 0.001 7.485 7.485 fft3d_ps 1301 14.7 2.688 2.922 7.355 7.445 rs_pw_transfer 1054 12.0 0.014 0.015 5.182 7.328 cp_fm_cholesky_decompose 22 10.9 7.269 7.312 7.269 7.312 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 7.068 7.196 dbcsr_complete_redistribute 397 12.7 1.166 1.197 4.907 6.713 calculate_dm_sparse 129 9.5 0.001 0.001 5.814 5.927 grid_collocate_task_list 129 9.7 4.947 5.713 4.947 5.713 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.749 5.549 mp_alltoall_d11v 2429 14.1 4.437 5.417 4.437 5.417 calculate_first_density_matrix 1 7.0 0.001 0.004 5.345 5.348 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.598 5.298 potential_pw2rs 129 12.3 0.020 0.023 4.892 4.904 mp_sum_l 7936 13.1 3.344 4.718 3.344 4.718 mp_alltoall_i22 720 14.1 2.611 4.646 2.611 4.646 mp_allgather_i34 2529 14.6 1.763 4.581 1.763 4.581 mp_waitany 11836 13.9 2.359 4.541 2.359 4.541 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.332 4.062 rs_pw_transfer_RS2PW_140 140 11.5 0.333 0.353 1.935 4.061 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.022 4.045 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.787 3.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=183.231000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.727273, yerr=12.990142 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.140343E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56895. MP_Allreduce 11306 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 83 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 0.081 0.161 191.411 191.414 qs_mol_dyn_low 1 2.0 0.005 0.017 190.842 190.855 qs_forces 11 3.9 0.003 0.008 190.720 190.725 qs_energies 11 4.9 0.003 0.008 183.440 183.450 scf_env_do_scf 11 5.9 0.001 0.002 164.333 164.340 velocity_verlet 10 3.0 0.001 0.001 126.030 126.033 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 117.810 117.831 dbcsr_multiply_generic 2527 12.6 0.191 0.197 82.060 82.812 qs_scf_new_mos 118 7.6 0.001 0.001 79.573 79.932 qs_scf_loop_do_ot 118 8.6 0.001 0.002 79.572 79.931 ot_scf_mini 118 9.6 0.004 0.004 75.063 75.395 multiply_cannon 2527 13.6 0.560 0.589 57.167 60.151 multiply_cannon_loop 2527 14.6 1.190 1.226 53.570 55.423 init_scf_loop 11 6.9 0.001 0.002 46.399 46.462 ot_mini 118 10.6 0.001 0.001 42.737 43.054 prepare_preconditioner 11 7.9 0.000 0.000 42.202 42.233 make_preconditioner 11 8.9 0.000 0.003 42.202 42.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.821 40.865 multiply_cannon_multrec 30324 15.6 14.387 19.154 28.150 32.658 rebuild_ks_matrix 129 8.3 0.001 0.001 30.476 30.744 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.475 30.744 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.416 27.651 mp_waitall_1 149060 16.7 17.527 27.500 17.527 27.500 qs_ot_get_derivative 118 11.6 0.001 0.002 23.050 23.380 make_m2s 5054 13.6 0.096 0.100 20.413 21.477 make_images 5054 14.6 1.967 2.267 20.104 21.169 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.108 19.662 apply_single 129 13.6 0.001 0.001 19.107 19.662 ot_diis_step 118 11.6 0.018 0.019 19.563 19.565 qs_ot_get_p 129 10.4 0.001 0.002 18.568 18.866 cp_fm_upper_to_full 105 14.7 11.123 16.417 11.123 16.417 cp_fm_cholesky_invert 11 10.9 16.053 16.062 16.053 16.062 multiply_cannon_metrocomm3 30324 15.6 0.048 0.050 6.453 15.315 sum_up_and_integrate 129 10.3 0.141 0.154 15.050 15.092 integrate_v_rspace 129 11.3 0.004 0.005 14.908 14.953 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.890 14.920 calculate_rho_elec 129 8.7 0.177 0.192 14.890 14.920 qs_ot_p2m_diag 83 11.4 0.343 0.390 14.371 14.424 dbcsr_mm_accdrv_process 62756 16.2 7.427 8.611 13.331 13.817 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.102 13.103 init_scf_run 11 5.9 0.000 0.001 12.931 12.932 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.930 12.932 dbcsr_complete_redistribute 395 12.7 1.504 1.627 8.921 12.762 make_images_data 5054 15.6 0.064 0.069 11.061 