=== 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: d168495ed9e6cddd42101d7df39d7697855199e2 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.04.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/01 job id: 46490078 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/02 job id: 46490080 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/03 job id: 46490081 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/04 job id: 46490082 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/05 job id: 46490083 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/06 job id: 46490084 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/07 job id: 46490085 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/08 job id: 46490086 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/09 job id: 46490087 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/10 job id: 46490088 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/11 job id: 46490089 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/12 job id: 46490091 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/13 job id: 46490093 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/14 job id: 46490096 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/15 job id: 46490097 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/16 job id: 46490098 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/17 job id: 46490101 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/18 job id: 46490102 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/19 job id: 46490103 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/20 job id: 46490104 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/21 job id: 46490105 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/22 job id: 46490106 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/23 job id: 46490107 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/24 job id: 46490108 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/25 job id: 46490109 --- 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/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/26 job id: 46490110 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/27 job id: 46490111 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 135.708 135.709 farming_run 1 2.0 134.745 134.747 135.680 135.683 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459266E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 115.531 115.532 qs_energies 1 2.0 0.000 0.000 115.168 115.169 mp2_main 1 3.0 0.000 0.000 113.131 113.133 mp2_gpw_main 1 4.0 0.020 0.026 112.122 112.124 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.490 93.582 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.375 55.467 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.652 46.595 get_2c_integrals 1 6.0 0.008 0.009 37.218 37.942 integrate_v_rspace 273 8.0 0.437 0.450 25.058 29.698 pw_transfer 6555 10.6 0.370 0.384 27.358 27.992 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.006 26.452 grid_integrate_task_list 273 9.0 20.884 25.961 20.884 25.961 fft_wrap_pw1pw2_100 2178 12.4 1.212 1.467 23.532 24.008 compute_2c_integrals 1 7.0 0.002 0.002 19.723 19.724 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.897 19.451 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.426 18.895 19.450 rpa_ri_compute_en 1 5.0 0.019 0.021 18.523 18.608 cp_fm_cholesky_decompose 12 8.2 17.436 18.193 17.436 18.193 cholesky_decomp 1 7.0 0.000 0.000 16.337 17.096 fft3d_s 5443 13.4 16.126 16.427 16.148 16.449 ao_to_mo_and_store_B_mult_1 272 7.0 10.857 15.580 10.857 15.580 calculate_wavefunction 272 8.0 5.420 5.500 12.521 13.174 rpa_num_int 1 6.0 0.000 0.000 10.634 10.643 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.621 10.630 calc_mat_Q 8 8.0 0.000 0.000 9.477 9.575 contract_S_to_Q 8 9.0 0.000 0.000 8.898 8.992 calc_potential_gpw 544 9.5 0.005 0.005 8.206 8.655 parallel_gemm_fm 14 9.1 0.000 0.000 8.496 8.581 parallel_gemm_fm_cosma 14 10.1 8.496 8.581 8.496 8.581 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.212 8.560 potential_pw2rs 545 10.0 0.107 0.109 7.710 8.339 collocate_single_gaussian 272 10.0 0.040 0.042 7.489 7.795 create_integ_mat 1 6.0 0.014 0.027 7.703 7.712 array2fm 1 7.0 0.000 0.000 6.633 7.142 pw_scatter_s 2720 13.7 4.413 4.644 4.413 4.644 pw_gather_s 2722 13.2 3.851 4.303 3.851 4.303 array2fm_buffer_send 1 8.0 2.915 3.121 2.915 3.121 pw_poisson_solve 545 10.5 1.108 1.143 2.169 2.339 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.121699, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.040 398.060 398.061 farming_run 1 2.0 397.156 397.167 398.021 398.023 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224511E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 3929151. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.029 209.709 209.709 qs_energies 1 2.0 0.086 0.346 209.430 209.445 scf_env_do_scf 1 3.0 0.000 0.000 105.684 105.684 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.815 104.824 rebuild_ks_matrix 4 6.0 0.000 0.000 104.814 104.823 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.063 104.814 104.823 hfx_ks_matrix 4 8.0 0.001 0.001 104.436 104.441 integrate_four_center 4 9.0 0.144 0.458 104.436 104.440 mp2_main 1 3.0 0.013 0.071 103.373 103.464 mp2_gpw_main 1 4.0 0.043 0.103 102.236 102.252 integrate_four_center_main 4 10.0 0.123 0.493 96.644 98.460 integrate_four_center_bin 270 11.0 96.521 98.454 96.521 98.454 init_scf_loop 1 4.0 0.000 0.000 91.446 91.447 mp2_ri_gpw_compute_in 1 5.0 0.071 0.103 74.966 76.074 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.004 54.626 55.730 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.164 42.309 47.139 integrate_v_rspace 95 8.0 0.398 0.560 28.687 33.315 pw_transfer 2240 10.6 0.145 0.166 29.926 30.375 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.937 29.396 ao_to_mo_and_store_B_mult_1 91 7.0 10.623 29.162 10.623 29.162 grid_integrate_task_list 95 9.0 23.976 28.793 23.976 28.793 mp2_ri_gpw_compute_en 1 5.0 0.059 0.119 27.042 28.748 fft_wrap_pw1pw2_100 730 12.4 1.272 1.459 26.655 27.130 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.836 1.916 25.224 25.234 get_2c_integrals 1 6.0 0.001 0.008 20.244 20.275 compute_2c_integrals 1 7.0 0.003 0.014 19.222 19.236 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.864 19.106 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.852 18.863 19.105 fft3d_s 1823 13.4 18.425 18.795 18.439 18.809 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.236 14.236 calculate_wavefunction 91 8.0 2.012 2.042 9.736 9.965 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.588 8.842 9.464 potential_pw2rs 186 10.0 0.034 0.035 8.628 9.103 local_gemm 172 8.0 8.286 8.892 8.286 8.892 mp2_ri_gpw_compute_en_comm 22 7.0 0.506 0.525 8.163 8.601 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.240 8.520 collocate_single_gaussian 91 10.0 0.017 0.026 7.887 8.202 calc_potential_gpw 182 9.5 0.002 0.002 7.920 8.116 mp_sendrecv_dm3 2068 8.0 6.185 6.631 6.185 6.631 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.404 6.337 6.404 pw_gather_s 912 13.2 4.925 5.390 4.925 5.390 mp_sync 37 10.5 2.458 4.646 2.458 4.646 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.239951, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 448.114688E+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 10329 270. MP_Sync 530 MP_Alltoall 2083 153398. 