12.692 multiply_cannon_sync_h2d 30324 15.6 11.782 12.649 11.782 12.649 hybrid_alltoall_any 5240 16.5 0.529 2.219 9.794 11.733 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.557 11.388 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.358 10.106 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.089 10.091 cp_fm_redistribute_end 83 14.4 1.743 10.024 1.755 10.030 cp_fm_diag_elpa_base 83 14.4 7.663 9.424 8.239 9.907 pw_transfer 1559 11.6 0.089 0.107 9.677 9.746 mp_alltoall_i22 718 14.1 5.620 9.596 5.620 9.596 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.366 9.596 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.447 9.524 density_rs2pw 129 9.7 0.006 0.006 7.229 8.742 fft_wrap_pw1pw2_140 527 13.2 0.487 0.501 8.393 8.495 grid_integrate_task_list 129 12.3 7.542 7.997 7.542 7.997 wfi_extrapolate 11 7.9 0.001 0.001 7.729 7.729 fft3d_ps 1301 14.7 2.789 2.832 7.655 7.711 cp_fm_cholesky_decompose 22 10.9 7.548 7.633 7.548 7.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.201 7.260 multiply_cannon_metrocomm4 25270 15.6 0.076 0.085 2.723 6.699 calculate_dm_sparse 129 9.5 0.001 0.001 6.529 6.612 mp_irecv_dv 76685 16.2 2.579 6.444 2.579 6.444 rs_pw_transfer 1054 12.0 0.013 0.014 4.708 6.294 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.128 6.231 mp_alltoall_d11v 2423 14.1 4.981 5.866 4.981 5.866 grid_collocate_task_list 129 9.7 5.109 5.798 5.109 5.798 calculate_first_density_matrix 1 7.0 0.000 0.003 4.997 5.001 potential_pw2rs 129 12.3 0.023 0.023 4.728 4.752 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.588 4.708 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.705 4.706 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.161 4.220 jit_kernel_multiply 13 15.8 3.173 4.116 3.173 4.116 mp_waitany 15064 13.9 2.283 3.855 2.283 3.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.414000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.545455, yerr=20.799317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410024443904 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 11444712984576 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 92.796586E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705502176 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.512206E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108700816 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828148. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.055 179.313 179.314 qs_mol_dyn_low 1 2.0 0.003 0.005 178.873 178.885 qs_forces 11 3.9 0.003 0.003 174.803 174.809 qs_energies 11 4.9 0.002 0.002 167.233 167.241 scf_env_do_scf 11 5.9 0.001 0.001 147.681 147.686 velocity_verlet 10 3.0 0.001 0.001 117.225 117.477 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 111.467 111.469 dbcsr_multiply_generic 2485 12.5 0.181 0.185 75.537 75.931 qs_scf_new_mos 116 7.6 0.001 0.001 72.798 72.900 qs_scf_loop_do_ot 116 8.6 0.001 0.001 72.797 72.900 ot_scf_mini 116 9.6 0.003 0.004 68.408 68.490 multiply_cannon 2485 13.5 0.585 0.620 56.045 60.423 multiply_cannon_loop 2485 14.5 0.442 0.450 51.580 52.335 ot_mini 116 10.6 0.001 0.001 39.000 39.087 init_scf_loop 11 6.9 0.000 0.000 36.068 36.070 mp_waitall_1 124680 16.7 25.637 32.651 25.637 32.651 prepare_preconditioner 11 7.9 0.000 0.000 31.884 31.914 make_preconditioner 11 8.9 0.000 0.000 31.884 31.914 rebuild_ks_matrix 127 8.3 0.001 0.001 29.873 29.939 qs_ks_build_kohn_sham_matrix 127 9.3 0.019 0.020 29.872 29.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.527 29.805 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.133 27.188 multiply_cannon_multrec 9940 15.5 10.311 14.478 19.758 22.446 ot_diis_step 116 11.6 0.019 0.020 19.539 19.539 qs_ot_get_derivative 116 11.6 0.001 0.002 19.396 19.469 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.150 19.385 apply_single 127 13.6 0.001 0.001 19.150 19.384 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.118 18.892 cp_fm_cholesky_invert 11 10.9 18.065 18.071 18.065 18.071 make_m2s 4970 13.5 0.065 0.070 15.988 18.019 make_images 4970 14.5 2.270 2.783 15.684 17.708 qs_ot_get_p 127 10.4 0.001 0.001 16.830 16.921 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.486 15.518 calculate_rho_elec 127 8.7 0.257 0.266 15.486 15.517 sum_up_and_integrate 127 10.3 0.179 0.191 15.048 15.093 integrate_v_rspace 127 11.3 0.004 0.004 14.868 14.922 qs_ot_p2m_diag 82 11.4 0.489 0.494 12.937 12.954 init_scf_run 11 5.9 0.000 0.001 12.686 12.687 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.686 12.686 make_images_data 4970 15.5 0.051 0.059 9.756 12.145 multiply_cannon_sync_h2d 9940 15.5 11.543 11.853 11.543 11.853 cp_dbcsr_syevd 82 12.4 0.005 0.005 11.804 11.805 hybrid_alltoall_any 5155 16.4 0.817 3.595 9.551 11.705 pw_transfer 1535 11.6 0.085 0.094 9.905 9.937 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.684 9.722 dbcsr_mm_accdrv_process 20590 16.1 2.651 3.583 9.078 9.701 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.873 8.883 cp_fm_diag_elpa_base 82 14.4 8.621 8.704 8.862 8.873 fft_wrap_pw1pw2_140 519 13.2 0.494 0.520 8.558 8.598 grid_integrate_task_list 127 12.3 7.656 8.219 7.656 8.219 cp_fm_cholesky_decompose 22 10.9 7.919 8.016 7.919 8.016 density_rs2pw 127 9.7 0.005 0.006 7.053 7.912 fft3d_ps 1281 14.7 2.693 2.762 7.853 7.905 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.703 7.761 wfi_extrapolate 11 7.9 0.001 0.001 7.410 7.410 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.514 7.306 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.217 7.225 mp_allgather_i34 2485 14.5 2.680 7.003 2.680 7.003 mp_alltoall_d11v 2401 14.1 5.291 6.270 5.291 6.270 calculate_dm_sparse 127 9.5 0.001 0.001 6.168 6.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.163 6.187 grid_collocate_task_list 127 9.7 5.353 6.057 5.353 6.057 dbcsr_complete_redistribute 393 12.7 2.126 2.247 5.372 5.742 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.348 5.349 calculate_first_density_matrix 1 7.0 0.000 0.000 5.061 5.065 rs_pw_transfer 1038 11.9 0.012 0.013 4.076 4.970 potential_pw2rs 127 12.3 0.026 0.026 4.560 4.575 jit_kernel_multiply 12 15.5 3.903 4.327 3.903 4.327 mp_sum_dm 438 4.9 3.830 4.094 3.830 4.094 update_particle_set 20 4.0 0.000 0.000 3.789 4.041 md_output 10 3.0 0.000 0.000 0.257 4.039 md_write_output 11 3.9 0.243 3.886 0.254 4.036 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.840 4.033 mp_irecv_dv 28618 15.9 1.803 3.961 1.803 3.961 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.592 3.896 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.884 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.670 3.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.640 3.652 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=179.314000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1403.181818, yerr=46.859205 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 11.785519E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 3.063190E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591807936 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 58746. MP_Allreduce 11072 1512. MP_Sync 87 MP_Alltoall 1724 36993717. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.064 305.507 305.509 qs_mol_dyn_low 1 2.0 0.003 0.004 304.470 304.481 qs_forces 11 3.9 0.003 0.003 302.330 302.339 qs_energies 11 4.9 0.002 0.002 293.028 293.040 scf_env_do_scf 11 5.9 0.001 0.001 267.504 267.515 velocity_verlet 10 3.0 0.001 0.002 217.335 217.598 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 142.519 142.520 init_scf_loop 11 6.9 0.000 0.000 124.732 124.735 prepare_preconditioner 11 7.9 0.000 0.000 119.250 119.280 make_preconditioner 11 8.9 0.000 0.000 119.250 119.280 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.342 116.399 qs_scf_new_mos 118 7.6 0.001 0.001 91.104 91.215 