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 0.016 0.032 53.863 53.865 qs_mol_dyn_low 1 2.0 0.004 0.006 53.553 53.561 qs_forces 11 3.9 0.002 0.003 53.284 53.285 qs_energies 11 4.9 0.001 0.002 51.813 51.826 scf_env_do_scf 11 5.9 0.000 0.001 45.791 45.791 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.779 43.779 qs_scf_new_mos 108 7.5 0.000 0.001 33.852 34.150 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.851 34.150 dbcsr_multiply_generic 2286 12.5 0.092 0.095 33.642 33.970 ot_scf_mini 108 9.5 0.002 0.003 32.161 32.344 multiply_cannon 2286 13.5 0.183 0.196 26.338 27.794 multiply_cannon_loop 2286 14.5 1.483 1.572 25.694 27.179 velocity_verlet 10 3.0 0.023 0.175 26.199 26.200 ot_mini 108 10.5 0.001 0.001 19.194 19.465 qs_ot_get_derivative 108 11.5 0.001 0.002 16.193 16.395 mp_waitall_1 245248 16.5 8.755 14.676 8.755 14.676 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.285 12.994 multiply_cannon_multrec 54864 15.5 4.252 6.559 7.702 11.202 qs_ot_get_p 119 10.4 0.001 0.001 8.185 8.482 rebuild_ks_matrix 119 8.3 0.000 0.000 7.788 7.917 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.787 7.917 mp_sum_l 7287 12.8 5.494 7.286 5.494 7.286 multiply_cannon_sync_h2d 54864 15.5 5.933 7.072 5.933 7.072 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.860 6.986 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.614 6.046 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.402 5.513 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.292 5.342 init_scf_run 11 5.9 0.000 0.001 4.815 4.815 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.814 4.815 dbcsr_mm_accdrv_process 76910 16.1 1.142 1.811 3.372 4.773 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.486 4.487 sum_up_and_integrate 119 10.3 0.012 0.016 4.418 4.426 integrate_v_rspace 119 11.3 0.002 0.003 4.406 4.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.299 4.299 cp_fm_redistribute_end 50 14.0 2.193 4.276 2.199 4.279 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.067 4.193 calculate_rho_elec 119 8.7 0.012 0.017 4.067 4.193 cp_fm_diag_elpa_base 50 14.0 2.075 4.172 2.078 4.180 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.916 3.121 apply_single 119 13.6 0.000 0.000 2.915 3.121 calculate_dm_sparse 119 9.5 0.000 0.001 2.924 3.066 multiply_cannon_metrocomm1 54864 15.5 0.052 0.056 1.885 3.006 jit_kernel_multiply 13 15.8 2.168 2.882 2.168 2.882 ot_diis_step 108 11.5 0.007 0.013 2.728 2.731 rs_pw_transfer 974 11.9 0.012 0.013 2.603 2.703 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.579 2.651 calculate_first_density_matrix 1 7.0 0.000 0.001 2.614 2.619 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.397 2.399 density_rs2pw 119 9.7 0.004 0.005 2.090 2.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.106 2.152 wfi_extrapolate 11 7.9 0.001 0.002 2.143 2.143 acc_transpose_blocks 54864 15.5 0.226 0.244 1.699 2.129 grid_integrate_task_list 119 12.3 2.013 2.086 2.013 2.086 init_scf_loop 11 6.9 0.000 0.000 1.997 1.997 mp_sum_d 4135 12.0 1.333 1.891 1.333 1.891 potential_pw2rs 119 12.3 0.004 0.004 1.778 1.785 pw_transfer 1439 11.6 0.052 0.057 1.644 1.714 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.568 1.642 make_m2s 4572 13.5 0.054 0.055 1.503 1.542 make_images 4572 14.5 0.132 0.138 1.420 1.458 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.389 1.412 mp_alltoall_d11v 2130 13.8 1.230 1.407 1.230 1.407 mp_waitany 12084 13.8 1.230 1.380 1.230 1.380 grid_collocate_task_list 119 9.7 1.290 1.374 1.290 1.374 fft_wrap_pw1pw2_140 487 13.2 0.183 0.200 1.207 1.281 fft3d_ps 1201 14.6 0.370 0.476 1.212 1.279 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.751 1.137 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.865000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.272727, yerr=1.135454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 484.315136E+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 10306 303. MP_Sync 54 MP_Alltoall 2060 1160182. 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 0.010 0.028 38.689 38.690 qs_mol_dyn_low 1 2.0 0.003 0.003 38.512 38.520 qs_forces 11 3.9 0.002 0.003 38.456 38.456 qs_energies 11 4.9 0.001 0.001 36.780 36.782 scf_env_do_scf 11 5.9 0.000 0.001 31.247 31.248 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.842 28.842 dbcsr_multiply_generic 2286 12.5 0.098 0.101 21.663 22.028 qs_scf_new_mos 108 7.5 0.001 0.001 19.814 20.051 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.813 20.051 ot_scf_mini 108 9.5 0.002 0.003 18.933 19.100 multiply_cannon 2286 13.5 0.211 0.219 16.780 18.180 velocity_verlet 10 3.0 0.001 0.001 17.955 17.956 multiply_cannon_loop 2286 14.5 0.903 0.974 15.687 17.066 ot_mini 108 10.5 0.001 0.001 11.654 11.884 mp_waitall_1 200699 16.5 5.674 11.210 5.674 11.210 multiply_cannon_metrocomm3 27432 15.5 0.066 0.069 4.261 9.836 qs_ot_get_derivative 108 11.5 0.001 0.001 9.202 9.370 multiply_cannon_multrec 27432 15.5 1.986 4.400 6.204 9.043 rebuild_ks_matrix 119 8.3 0.000 0.000 7.107 7.245 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.107 7.244 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.274 6.399 dbcsr_mm_accdrv_process 47894 16.0 3.293 5.631 4.147 6.108 qs_ot_get_p 119 10.4 0.001 0.001 4.404 4.635 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.585 4.421 init_scf_run 11 5.9 0.000 0.001 4.344 4.345 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.344 4.344 sum_up_and_integrate 119 10.3 0.024 0.027 4.130 4.135 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.033 4.113 apply_single 119 13.6 0.000 0.000 3.033 4.112 integrate_v_rspace 119 11.3 0.002 0.002 4.105 4.112 mp_sum_l 7287 12.8 2.004 3.929 2.004 3.929 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.703 3.733 calculate_rho_elec 119 8.7 0.021 0.024 3.702 3.732 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.992 3.011 rs_pw_transfer 974 11.9 0.010 0.011 2.463 2.865 calculate_first_density_matrix 1 7.0 0.000 0.000 2.861 2.862 multiply_cannon_sync_h2d 27432 15.5 2.197 2.778 2.197 2.778 make_m2s 4572 13.5 0.053 0.055 2.429 2.675 make_images 4572 14.5 0.199 0.236 2.341 2.584 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.563 2.563 density_rs2pw 119 9.7 0.004 0.004 2.036 2.454 ot_diis_step 108 11.5 0.010 0.011 2.402 2.403 init_scf_loop 11 6.9 0.000 0.000 2.385 2.385 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.334 2.336 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.243 2.243 cp_fm_redistribute_end 50 14.0 1.137 2.217 1.140 2.220 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.152 2.191 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.099 2.186 calculate_dm_sparse 119 9.5 0.000 0.001 2.097 2.171 cp_fm_diag_elpa_base 50 14.0 1.047 2.127 1.076 2.164 jit_kernel_multiply 10 16.2 0.801 2.020 0.801 2.020 grid_integrate_task_list 119 12.3 1.841 1.949 1.841 1.949 pw_transfer 1439 11.6 0.065 0.070 1.896 1.931 potential_pw2rs 119 12.3 0.006 0.006 1.850 1.858 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.804 1.840 acc_transpose_blocks 27432 15.5 0.109 0.114 1.257 1.602 prepare_preconditioner 11 7.9 0.000 0.000 1.497 1.524 make_preconditioner 11 8.9 0.000 0.000 1.497 1.524 make_images_data 4572 15.5 0.046 0.051 1.119 1.518 fft_wrap_pw1pw2_140 487 13.2 0.201 0.210 1.430 1.465 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.392 1.450 wfi_extrapolate 11 7.9 0.001 0.001 1.433 1.433 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.974 1.413 fft3d_ps 1201 14.6 0.519 0.574 1.355 1.383 grid_collocate_task_list 119 9.7 1.229 1.360 1.229 1.360 mp_alltoall_d11v 2130 13.8 1.161 1.338 1.161 1.338 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.324 1.330 mp_allgather_i34 2286 14.5 0.532 1.259 0.532 1.259 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.166 1.215 mp_sum_d 4135 12.0 0.554 0.996 0.554 0.996 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.725 0.985 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.149 0.530 0.930 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.927 0.928 mp_waitany 5720 13.7 0.501 0.911 0.501 0.911 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.690000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.454545, yerr=1.724879 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 516.874240E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 514782. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.036 32.947 32.949 qs_mol_dyn_low 1 2.0 0.003 0.004 32.617 32.624 qs_forces 11 3.9 0.007 0.035 32.561 32.562 qs_energies 11 4.9 0.009 0.066 30.965 30.968 scf_env_do_scf 11 5.9 0.001 0.004 26.008 26.009 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 22.980 22.980 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.277 16.365 velocity_verlet 10 3.0 0.001 0.002 14.993 14.994 qs_scf_new_mos 108 7.5 0.001 0.001 14.916 14.926 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.916 14.925 ot_scf_mini 108 9.5 0.003 0.005 14.205 14.215 multiply_cannon 2286 13.5 0.195 0.204 13.114 13.888 multiply_cannon_loop 2286 14.5 0.634 0.667 12.346 13.153 ot_mini 108 10.5 0.001 0.001 8.544 8.558 multiply_cannon_multrec 18288 15.5 1.928 2.905 6.941 7.284 qs_ot_get_derivative 108 11.5 0.001 0.001 7.059 7.071 rebuild_ks_matrix 119 8.3 0.000 0.000 6.408 6.426 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.408 6.426 dbcsr_mm_accdrv_process 38222 16.0 4.016 5.608 4.929 5.744 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.683 5.700 sum_up_and_integrate 119 10.3 0.030 0.031 3.896 3.900 integrate_v_rspace 119 11.3 0.003 0.004 3.865 3.873 init_scf_run 11 5.9 0.000 0.001 3.610 3.610 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.610 3.610 mp_waitall_1 158411 16.6 2.464 3.607 2.464 3.607 qs_ot_get_p 119 10.4 0.036 0.281 3.503 3.522 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.386 3.399 calculate_rho_elec 119 8.7 0.031 0.031 3.386 3.398 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.640 3.268 init_scf_loop 11 6.9 0.052 0.411 3.012 3.013 rs_pw_transfer 974 11.9 0.009 0.010 2.200 2.450 calculate_first_density_matrix 1 7.0 0.000 0.003 2.411 2.412 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.031 2.359 apply_single 119 13.6 0.000 0.000 2.031 2.359 multiply_cannon_metrocomm3 18288 15.5 0.043 0.045 1.383 2.347 jit_kernel_multiply 11 16.1 0.861 2.218 0.861 2.218 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.207 2.212 density_rs2pw 119 9.7 0.004 0.004 1.898 2.132 calculate_dm_sparse 119 9.5 0.000 0.000 1.980 1.987 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.948 1.948 grid_integrate_task_list 119 12.3 1.803 1.900 1.803 1.900 pw_transfer 1439 11.6 0.066 0.070 1.874 1.883 make_m2s 4572 13.5 0.045 0.046 1.707 1.867 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.781 1.793 make_images 4572 14.5 0.189 0.199 1.621 1.778 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.712 1.721 cp_fm_diag_elpa_base 50 14.0 1.689 1.698 1.710 1.719 prepare_preconditioner 11 7.9 0.000 0.000 1.699 1.701 make_preconditioner 11 8.9 0.000 0.001 1.699 1.701 potential_pw2rs 119 12.3 0.007 0.008 1.662 1.667 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.649 1.651 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.553 1.639 multiply_cannon_sync_h2d 18288 15.5 1.406 1.581 1.406 1.581 qs_ot_get_derivative_diag 49 12.0 0.001 0.004 1.520 1.524 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.514 1.521 ot_diis_step 108 11.5 0.011 0.011 1.467 1.467 mp_sum_l 7287 12.8 1.074 1.461 1.074 1.461 fft_wrap_pw1pw2_140 487 13.2 0.254 0.260 1.443 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.359 1.363 grid_collocate_task_list 119 9.7 1.216 1.345 1.216 1.345 fft3d_ps 1201 14.6 0.528 0.546 1.267 1.279 acc_transpose_blocks 18288 15.5 0.076 0.078 1.230 1.251 wfi_extrapolate 11 7.9 0.001 0.001 1.147 1.147 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.986 0.987 make_images_data 4572 15.5 0.045 0.049 0.776 0.955 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.861 0.883 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.668 0.865 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.810 mp_waitany 9880 13.7 0.539 0.799 0.539 0.799 acc_transpose_blocks_kernels 18288 16.5 0.209 0.221 0.789 0.799 rs_pw_transfer_RS2PW_140 130 11.5 0.120 0.128 0.537 0.785 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.651 0.742 mp_alltoall_i22 627 13.8 0.539 0.731 0.539 0.731 mp_alltoall_d11v 2130 13.8 0.641 0.719 0.641 0.719 cp_fm_cholesky_invert 11 10.9 0.701 0.705 0.701 0.705 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.640 0.702 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.949000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.000000, yerr=1.705606 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 554.450944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.064 34.584 34.585 qs_mol_dyn_low 1 2.0 0.003 0.003 34.336 34.359 qs_forces 11 3.9 0.003 0.008 34.275 34.276 qs_energies 11 4.9 0.002 0.006 32.562 32.567 scf_env_do_scf 11 5.9 0.001 0.001 27.500 27.502 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.168 24.168 dbcsr_multiply_generic 2286 12.5 0.098 0.103 17.913 18.017 velocity_verlet 10 3.0 0.001 0.002 17.502 17.504 qs_scf_new_mos 108 7.5 0.001 0.001 15.843 15.892 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.842 15.891 ot_scf_mini 108 9.5 0.002 0.003 14.916 14.963 multiply_cannon 2286 13.5 0.231 0.259 14.532 14.864 multiply_cannon_loop 2286 14.5 0.939 0.964 13.630 13.956 ot_mini 108 10.5 0.001 0.001 9.085 9.149 multiply_cannon_multrec 27432 15.5 2.332 2.971 8.674 9.028 dbcsr_mm_accdrv_process 47916 15.9 5.377 7.385 6.248 7.443 qs_ot_get_derivative 108 11.5 0.001 0.001 7.289 7.339 rebuild_ks_matrix 119 8.3 0.000 0.000 6.492 6.543 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.491 6.542 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.762 5.806 sum_up_and_integrate 119 10.3 0.035 0.038 3.737 3.745 init_scf_run 11 5.9 0.000 0.001 3.731 3.731 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.730 3.731 integrate_v_rspace 119 11.3 0.003 0.004 3.701 3.710 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.470 3.501 calculate_rho_elec 119 8.7 0.040 0.046 3.470 3.500 qs_ot_get_p 119 10.4 0.001 0.001 3.249 3.327 init_scf_loop 11 6.9 0.000 0.003 3.314 3.314 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.770 3.102 prepare_preconditioner 11 7.9 0.000 0.000 2.482 2.492 make_preconditioner 11 8.9 0.000 0.001 2.482 2.492 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.096 2.423 calculate_first_density_matrix 1 7.0 0.000 0.003 2.382 2.384 mp_waitall_1 137007 16.6 1.766 2.320 1.766 2.320 make_m2s 4572 13.5 0.054 0.056 2.085 2.198 calculate_dm_sparse 119 9.5 0.000 0.000 2.109 2.162 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.961 2.152 apply_single 119 13.6 0.000 0.000 1.961 2.151 rs_pw_transfer 974 11.9 0.009 0.009 1.917 2.107 make_images 4572 14.5 0.269 0.328 1.978 2.090 density_rs2pw 119 9.7 0.004 0.004 1.887 2.067 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.032 2.040 pw_transfer 1439 11.6 0.066 0.070 1.958 1.992 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.865 1.904 grid_integrate_task_list 119 12.3 1.824 1.889 1.824 1.889 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.849 1.873 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.795 1.806 acc_transpose_blocks 27432 15.5 0.111 0.112 1.487 1.767 ot_diis_step 108 11.5 0.012 0.012 1.755 1.756 cp_dbcsr_syevd 50 12.0 0.003 0.004 1.726 1.726 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.618 1.630 fft_wrap_pw1pw2_140 487 13.2 0.287 0.300 1.550 1.589 jit_kernel_multiply 8 15.9 0.809 1.576 0.809 1.576 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.494 1.502 cp_fm_diag_elpa_base 50 14.0 1.460 1.475 1.491 1.500 potential_pw2rs 119 12.3 0.009 0.010 1.495 1.497 grid_collocate_task_list 119 9.7 1.222 1.338 1.222 1.338 fft3d_ps 1201 14.6 0.555 0.602 1.304 1.330 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.836 1.313 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.299 1.308 wfi_extrapolate 11 7.9 0.001 0.001 1.300 1.300 mp_sum_l 7287 12.8 0.896 1.260 0.896 1.260 cp_fm_upper_to_full 72 14.2 0.810 