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.103 91.214 dbcsr_multiply_generic 2527 12.6 0.214 0.220 86.994 87.586 ot_scf_mini 118 9.6 0.004 0.004 86.323 86.356 cp_fm_upper_to_full 105 14.8 52.920 75.777 52.920 75.777 multiply_cannon 2527 13.6 0.714 0.755 61.571 62.145 multiply_cannon_loop 2527 14.6 0.476 0.483 57.950 59.281 ot_mini 118 10.6 0.001 0.001 44.276 44.307 dbcsr_complete_redistribute 395 12.7 3.970 4.008 29.460 42.360 rebuild_ks_matrix 129 8.3 0.001 0.001 39.015 39.053 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.015 39.053 copy_fm_to_dbcsr 209 11.7 0.002 0.002 26.127 38.988 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.862 36.636 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.019 36.056 mp_alltoall_i22 718 14.1 21.672 34.494 21.672 34.494 cp_fm_cholesky_invert 11 10.9 33.237 33.244 33.237 33.244 mp_waitall_1 104500 16.8 28.635 32.760 28.635 32.760 qs_ot_get_p 129 10.4 0.001 0.001 26.568 26.577 qs_ot_get_derivative 118 11.6 0.002 0.002 24.149 24.182 qs_ot_p2m_diag 83 11.4 0.878 0.883 22.561 22.590 make_m2s 5054 13.6 0.076 0.078 20.901 21.789 make_images 5054 14.6 4.119 4.947 20.419 21.309 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.927 20.942 calculate_rho_elec 129 8.7 0.486 0.488 20.926 20.941 multiply_cannon_multrec 10108 15.6 10.554 12.320 20.736 20.846 ot_diis_step 118 11.6 0.022 0.022 20.102 20.102 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 18.771 19.852 sum_up_and_integrate 129 10.3 0.323 0.323 19.752 19.834 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.782 19.782 integrate_v_rspace 129 11.3 0.004 0.004 19.429 19.510 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.270 19.471 apply_single 129 13.6 0.001 0.001 19.269 19.471 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.623 16.625 cp_fm_diag_elpa_base 83 14.4 12.261 13.842 16.620 16.620 multiply_cannon_sync_h2d 10108 15.6 15.774 15.787 15.774 15.787 init_scf_run 11 5.9 0.000 0.001 15.431 15.431 scf_env_initial_rho_setup 11 6.9 0.001 0.001 15.431 15.431 make_images_data 5054 15.6 0.060 0.064 11.129 13.045 hybrid_alltoall_any 5240 16.5 1.314 3.090 11.054 12.990 pw_transfer 1559 11.6 0.094 0.094 12.408 12.414 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.171 12.175 dbcsr_mm_accdrv_process 20926 16.1 3.874 5.725 9.941 11.776 fft_wrap_pw1pw2_140 527 13.2 0.547 0.549 10.805 10.812 fft3d_ps 1301 14.7 2.763 2.770 10.191 10.201 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.343 9.378 wfi_extrapolate 11 7.9 0.001 0.001 9.034 9.034 mp_alltoall_d11v 2423 14.1 8.090 9.003 8.090 9.003 cp_fm_cholesky_decompose 22 10.9 8.920 8.937 8.920 8.937 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.779 8.833 grid_integrate_task_list 129 12.3 8.644 8.801 8.644 8.801 density_rs2pw 129 9.7 0.005 0.005 8.420 8.554 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.133 8.137 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.009 7.048 calculate_dm_sparse 129 9.5 0.001 0.001 6.909 7.009 grid_collocate_task_list 129 9.7 6.414 6.443 6.414 6.443 rs_scatter_matrices 140 9.7 3.615 4.552 6.055 6.287 calculate_first_density_matrix 1 7.0 0.000 0.000 6.114 6.115 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=305.509000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2708.181818, yerr=182.500719 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 1.261830E+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 6805335. 