1.153 0.810 1.153 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.862 1.145 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.080 1.102 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.079 1.079 multiply_cannon_sync_h2d 27432 15.5 0.997 1.056 0.997 1.056 dbcsr_complete_redistribute 329 12.2 0.121 0.146 0.748 1.027 make_images_data 4572 15.5 0.045 0.048 0.818 0.936 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.697 0.899 jit_kernel_transpose 5 15.6 0.595 0.881 0.595 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.872 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.560 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.825 mp_alltoall_d11v 2130 13.8 0.690 0.774 0.690 0.774 mp_alltoall_i22 627 13.8 0.426 0.728 0.426 0.728 cp_fm_cholesky_invert 11 10.9 0.725 0.728 0.725 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.585000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.363636, yerr=5.645154 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 599.928832E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 27.758 27.759 qs_mol_dyn_low 1 2.0 0.003 0.003 27.583 27.591 qs_forces 11 3.9 0.002 0.003 27.525 27.526 qs_energies 11 4.9 0.001 0.001 25.815 25.817 scf_env_do_scf 11 5.9 0.000 0.001 20.922 20.922 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.512 18.513 velocity_verlet 10 3.0 0.001 0.001 14.101 14.104 dbcsr_multiply_generic 2286 12.5 0.090 0.095 12.174 12.312 qs_scf_new_mos 108 7.5 0.001 0.001 10.655 10.677 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.654 10.676 multiply_cannon 2286 13.5 0.231 0.247 9.783 10.071 ot_scf_mini 108 9.5 0.002 0.002 9.991 10.016 multiply_cannon_loop 2286 14.5 0.331 0.344 8.879 9.128 multiply_cannon_multrec 9144 15.5 1.579 1.820 6.024 6.234 rebuild_ks_matrix 119 8.3 0.000 0.000 5.896 5.916 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.895 5.915 ot_mini 108 10.5 0.001 0.001 5.631 5.661 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.244 5.262 dbcsr_mm_accdrv_process 12550 15.8 3.320 4.159 4.345 4.412 qs_ot_get_derivative 108 11.5 0.001 0.001 4.354 4.379 sum_up_and_integrate 119 10.3 0.038 0.042 3.607 3.613 integrate_v_rspace 119 11.3 0.003 0.010 3.569 3.576 init_scf_run 11 5.9 0.000 0.001 3.478 3.478 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.477 3.477 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.465 3.472 calculate_rho_elec 119 8.7 0.060 0.061 3.465 3.471 qs_ot_get_p 119 10.4 0.001 0.001 2.614 2.655 init_scf_loop 11 6.9 0.000 0.000 2.391 2.392 calculate_first_density_matrix 1 7.0 0.000 0.000 2.359 2.360 jit_kernel_multiply 10 15.6 0.987 2.047 0.987 2.047 pw_transfer 1439 11.6 0.066 0.070 2.005 2.014 calculate_dm_sparse 119 9.5 0.000 0.000 1.980 1.997 grid_integrate_task_list 119 12.3 1.844 1.951 1.844 1.951 mp_waitall_1 115863 16.7 1.359 1.944 1.359 1.944 density_rs2pw 119 9.7 0.004 0.004 1.782 1.923 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.911 1.922 make_m2s 4572 13.5 0.035 0.036 1.658 1.831 make_images 4572 14.5 0.267 0.303 1.568 1.738 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.733 1.735 rs_pw_transfer 974 11.9 0.008 0.008 1.600 1.723 prepare_preconditioner 11 7.9 0.000 0.000 1.673 1.677 make_preconditioner 11 8.9 0.000 0.000 1.673 1.677 fft_wrap_pw1pw2_140 487 13.2 0.366 0.374 1.583 1.596 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.565 1.590 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.543 1.544 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.509 1.511 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.457 1.476 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.406 1.413 grid_collocate_task_list 119 9.7 1.271 1.388 1.271 1.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.362 1.369 potential_pw2rs 119 12.3 0.010 0.011 1.344 1.346 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.263 1.271 cp_fm_diag_elpa_base 50 14.0 1.237 1.253 1.261 1.269 ot_diis_step 108 11.5 0.013 0.013 1.264 1.264 fft3d_ps 1201 14.6 0.560 0.574 1.244 1.255 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.213 1.213 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.146 1.168 apply_single 119 13.6 0.000 0.000 1.146 1.167 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.090 1.102 wfi_extrapolate 11 7.9 0.001 0.001 1.072 1.072 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.757 1.013 make_images_data 4572 15.5 0.039 0.043 0.770 0.966 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.919 acc_transpose_blocks 9144 15.5 0.038 0.039 0.890 0.900 mp_alltoall_d11v 2130 13.8 0.750 0.862 0.750 0.862 cp_fm_cholesky_invert 11 10.9 0.830 0.832 0.830 0.832 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.400 0.821 multiply_cannon_sync_h2d 9144 15.5 0.707 0.788 0.707 0.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.772 0.774 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.666 0.716 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.701 0.709 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.650 0.655 mp_allgather_i34 2286 14.5 0.236 0.633 0.236 0.633 mp_alltoall_z22v 1201 16.6 0.559 0.589 0.559 0.589 qs_create_task_list 11 7.9 0.000 0.000 0.542 0.568 generate_qs_task_list 11 8.9 0.189 0.211 0.542 0.568 mp_waitany 5200 13.7 0.445 0.563 0.445 0.563 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.237 0.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.759000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.363636, yerr=5.661235 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 748.314624E+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 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 41.217 41.217 qs_mol_dyn_low 1 2.0 0.003 0.003 40.847 40.854 qs_forces 11 3.9 0.002 0.003 40.789 40.789 qs_energies 11 4.9 0.001 0.001 38.822 38.825 scf_env_do_scf 11 5.9 0.001 0.001 32.628 32.628 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.960 24.961 velocity_verlet 10 3.0 0.001 0.001 22.922 22.927 dbcsr_multiply_generic 2286 12.5 0.098 0.101 17.283 17.387 qs_scf_new_mos 108 7.5 0.001 0.001 15.363 15.466 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.362 15.465 multiply_cannon 2286 13.5 0.304 0.312 13.651 14.421 ot_scf_mini 108 9.5 0.002 0.002 14.296 14.400 multiply_cannon_loop 2286 14.5 0.344 0.350 12.428 13.220 multiply_cannon_multrec 9144 15.5 3.378 4.689 8.708 8.819 ot_mini 108 10.5 0.001 0.001 8.560 8.679 init_scf_loop 11 6.9 0.000 0.000 7.642 7.644 rebuild_ks_matrix 119 8.3 0.000 0.000 7.177 7.325 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.177 7.325 prepare_preconditioner 11 7.9 0.000 0.000 6.698 6.711 make_preconditioner 11 8.9 0.000 0.000 6.698 6.711 qs_ot_get_derivative 108 11.5 0.001 0.001 6.567 6.670 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.483 6.617 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.248 6.594 dbcsr_mm_accdrv_process 12550 15.8 4.393 6.100 5.210 6.452 cp_fm_upper_to_full 72 14.2 3.156 4.554 3.156 4.554 init_scf_run 11 5.9 0.000 0.001 4.188 4.188 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.187 4.187 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.155 4.160 calculate_rho_elec 119 8.7 0.118 0.121 4.154 4.160 sum_up_and_integrate 119 10.3 0.064 0.065 3.932 3.938 integrate_v_rspace 119 11.3 0.003 0.004 3.867 3.874 mp_waitall_1 94719 16.7 2.204 3.201 2.204 3.201 qs_ot_get_p 119 10.4 0.001 0.001 3.045 3.189 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.432 2.814 calculate_first_density_matrix 1 7.0 0.000 0.000 2.769 2.770 dbcsr_complete_redistribute 329 12.2 0.295 0.300 1.958 2.755 pw_transfer 1439 11.6 0.069 0.069 2.614 2.620 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.515 2.522 mp_alltoall_i22 627 13.8 1.583 2.486 1.583 2.486 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.667 2.467 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.116 2.361 apply_single 119 13.6 0.000 0.000 2.116 2.361 make_m2s 4572 13.5 0.038 0.038 2.136 2.288 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.444 2.238 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.218 2.219 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.284 2.204 make_images 4572 14.5 0.353 0.387 2.015 2.167 calculate_dm_sparse 119 9.5 0.000 0.000 2.123 2.141 fft_wrap_pw1pw2_140 487 13.2 0.617 0.623 2.131 2.139 density_rs2pw 119 9.7 0.004 0.004 2.101 2.118 grid_integrate_task_list 119 12.3 2.041 2.048 2.041 2.048 ot_diis_step 108 11.5 0.014 0.015 1.969 1.969 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.804 1.924 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.835 1.835 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.747 1.747 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.633 1.686 mp_sum_l 7287 12.8 0.953 1.658 0.953 1.658 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.569 1.570 fft3d_ps 1201 14.6 0.595 0.606 1.531 1.539 grid_collocate_task_list 119 9.7 1.474 1.483 1.474 1.483 rs_pw_transfer 974 11.9 0.009 0.009 1.400 1.429 cp_fm_cholesky_invert 11 10.9 1.362 1.365 1.362 1.365 potential_pw2rs 119 12.3 0.014 0.014 1.356 1.358 wfi_extrapolate 11 7.9 0.001 0.001 1.357 1.357 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.302 1.302 cp_fm_diag_elpa_base 50 14.0 1.157 1.210 1.300 1.301 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.037 1.215 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.201 1.208 make_images_data 4572 15.5 0.043 0.046 0.975 1.161 jit_kernel_multiply 6 15.8 0.790 1.134 0.790 1.134 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.112 1.130 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.085 1.096 multiply_cannon_sync_h2d 9144 15.5 1.043 1.046 1.043 1.046 mp_alltoall_d11v 2130 13.8 1.008 1.035 1.008 1.035 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.983 1.021 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.826 0.970 qs_create_task_list 11 7.9 0.000 0.000 0.932 0.946 generate_qs_task_list 11 8.9 0.369 0.388 0.932 0.946 acc_transpose_blocks 9144 15.5 0.038 0.039 0.901 0.934 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.906 0.920 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.217000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.090909, yerr=12.637999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 497.946624E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2439263. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.062 81.731 81.747 qs_mol_dyn_low 1 2.0 0.003 0.003 81.452 81.461 qs_forces 11 3.9 0.005 0.006 81.382 81.382 qs_energies 11 4.9 0.001 0.002 78.515 78.531 scf_env_do_scf 11 5.9 0.001 0.002 68.973 68.976 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 63.588 63.588 dbcsr_multiply_generic 2055 12.4 0.106 0.108 51.026 51.257 qs_scf_new_mos 99 7.5 0.000 0.001 46.356 46.486 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.355 46.485 ot_scf_mini 99 9.5 0.002 0.002 43.981 44.106 multiply_cannon 2055 13.4 0.227 0.316 42.744 43.604 multiply_cannon_loop 2055 14.4 1.519 1.560 41.696 42.712 velocity_verlet 10 3.0 0.001 0.001 42.336 42.337 ot_mini 99 10.5 0.001 0.001 25.922 26.018 qs_ot_get_derivative 99 11.5 0.001 0.001 19.209 19.303 multiply_cannon_multrec 49320 15.4 12.498 13.112 17.340 17.897 rebuild_ks_matrix 110 8.3 0.000 0.000 14.515 14.636 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.514 14.636 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.722 12.835 mp_waitall_1 220248 16.4 11.029 11.938 11.029 11.938 multiply_cannon_sync_h2d 49320 15.4 10.332 10.880 10.332 10.880 qs_ot_get_p 110 10.4 0.001 0.001 9.595 9.721 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.240 7.802 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.164 7.702 apply_single 110 13.6 0.000 0.001 7.163 7.702 init_scf_run 11 5.9 0.000 0.001 7.501 7.501 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.501 7.501 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.392 7.468 sum_up_and_integrate 110 10.3 0.036 0.043 7.067 7.082 integrate_v_rspace 110 11.3 0.003 0.004 7.031 7.054 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.512 6.653 calculate_rho_elec 110 8.6 0.021 0.026 6.511 6.652 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.519 6.560 ot_diis_step 99 11.5 0.005 0.006 6.523 6.523 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.635 5.635 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.416 5.487 init_scf_loop 11 6.9 0.000 0.000 5.359 5.360 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.074 5.101 cp_fm_diag_elpa_base 48 14.0 5.060 5.087 5.072 5.099 dbcsr_mm_accdrv_process 87628 16.1 1.903 1.977 4.719 4.976 mp_sum_l 6594 12.7 4.051 4.802 4.051 4.802 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.311 4.314 rs_pw_transfer 902 11.9 0.012 0.013 3.660 4.213 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.817 4.083 wfi_extrapolate 11 7.9 0.001 0.001 4.056 4.056 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.960 4.017 density_rs2pw 110 9.6 0.005 0.005 3.335 3.916 make_m2s 4110 13.4 0.061 0.066 3.776 3.892 calculate_dm_sparse 110 9.5 0.001 0.001 3.742 3.858 make_images 4110 14.4 0.176 0.191 3.680 3.799 grid_integrate_task_list 110 12.3 3.246 3.390 3.246 3.390 calculate_first_density_matrix 1 7.0 0.000 0.001 3.351 3.363 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.199 3.244 prepare_preconditioner 11 7.9 0.000 0.000 3.208 3.227 make_preconditioner 11 8.9 0.000 0.000 3.208 3.227 pw_transfer 1331 11.6 0.055 0.067 3.081 3.157 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 2.993 3.070 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.999 3.049 fft_wrap_pw1pw2_140 451 13.1 0.450 0.496 2.557 2.643 potential_pw2rs 110 12.3 0.006 0.007 2.600 2.622 jit_kernel_multiply 13 15.9 2.538 2.609 2.538 2.609 acc_transpose_blocks 49320 15.4 0.208 0.216 2.198 2.483 mp_waitany 14300 13.8 1.835 2.480 1.835 2.480 mp_alltoall_d11v 2046 13.8 1.994 2.427 1.994 2.427 grid_collocate_task_list 110 9.6 2.083 2.293 2.083 2.293 fft3d_ps 1111 14.6 0.792 0.879 2.189 2.241 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.922 1.948 mp_sum_d 3889 11.9 1.309 1.877 1.309 1.877 make_images_data 4110 15.4 0.043 0.046 1.747 1.866 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.503 1.801 cp_fm_cholesky_invert 11 10.9 1.755 1.759 1.755 1.759 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.646 1.673 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.747000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.000000, yerr=2.696799 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 585.142272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2690344. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.043 69.180 69.181 qs_mol_dyn_low 1 2.0 0.003 0.006 68.611 68.621 qs_forces 11 3.9 0.008 0.009 68.542 68.543 qs_energies 11 4.9 0.001 0.002 65.197 65.200 scf_env_do_scf 11 5.9 0.001 0.001 56.458 56.461 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.896 48.897 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.714 37.932 velocity_verlet 10 3.0 0.001 0.002 36.088 36.089 qs_scf_new_mos 99 7.5 0.001 0.001 32.595 32.757 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.595 32.756 multiply_cannon 2055 13.4 0.227 0.250 31.278 32.457 ot_scf_mini 99 9.5 0.003 0.004 30.949 31.121 multiply_cannon_loop 2055 14.4 0.924 0.947 29.996 30.908 ot_mini 99 10.5 0.001 0.001 18.071 18.244 multiply_cannon_multrec 24660 15.4 7.627 9.807 13.875 15.825 rebuild_ks_matrix 110 8.3 0.000 0.000 13.810 13.949 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.810 13.948 qs_ot_get_derivative 99 11.5 0.009 0.080 12.181 12.356 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.151 12.264 mp_waitall_1 176588 16.5 7.534 10.219 7.534 10.219 multiply_cannon_sync_h2d 24660 15.4 7.090 8.296 7.090 8.296 