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 0.021 0.067 85.269 85.270 qs_energies 1 2.0 0.000 0.000 84.739 84.751 ls_scf 1 3.0 0.001 0.009 83.562 83.576 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.643 72.836 multiply_cannon 111 7.7 0.018 0.021 56.069 57.124 multiply_cannon_loop 111 8.7 0.210 0.228 52.686 53.912 ls_scf_main 1 4.0 0.000 0.002 52.087 52.090 density_matrix_trs4 2 5.0 0.002 0.003 46.627 46.685 ls_scf_init_scf 1 4.0 0.000 0.001 28.466 28.468 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.358 27.415 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.251 25.263 mp_waitall_1 11031 10.9 22.204 24.468 22.204 24.468 multiply_cannon_multrec 2664 9.7 8.165 8.866 15.489 17.167 multiply_cannon_sync_h2d 2664 9.7 13.676 15.621 13.676 15.621 make_m2s 222 7.7 0.008 0.011 12.924 13.429 make_images 222 8.7 0.099 0.109 12.901 13.410 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.577 12.239 make_images_data 222 9.7 0.004 0.005 7.472 8.025 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.474 7.914 dbcsr_mm_accdrv_process 4760 10.4 0.513 0.613 6.942 7.880 hybrid_alltoall_any 227 10.6 0.215 1.830 6.439 7.665 dbcsr_mm_accdrv_process_sort 4760 11.4 6.227 7.090 6.227 7.090 calculate_norms 4752 9.8 5.534 6.081 5.534 6.081 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.976 5.105 mp_sum_l 807 5.4 3.134 4.612 3.134 4.612 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.370 3.584 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.076 3.549 mp_irecv_dv 6231 10.9 2.059 3.523 2.059 3.523 make_images_sizes 222 9.7 0.000 0.000 0.666 3.293 mp_alltoall_i44 222 10.7 0.666 3.292 0.666 3.292 arnoldi_extremal 4 6.8 0.000 0.000 3.221 3.244 arnoldi_normal_ev 4 7.8 0.001 0.003 3.221 3.244 build_subspace 16 8.4 0.009 0.012 3.131 3.133 ls_scf_post 1 4.0 0.000 0.003 3.008 3.024 ls_scf_store_result 1 5.0 0.000 0.000 2.831 2.871 dbcsr_special_finalize 555 9.7 0.005 0.006 2.429 2.844 dbcsr_merge_single_wm 555 10.7 0.467 0.591 2.421 2.836 make_images_pack 222 9.7 2.210 2.629 2.212 2.631 dbcsr_sort_data 658 11.4 2.211 2.590 2.211 2.590 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.319 2.555 dbcsr_matrix_vector_mult_local 304 10.0 2.073 2.461 2.075 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.274 2.342 buffer_matrices_ensure_size 222 8.7 1.760 2.047 1.760 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.792 1.797 rebuild_ks_matrix 3 7.3 0.000 0.000 1.783 1.788 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.783 1.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.270000, 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/03b41477b2e102c540934725bb0f08de7bd8b995_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 2.093691E+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 0.281 0.299 91.694 91.694 qs_energies 1 2.0 0.000 0.000 90.953 90.958 ls_scf 1 3.0 0.000 0.000 89.640 89.647 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.555 75.878 multiply_cannon 111 7.7 0.029 0.041 53.692 57.312 ls_scf_main 1 4.0 0.000 0.000 54.885 54.891 multiply_cannon_loop 111 8.7 0.116 0.123 50.397 53.143 density_matrix_trs4 2 5.0 0.002 0.003 49.199 49.375 ls_scf_init_scf 1 4.0 0.000 0.000 31.134 31.135 mp_waitall_1 9105 10.9 21.530 30.522 21.530 30.522 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.744 29.839 multiply_cannon_multrec 1332 9.7 13.095 16.863 22.393 27.591 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.347 27.358 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.067 21.302 make_m2s 222 7.7 0.006 0.007 15.267 15.984 make_images 222 8.7 1.584 1.925 15.236 15.954 dbcsr_mm_accdrv_process 4041 10.4 0.293 0.794 8.902 10.373 dbcsr_mm_accdrv_process_sort 4041 11.4 8.407 9.922 8.407 9.922 make_images_data 222 9.7 0.004 0.004 8.779 9.625 hybrid_alltoall_any 227 10.6 0.519 2.423 8.220 9.238 mp_sum_l 807 5.4 5.295 8.256 5.295 8.256 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.234 7.597 mp_irecv_dv 3311 11.0 3.214 7.531 3.214 7.531 calculate_norms 2376 9.8 5.998 6.803 5.998 6.803 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.989 6.634 multiply_cannon_sync_h2d 1332 9.7 4.783 5.894 4.783 5.894 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.005 5.170 arnoldi_extremal 4 6.8 0.000 0.000 4.602 4.620 arnoldi_normal_ev 4 7.8 0.001 0.004 4.602 4.620 build_subspace 16 8.4 0.014 0.021 4.345 4.348 ls_scf_post 1 4.0 0.000 