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.132 7.658 init_scf_loop 11 6.9 0.000 0.000 7.527 7.528 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.531 7.150 apply_single 110 13.6 0.000 0.001 6.530 7.150 sum_up_and_integrate 110 10.3 0.052 0.059 6.622 6.635 integrate_v_rspace 110 11.3 0.002 0.003 6.570 6.583 qs_ot_get_p 110 10.4 0.001 0.001 6.250 6.450 init_scf_run 11 5.9 0.000 0.001 6.388 6.389 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.388 6.389 dbcsr_mm_accdrv_process 52282 16.1 4.682 5.883 6.088 6.362 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.127 6.136 calculate_rho_elec 110 8.6 0.040 0.048 6.126 6.135 ot_diis_step 99 11.5 0.015 0.049 5.841 5.841 prepare_preconditioner 11 7.9 0.000 0.000 5.503 5.529 make_preconditioner 11 8.9 0.000 0.000 5.503 5.529 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.597 5.323 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.064 5.225 make_m2s 4110 13.4 0.057 0.059 4.112 4.596 make_images 4110 14.4 0.398 0.442 4.004 4.484 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.439 4.459 cp_dbcsr_syevd 48 12.0 0.004 0.009 4.018 4.018 density_rs2pw 110 9.6 0.004 0.005 3.255 3.770 pw_transfer 1331 11.6 0.067 0.074 3.615 3.747 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.507 3.643 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.522 3.524 wfi_extrapolate 11 7.9 0.001 0.001 3.506 3.506 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.464 3.481 cp_fm_diag_elpa_base 48 14.0 3.413 3.428 3.456 3.472 rs_pw_transfer 902 11.9 0.012 0.013 2.929 3.446 grid_integrate_task_list 110 12.3 3.163 3.370 3.163 3.370 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.203 3.250 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.111 3.202 fft_wrap_pw1pw2_140 451 13.1 0.520 0.535 3.010 3.144 calculate_dm_sparse 110 9.5 0.001 0.001 2.969 2.998 make_images_data 4110 15.4 0.047 0.050 2.325 2.832 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.070 2.824 calculate_first_density_matrix 1 7.0 0.000 0.000 2.795 2.796 fft3d_ps 1111 14.6 1.108 1.326 2.480 2.628 cp_fm_cholesky_invert 11 10.9 2.539 2.546 2.539 2.546 mp_sum_l 6594 12.7 1.698 2.533 1.698 2.533 grid_collocate_task_list 110 9.6 2.050 2.479 2.050 2.479 potential_pw2rs 110 12.3 0.008 0.009 2.405 2.420 mp_alltoall_d11v 2046 13.8 1.769 2.017 1.769 2.017 jit_kernel_multiply 10 16.2 1.051 1.998 1.051 1.998 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.933 1.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.847 1.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.815 1.837 acc_transpose_blocks 24660 15.4 0.111 0.115 1.790 1.812 mp_waitany 10164 13.8 1.221 1.753 1.221 1.753 multiply_cannon_metrocomm4 22605 15.4 0.077 0.080 0.774 1.693 mp_allgather_i34 2055 14.4 0.606 1.659 0.606 1.659 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.556 1.566 mp_irecv_dv 57340 16.2 0.648 1.483 0.648 1.483 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.221 0.963 1.474 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.181000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=552.818182, yerr=7.132744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 658.173952E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.089 60.639 60.640 qs_mol_dyn_low 1 2.0 0.003 0.004 60.225 60.291 qs_forces 11 3.9 0.003 0.004 60.011 60.012 qs_energies 11 4.9 0.003 0.016 56.792 56.794 scf_env_do_scf 11 5.9 0.001 0.002 48.141 48.143 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.603 39.604 velocity_verlet 10 3.0 0.001 0.001 32.354 32.360 dbcsr_multiply_generic 2055 12.4 0.105 0.108 28.672 28.905 qs_scf_new_mos 99 7.5 0.001 0.001 24.625 24.701 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.625 24.701 multiply_cannon 2055 13.4 0.213 0.224 22.556 23.857 ot_scf_mini 99 9.5 0.002 0.002 23.419 23.507 multiply_cannon_loop 2055 14.4 0.618 0.632 21.361 22.398 ot_mini 99 10.5 0.001 0.001 13.567 13.659 rebuild_ks_matrix 110 8.3 0.000 0.000 12.344 12.456 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.344 12.456 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.865 10.966 multiply_cannon_multrec 16440 15.4 3.913 4.924 9.919 10.829 mp_waitall_1 139946 16.5 6.989 10.174 6.989 10.174 qs_ot_get_derivative 99 11.5 0.001 0.001 9.175 9.264 init_scf_loop 11 6.9 0.008 0.060 8.505 8.507 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.480 7.503 prepare_preconditioner 11 7.9 0.002 0.015 6.628 6.650 make_preconditioner 11 8.9 0.000 0.001 6.626 6.650 sum_up_and_integrate 110 10.3 0.060 0.061 6.501 6.515 integrate_v_rspace 110 11.3 0.003 0.003 6.440 6.454 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.411 6.421 calculate_rho_elec 110 8.6 0.059 0.059 6.411 6.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.931 6.294 init_scf_run 11 5.9 0.000 0.001 6.187 6.187 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.187 6.187 dbcsr_mm_accdrv_process 34862 16.1 5.071 5.593 5.860 6.072 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.918 5.354 apply_single 110 13.6 0.000 0.000 4.917 5.354 qs_ot_get_p 110 10.4 0.001 0.001 5.081 5.215 density_rs2pw 110 9.6 0.005 0.005 3.453 4.685 make_m2s 4110 13.4 0.049 0.051 4.018 4.409 multiply_cannon_sync_h2d 16440 15.4 3.706 4.369 3.706 4.369 ot_diis_step 99 11.5 0.010 0.011 4.366 4.366 make_images 4110 14.4 0.392 0.510 3.904 4.293 rs_pw_transfer 902 11.9 0.010 0.011 3.035 4.274 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.386 4.036 pw_transfer 1331 11.6 0.066 0.073 3.611 3.618 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.546 3.550 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.504 3.513 grid_integrate_task_list 110 12.3 3.169 3.403 3.169 3.403 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.214 3.214 fft_wrap_pw1pw2_140 451 13.1 0.637 0.646 3.046 3.058 wfi_extrapolate 11 7.9 0.001 0.001 2.960 2.960 mp_waitany 17072 13.8 1.614 2.909 1.614 2.909 make_images_data 4110 15.4 0.043 0.047 2.337 2.818 grid_collocate_task_list 110 9.6 2.135 2.779 2.135 2.779 calculate_dm_sparse 110 9.5 0.001 0.001 2.736 2.766 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.749 2.751 hybrid_alltoall_any 4261 16.3 0.106 0.378 2.096 2.738 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.699 2.707 calculate_first_density_matrix 1 7.0 0.002 0.015 2.706 2.707 cp_fm_diag_elpa_base 48 14.0 2.634 2.658 2.697 2.706 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.194 1.321 2.553 cp_fm_cholesky_invert 11 10.9 2.520 2.526 2.520 2.526 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.442 2.479 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.872 2.358 fft3d_ps 1111 14.6 1.091 1.101 2.326 2.342 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.251 2.304 potential_pw2rs 110 12.3 0.011 0.011 2.228 2.232 mp_irecv_dv 48980 15.7 0.803 2.230 0.803 2.230 mp_alltoall_d11v 2046 13.8 1.742 2.191 1.742 2.191 mp_sum_l 6594 12.7 1.531 2.105 1.531 2.105 qs_energies_init_hamiltonians 11 5.9 0.019 0.047 2.035 2.035 dbcsr_complete_redistribute 325 12.2 0.327 0.374 1.421 1.888 cp_fm_upper_to_full 70 14.2 1.413 1.789 1.413 1.789 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.709 1.723 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.675 1.688 acc_transpose_blocks 16440 15.4 0.075 0.079 1.290 1.578 mp_allgather_i34 2055 14.4 0.519 1.554 0.519 1.554 cp_fm_cholesky_decompose 22 10.9 1.520 1.542 1.520 1.542 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.375 1.501 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.986 1.443 rs_gather_matrices 110 12.3 0.233 0.258 0.964 1.342 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.220 1.226 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.640000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.727273, yerr=8.068918 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 723.968000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.087 66.602 