0.000 3.620 3.626 ls_scf_store_result 1 5.0 0.000 0.000 3.322 3.448 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.109 3.343 dbcsr_matrix_vector_mult_local 304 10.0 2.730 3.213 2.732 3.215 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.318 3.038 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.598 2.688 make_images_pack 222 9.7 2.028 2.436 2.030 2.438 mp_allgather_i34 111 8.7 0.978 2.434 0.978 2.434 dbcsr_sort_data 436 11.2 1.786 2.062 1.786 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.981 1.983 rebuild_ks_matrix 3 7.3 0.000 0.000 1.969 1.971 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.969 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.694000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1805.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 2.713817E+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 0.035 0.064 97.129 97.130 qs_energies 1 2.0 0.000 0.000 96.144 96.147 ls_scf 1 3.0 0.000 0.001 94.733 94.737 dbcsr_multiply_generic 111 6.7 0.016 0.017 79.381 79.655 multiply_cannon 111 7.7 0.041 0.100 55.173 58.921 ls_scf_main 1 4.0 0.000 0.000 58.042 58.047 multiply_cannon_loop 111 8.7 0.099 0.104 51.621 55.781 density_matrix_trs4 2 5.0 0.002 0.003 52.085 52.231 mp_waitall_1 7281 11.0 24.651 34.601 24.651 34.601 ls_scf_init_scf 1 4.0 0.000 0.002 33.114 33.116 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.784 31.853 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.361 29.381 multiply_cannon_multrec 888 9.7 12.535 15.260 21.066 24.430 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.578 23.680 make_m2s 222 7.7 0.006 0.007 17.322 18.564 make_images 222 8.7 1.983 2.300 17.284 18.525 hybrid_alltoall_any 227 10.6 0.622 2.873 9.560 11.040 make_images_data 222 9.7 0.003 0.005 9.929 10.948 dbcsr_mm_accdrv_process 3754 10.4 0.262 0.466 8.061 9.304 dbcsr_mm_accdrv_process_sort 3754 11.4 7.681 8.837 7.681 8.837 mp_sum_l 807 5.4 5.328 8.811 5.328 8.811 multiply_cannon_sync_h2d 888 9.7 6.006 7.397 6.006 7.397 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.498 7.164 mp_irecv_dv 2335 11.1 2.483 7.112 2.483 7.112 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.768 7.043 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.105 7.015 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.935 5.144 arnoldi_extremal 4 6.8 0.000 0.000 5.023 5.046 arnoldi_normal_ev 4 7.8 0.001 0.005 5.023 5.046 build_subspace 16 8.4 0.014 0.020 4.715 4.721 calculate_norms 1584 9.8 4.291 4.680 4.291 4.680 mp_allgather_i34 111 8.7 1.393 3.854 1.393 3.854 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.412 3.726 ls_scf_post 1 4.0 0.000 0.001 3.578 3.582 dbcsr_matrix_vector_mult_local 304 10.0 3.010 3.567 3.012 3.569 ls_scf_store_result 1 5.0 0.000 0.000 3.311 3.404 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.853 2.966 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.775 2.782 acc_transpose_blocks 888 9.7 0.004 0.005 1.988 2.234 dbcsr_sort_data 325 11.1 1.882 2.222 1.882 2.222 make_images_sizes 222 9.7 0.000 0.000 1.050 2.210 mp_alltoall_i44 222 10.7 1.049 2.210 1.049 2.210 acc_transpose_blocks_kernels 888 10.7 0.012 0.012 1.948 2.193 jit_kernel_transpose 1 13.0 1.937 2.182 1.937 2.182 make_images_pack 222 9.7 1.827 2.138 1.830 2.140 dbcsr_data_release 9322 10.9 1.317 1.978 1.317 1.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=97.130000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 3.360350E+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 0.042 0.068 98.071 98.072 qs_energies 1 2.0 0.000 0.000 97.351 97.357 ls_scf 1 3.0 0.041 0.114 95.700 95.712 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.410 78.661 ls_scf_main 1 4.0 0.000 0.000 58.846 59.145 multiply_cannon 111 7.7 0.073 0.193 51.753 56.124 density_matrix_trs4 2 5.0 0.002 0.003 52.591 52.688 multiply_cannon_loop 111 8.7 0.114 0.126 46.752 49.461 ls_scf_init_scf 1 4.0 0.001 0.003 33.559 33.598 ls_scf_init_matrix_S 1 5.0 0.000 0.005 31.413 31.494 mp_waitall_1 6369 11.0 23.024 30.376 23.024 30.376 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.947 28.958 multiply_cannon_multrec 1332 9.7 14.146 17.636 22.033 24.805 make_m2s 222 7.7 0.006 0.007 21.203 22.568 make_images 222 8.7 3.137 3.605 21.153 22.520 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.538 16.827 make_images_data 222 9.7 0.004 0.004 11.880 13.484 hybrid_alltoall_any 227 10.6 0.797 3.806 11.275 12.992 dbcsr_mm_accdrv_process 3641 10.4 0.217 0.523 7.524 9.019 dbcsr_mm_accdrv_process_sort 3641 11.4 7.112 8.608 7.112 8.608 mp_sum_l 807 5.4 3.883 7.065 3.883 7.065 multiply_cannon_sync_h2d 1332 9.7 5.503 6.319 5.503 6.319 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.087 6.114 mp_irecv_dv 3229 10.9 2.065 6.030 2.065 6.030 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.958 5.535 arnoldi_extremal 4 6.8 0.000 0.000 5.186 5.199 arnoldi_normal_ev 4 7.8 0.001 0.004 5.185 5.199 build_subspace 16 8.4 0.014 0.021 4.851 4.859 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.534 4.707 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.438 4.673 calculate_norms 2376 9.8 4.220 4.635 4.220 4.635 mp_allgather_i34 111 8.7 2.133 4.442 2.133 4.442 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.560 3.866 dbcsr_matrix_vector_mult_local 304 10.0 3.175 3.684 3.177 3.686 dbcsr_sort_data 658 11.4 3.103 3.462 3.103 3.462 ls_scf_post 1 4.0 0.003 0.009 3.253 3.264 dbcsr_special_finalize 555 9.7 0.006 0.007 2.855 3.252 dbcsr_merge_single_wm 555 10.7 0.538 0.668 2.846 3.245 ls_scf_dm_to_ks 2 5.0 0.006 0.029 3.063 3.114 ls_scf_store_result 1 5.0 0.000 0.000 2.998 3.044 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.520 2.559 rebuild_ks_matrix 3 7.3 0.000 0.000 2.497 2.535 qs_ks_build_kohn_sham_matrix 3 8.3 0.265 0.323 2.497 2.535 dbcsr_data_release 10477 10.7 1.577 2.407 1.577 2.407 ls_scf_initial_guess 1 5.0 0.000 0.000 2.145 2.276 ls_scf_qs_atomic_guess 1 6.0 0.000 0.001 2.145 2.276 dbcsr_finalize 304 7.8 0.049 0.061 1.797 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.072000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 4.624548E+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 0.046 0.063 100.114 100.114 qs_energies 1 2.0 0.000 0.000 99.009 99.016 ls_scf 1 3.0 0.000 0.000 96.775 96.787 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.004 78.180 ls_scf_main 1 4.0 0.000 0.000 62.353 62.354 multiply_cannon 111 7.7 0.086 0.202 55.409 60.849 density_matrix_trs4 2 5.0 0.002 0.003 54.994 55.086 multiply_cannon_loop 111 8.7 0.070 0.076 51.017 52.522 mp_waitall_1 5436 11.0 26.712 31.988 26.712 31.988 ls_scf_init_scf 1 4.0 0.000 0.000 30.846 30.851 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.440 29.497 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.291 27.306 multiply_cannon_multrec 444 9.7 14.047 16.596 21.085 22.570 make_m2s 222 7.7 0.004 0.005 17.829 20.234 make_images 222 8.7 3.723 4.415 17.768 20.175 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.978 16.584 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.498 14.539 make_images_data 222 9.7 0.003 0.004 10.038 12.390 hybrid_alltoall_any 227 10.6 0.788 3.763 9.880 12.280 dbcsr_mm_accdrv_process 3003 10.4 0.191 0.358 6.740 7.869 multiply_cannon_sync_h2d 444 9.7 6.589 7.545 6.589 7.545 dbcsr_mm_accdrv_process_sort 3003 11.4 6.421 7.524 6.421 7.524 mp_allgather_i34 111 8.7 2.627 7.009 2.627 7.009 arnoldi_extremal 4 6.8 0.000 0.000 5.755 5.764 arnoldi_normal_ev 4 7.8 0.002 0.005 5.755 5.763 build_subspace 16 8.4 0.015 0.020 5.360 5.368 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.569 4.735 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.677 4.443 mp_irecv_dv 1241 11.2 1.659 4.429 1.659 4.429 mp_sum_l 807 5.4 2.891 4.329 2.891 4.329 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.118 4.310 