66.611 qs_mol_dyn_low 1 2.0 0.003 0.006 65.926 65.935 qs_forces 11 3.9 0.004 0.011 65.862 65.862 qs_energies 11 4.9 0.003 0.019 62.440 62.444 scf_env_do_scf 11 5.9 0.001 0.001 53.108 53.112 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.412 41.413 velocity_verlet 10 3.0 0.001 0.001 36.814 36.816 dbcsr_multiply_generic 2055 12.4 0.112 0.117 29.983 30.206 qs_scf_new_mos 99 7.5 0.001 0.001 26.381 26.507 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.381 26.506 ot_scf_mini 99 9.5 0.003 0.003 24.759 24.891 multiply_cannon 2055 13.4 0.246 0.266 23.097 24.484 multiply_cannon_loop 2055 14.4 0.885 0.906 21.688 22.205 ot_mini 99 10.5 0.001 0.001 14.113 14.267 multiply_cannon_multrec 24660 15.4 4.244 7.016 12.716 14.248 rebuild_ks_matrix 110 8.3 0.000 0.000 12.291 12.421 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.291 12.421 init_scf_loop 11 6.9 0.003 0.021 11.657 11.660 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.858 10.972 qs_ot_get_derivative 99 11.5 0.001 0.001 9.965 10.100 prepare_preconditioner 11 7.9 0.000 0.001 9.812 9.832 make_preconditioner 11 8.9 0.000 0.001 9.812 9.832 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.022 9.497 dbcsr_mm_accdrv_process 52304 16.0 7.003 8.354 8.324 9.403 init_scf_run 11 5.9 0.000 0.001 6.557 6.558 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.557 6.558 sum_up_and_integrate 110 10.3 0.068 0.072 6.538 6.552 mp_waitall_1 121746 16.5 4.661 6.496 4.661 6.496 integrate_v_rspace 110 11.3 0.006 0.033 6.471 6.483 qs_rho_update_rho_low 110 7.6 0.001 0.007 6.282 6.293 calculate_rho_elec 110 8.6 0.079 0.091 6.280 6.292 qs_ot_get_p 110 10.4 0.001 0.002 5.669 5.834 make_m2s 4110 13.4 0.059 0.063 5.251 5.600 make_images 4110 14.4 0.575 0.697 5.110 5.456 cp_fm_upper_to_full 70 14.2 3.355 4.874 3.355 4.874 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.019 4.146 apply_single 110 13.6 0.000 0.000 4.019 4.146 ot_diis_step 99 11.5 0.011 0.011 4.111 4.111 pw_transfer 1331 11.6 0.066 0.075 3.895 3.952 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.914 3.929 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.788 3.850 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.720 3.799 dbcsr_complete_redistribute 325 12.2 0.425 0.482 2.646 3.797 density_rs2pw 110 9.6 0.005 0.009 3.160 3.673 grid_integrate_task_list 110 12.3 3.258 3.493 3.258 3.493 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.442 3.443 multiply_cannon_sync_h2d 24660 15.4 3.199 3.441 3.199 3.441 calculate_dm_sparse 110 9.5 0.001 0.001 3.313 3.343 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.242 3.308 calculate_first_density_matrix 1 7.0 0.000 0.002 3.296 3.299 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.131 3.268 fft_wrap_pw1pw2_140 451 13.1 0.668 0.689 3.162 3.207 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.411 3.152 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.095 3.097 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 2.985 3.057 wfi_extrapolate 11 7.9 0.001 0.001 2.975 2.975 make_images_data 4110 15.4 0.046 0.050 2.647 2.963 rs_pw_transfer 902 11.9 0.010 0.011 2.387 2.910 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.781 2.907 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.844 2.855 cp_fm_diag_elpa_base 48 14.0 2.699 2.757 2.842 2.854 mp_alltoall_i22 605 13.7 1.678 2.849 1.678 2.849 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.654 2.699 cp_fm_cholesky_invert 11 10.9 2.606 2.614 2.606 2.614 fft3d_ps 1111 14.6 1.085 1.120 2.549 2.587 grid_collocate_task_list 110 9.6 2.184 2.496 2.184 2.496 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 2.266 2.267 potential_pw2rs 110 12.3 0.013 0.013 2.200 2.214 mp_alltoall_d11v 2046 13.8 1.921 2.147 1.921 2.147 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.927 1.944 jit_kernel_multiply 10 15.8 0.986 1.862 0.986 1.862 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.687 1.723 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.697 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.692 acc_transpose_blocks 24660 15.4 0.104 0.107 1.660 1.681 cp_fm_cholesky_decompose 22 10.9 1.633 1.680 1.633 1.680 mp_waitany 13376 13.8 1.080 1.580 1.080 1.580 mp_allgather_i34 2055 14.4 0.502 1.575 0.502 1.575 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.840 1.504 mp_sum_l 6594 12.7 0.997 1.493 0.997 1.493 mp_irecv_dv 62702 16.1 0.741 1.421 0.741 1.421 rs_pw_transfer_RS2PW_140 121 11.5 0.168 0.176 0.831 1.360 mp_alltoall_z22v 1111 16.6 1.249 1.334 1.249 1.334 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.611000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=687.272727, yerr=7.580968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 828.772352E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 54.880 54.881 qs_mol_dyn_low 1 2.0 0.003 0.003 54.339 54.349 qs_forces 11 3.9 0.012 0.012 54.274 54.275 qs_energies 11 4.9 0.001 0.002 50.583 50.588 scf_env_do_scf 11 5.9 0.000 0.001 42.086 42.086 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.528 34.528 velocity_verlet 10 3.0 0.001 0.001 30.770 30.773 dbcsr_multiply_generic 2055 12.4 0.103 0.108 22.360 22.519 qs_scf_new_mos 99 7.5 0.001 0.001 19.574 19.605 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.573 19.605 ot_scf_mini 99 9.5 0.002 0.002 18.346 18.363 multiply_cannon 2055 13.4 0.247 0.259 17.100 18.311 multiply_cannon_loop 2055 14.4 0.321 0.334 15.757 16.055 rebuild_ks_matrix 110 8.3 0.000 0.000 11.866 11.894 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.865 11.894 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.524 10.549 ot_mini 99 10.5 0.001 0.001 10.102 10.114 multiply_cannon_multrec 8220 15.4 3.199 4.498 7.403 8.341 init_scf_loop 11 6.9 0.000 0.000 7.513 7.514 mp_waitall_1 103326 16.6 5.765 7.423 5.765 7.423 sum_up_and_integrate 110 10.3 0.080 0.081 6.474 6.487 integrate_v_rspace 110 11.3 0.003 0.004 6.394 6.406 qs_ot_get_derivative 99 11.5 0.001 0.001 6.330 6.347 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.238 6.252 calculate_rho_elec 110 8.6 0.114 0.115 6.237 6.252 prepare_preconditioner 11 7.9 0.000 0.000 5.863 5.864 make_preconditioner 11 8.9 0.000 0.000 5.863 5.864 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.456 5.528 init_scf_run 11 5.9 0.000 0.001 5.320 5.320 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.320 5.320 dbcsr_mm_accdrv_process 17442 15.9 2.814 3.754 4.075 4.954 qs_ot_get_p 110 10.4 0.001 0.001 4.362 4.376 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.976 4.374 make_m2s 4110 13.4 0.039 0.040 4.088 4.369 make_images 4110 14.4 0.636 0.691 3.959 4.242 pw_transfer 1331 11.6 0.066 0.073 3.886 3.902 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.778 3.798 ot_diis_step 99 11.5 0.012 0.012 3.753 3.753 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.674 3.691 apply_single 110 13.6 0.000 0.000 3.673 3.691 grid_integrate_task_list 110 12.3 3.340 3.537 3.340 3.537 density_rs2pw 110 9.6 0.004 0.005 2.992 3.364 fft_wrap_pw1pw2_140 451 13.1 0.833 0.848 3.283 3.314 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.094 3.097 multiply_cannon_sync_h2d 8220 15.4 2.926 3.014 2.926 3.014 cp_fm_cholesky_invert 11 10.9 2.800 2.803 2.800 2.803 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.803 2.803 make_images_data 4110 15.4 0.038 0.044 2.345 2.733 qs_energies_init_hamiltonians 11 5.9 0.007 0.042 2.729 2.730 wfi_extrapolate 11 7.9 0.001 0.001 2.714 2.714 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.275 2.