dbcsr_matrix_vector_mult_local 304 10.0 3.652 4.116 3.654 4.118 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.755 3.844 calculate_norms 792 9.8 3.545 3.650 3.545 3.650 ls_scf_post 1 4.0 0.000 0.000 3.577 3.583 make_images_sizes 222 9.7 0.000 0.000 1.152 3.507 mp_alltoall_i44 222 10.7 1.151 3.507 1.151 3.507 ls_scf_store_result 1 5.0 0.000 0.000 3.363 3.404 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.907 3.059 dbcsr_finalize 304 7.8 0.062 0.078 2.195 2.245 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.240 2.242 qs_energies_init_hamiltonians 1 3.0 0.004 0.008 2.218 2.218 rebuild_ks_matrix 3 7.3 0.000 0.000 2.208 2.209 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.208 2.209 dbcsr_merge_all 275 8.9 0.473 0.523 2.046 2.091 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.114000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3686.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/03b41477b2e102c540934725bb0f08de7bd8b995_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 8.769987E+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 0.105 0.141 108.825 108.826 qs_energies 1 2.0 0.000 0.000 107.337 107.352 ls_scf 1 3.0 0.000 0.000 104.404 104.419 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.767 77.903 ls_scf_main 1 4.0 0.000 0.000 65.884 65.885 density_matrix_trs4 2 5.0 0.002 0.003 56.527 56.585 multiply_cannon 111 7.7 0.141 0.328 50.172 52.160 multiply_cannon_loop 111 8.7 0.068 0.072 46.645 47.758 ls_scf_init_scf 1 4.0 0.000 0.000 34.859 34.859 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.063 33.074 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.233 30.241 mp_waitall_1 4527 11.1 21.953 25.830 21.953 25.830 make_m2s 222 7.7 0.005 0.005 23.902 24.953 make_images 222 8.7 4.589 5.002 23.795 24.845 multiply_cannon_multrec 444 9.7 17.928 18.707 22.572 23.364 hybrid_alltoall_any 227 10.6 1.662 3.632 12.990 15.789 make_images_data 222 9.7 0.003 0.004 13.148 15.618 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.220 11.036 multiply_cannon_sync_h2d 444 9.7 8.849 8.894 8.849 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.228 7.238 arnoldi_normal_ev 4 7.8 0.003 0.009 7.228 7.238 build_subspace 16 8.4 0.026 0.036 6.682 6.697 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.331 5.473 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.316 5.404 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.025 5.279 dbcsr_matrix_vector_mult_local 304 10.0 4.919 5.202 4.922 5.204 dbcsr_mm_accdrv_process 1814 10.4 0.301 0.411 4.481 4.622 dbcsr_mm_accdrv_process_sort 1814 11.4 4.129 4.260 4.129 4.260 mp_allgather_i34 111 8.7 1.181 3.695 1.181 3.695 ls_scf_post 1 4.0 0.000 0.000 3.662 3.676 make_images_sizes 222 9.7 0.000 0.000 1.416 3.477 mp_alltoall_i44 222 10.7 1.416 3.476 1.416 3.476 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.419 calculate_norms 792 9.8 3.242 3.284 3.242 3.284 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.188 dbcsr_merge_all 275 8.9 0.893 0.923 2.874 2.969 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.902 2.902 dbcsr_complete_redistribute 5 7.6 1.434 1.487 2.766 2.873 dbcsr_data_release 12724 10.6 2.334 2.833 2.334 2.833 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.678 2.679 rebuild_ks_matrix 3 7.3 0.000 0.000 2.612 2.613 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.009 2.612 2.613 matrix_ls_to_qs 2 6.0 0.000 0.000 2.411 2.536 dbcsr_sort_data 325 11.1 2.444 2.516 2.444 2.516 dbcsr_new_transposed 4 7.5 0.244 0.253 2.302 2.324 dbcsr_frobenius_norm 74 6.6 2.059 2.128 2.195 2.238 dbcsr_add_d 103 6.2 0.000 0.000 2.137 2.209 dbcsr_add_anytype 103 7.2 0.861 0.891 2.137 2.209 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.826000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6866.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 03b41477b2e102c540934725bb0f08de7bd8b995 Summary: empty Status: OK