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.549 2.551 grid_collocate_task_list 110 9.6 2.278 2.541 2.278 2.541 rs_pw_transfer 902 11.9 0.010 0.011 2.148 2.535 calculate_first_density_matrix 1 7.0 0.000 0.000 2.511 2.513 calculate_dm_sparse 110 9.5 0.001 0.001 2.448 2.485 fft3d_ps 1111 14.6 1.135 1.180 2.321 2.346 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.310 2.317 cp_fm_diag_elpa_base 48 14.0 2.255 2.275 2.308 2.315 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.125 2.133 potential_pw2rs 110 12.3 0.016 0.016 2.025 2.031 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.777 1.984 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.868 1.878 mp_alltoall_d11v 2046 13.8 1.571 1.875 1.571 1.875 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.710 1.719 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.670 1.673 cp_fm_cholesky_decompose 22 10.9 1.644 1.658 1.644 1.658 qs_env_update_s_mstruct 11 6.9 0.001 0.003 1.526 1.656 mp_allgather_i34 2055 14.4 0.526 1.618 0.526 1.618 dbcsr_complete_redistribute 325 12.2 0.634 0.678 1.480 1.579 mp_waitany 9240 13.8 1.088 1.502 1.088 1.502 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.431 jit_kernel_multiply 8 15.8 0.948 1.371 0.948 1.371 qs_create_task_list 11 7.9 0.000 0.000 1.226 1.332 generate_qs_task_list 11 8.9 0.376 0.443 1.226 1.332 rs_gather_matrices 110 12.3 0.325 0.363 0.939 1.232 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.178 0.816 1.203 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.090 1.110 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.881000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.818182, yerr=9.870232 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 1.359778E+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 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.059 88.189 88.190 qs_mol_dyn_low 1 2.0 0.003 0.003 87.508 87.517 qs_forces 11 3.9 0.043 0.080 87.445 87.446 qs_energies 11 4.9 0.029 0.030 83.267 83.291 scf_env_do_scf 11 5.9 0.000 0.001 72.198 72.198 velocity_verlet 10 3.0 0.001 0.001 55.598 55.604 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.023 44.024 dbcsr_multiply_generic 2055 12.4 0.118 0.124 28.967 29.120 init_scf_loop 11 6.9 0.000 0.000 28.105 28.108 prepare_preconditioner 11 7.9 0.000 0.000 26.091 26.104 make_preconditioner 11 8.9 0.000 0.000 26.091 26.104 qs_scf_new_mos 99 7.5 0.001 0.001 25.931 26.022 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.930 26.021 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.381 25.573 ot_scf_mini 99 9.5 0.002 0.002 24.196 24.272 multiply_cannon 2055 13.4 0.346 0.365 21.848 22.519 multiply_cannon_loop 2055 14.4 0.341 0.344 20.048 20.423 cp_fm_upper_to_full 70 14.2 12.829 18.268 12.829 18.268 rebuild_ks_matrix 110 8.3 0.000 0.001 13.998 14.092 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.998 14.092 ot_mini 99 10.5 0.001 0.001 13.674 13.751 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.682 12.767 dbcsr_complete_redistribute 325 12.2 1.029 1.061 7.262 10.469 multiply_cannon_multrec 8220 15.4 4.365 4.553 10.008 10.131 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.262 9.461 qs_ot_get_derivative 99 11.5 0.001 0.001 9.023 9.102 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.695 8.867 mp_alltoall_i22 605 13.7 5.342 8.512 5.342 8.512 mp_waitall_1 84994 16.7 7.353 8.362 7.353 8.362 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.638 7.673 calculate_rho_elec 110 8.6 0.225 0.226 7.637 7.672 sum_up_and_integrate 110 10.3 0.150 0.151 7.246 7.259 integrate_v_rspace 110 11.3 0.004 0.004 7.096 7.108 init_scf_run 11 5.9 0.000 0.001 6.722 6.722 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.721 6.722 dbcsr_mm_accdrv_process 11614 15.7 3.392 4.304 5.501 5.777 make_m2s 4110 13.4 0.043 0.044 5.133 5.679 make_images 4110 14.4 0.880 0.931 4.946 5.490 qs_ot_get_p 110 10.4 0.001 0.001 5.382 5.479 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.910 5.475 apply_single 110 13.6 0.000 0.000 4.910 5.475 cp_fm_cholesky_invert 11 10.9 5.113 5.117 5.113 5.117 pw_transfer 1331 11.6 0.075 0.076 4.978 4.988 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.676 4.975 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 4.860 4.871 ot_diis_step 99 11.5 0.016 0.017 4.633 4.633 fft_wrap_pw1pw2_140 451 13.1 1.336 1.341 4.248 4.258 multiply_cannon_sync_h2d 8220 15.4 3.949 3.956 3.949 3.956 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.844 3.851 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.725 3.748 grid_integrate_task_list 110 12.3 3.661 3.717 3.661 3.717 density_rs2pw 110 9.6 0.004 0.005 3.682 3.708 hybrid_alltoall_any 4261 16.3 0.256 0.552 2.798 3.534 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.861 3.529 make_images_data 4110 15.4 0.042 0.044 2.806 3.429 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.410 3.410 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.321 3.326 wfi_extrapolate 11 7.9 0.001 0.001 3.323 3.323 calculate_first_density_matrix 1 7.0 0.000 0.000 3.280 3.283 calculate_dm_sparse 110 9.5 0.001 0.001 3.101 3.122 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.838 2.839 cp_fm_diag_elpa_base 48 14.0 2.302 2.494 2.837 2.837 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.800 2.834 fft3d_ps 1111 14.6 1.297 1.308 2.788 2.801 grid_collocate_task_list 110 9.6 2.629 2.654 2.629 2.654 potential_pw2rs 110 12.3 0.021 0.022 2.371 2.377 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.332 2.372 rs_pw_transfer 902 11.9 0.010 0.011 2.288 2.337 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.248 2.301 jit_kernel_multiply 10 15.3 1.907 2.198 1.907 2.198 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.152 mp_alltoall_d11v 2046 13.8 1.874 1.966 1.874 1.966 cp_fm_cholesky_decompose 22 10.9 1.936 1.959 1.936 1.959 qs_create_task_list 11 7.9 0.000 0.000 1.894 1.940 generate_qs_task_list 11 8.9 0.732 0.790 1.894 1.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.916 1.925 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.724 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.190000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1218.363636, yerr=64.513949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 626.683904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57511. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2146788. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.074 208.211 208.212 qs_mol_dyn_low 1 2.0 0.003 0.003 206.969 206.982 qs_forces 11 3.9 0.027 0.031 206.757 206.757 qs_energies 11 4.9 0.001 0.002 201.137 201.152 scf_env_do_scf 11 5.9 0.001 0.001 184.136 184.139 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 163.963 163.965 dbcsr_multiply_generic 2507 12.6 0.177 0.181 124.944 125.972 qs_scf_new_mos 117 7.6 0.001 0.001 124.125 124.558 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.125 124.558 velocity_verlet 10 3.0 0.001 0.001 123.154 123.156 ot_scf_mini 117 9.6 0.003 0.003 117.529 117.971 multiply_cannon 2507 13.6 0.239 0.247 101.874 104.110 multiply_cannon_loop 2507 14.6 2.079 2.133 99.727 101.787 ot_mini 117 10.6 0.001 0.001 65.730 66.165 multiply_cannon_multrec 60168 15.6 33.749 36.049 42.001 44.311 qs_ot_get_derivative 117 11.6 0.001 0.001 40.420 40.851 rebuild_ks_matrix 128 8.3 0.001 0.001 33.716 34.173 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.715 34.172 mp_waitall_1 267128 16.5 28.382 31.015 28.382 31.015 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.268 30.669 multiply_cannon_sync_h2d 60168 15.6 27.867 30.219 27.867 30.219 qs_ot_get_p 128 10.4 0.001 0.001 29.735 30.216 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.765 25.935 apply_single 128 13.6 0.001 0.001 24.765 25.935 ot_diis_step 117 11.6 0.007 0.008 25.083 25.084 qs_ot_p2m_diag 83 11.4 0.079 0.091 23.235 23.303 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.539 20.539 init_scf_loop 11 6.9 0.000 0.000 20.102 20.104 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.318 19.624 multiply_cannon_metrocomm3 60168 15.6 0.114 0.117 16.079 18.036 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.590 17.620 cp_fm_diag_elpa_base 83 14.4 17.525 17.554 17.586 17.615 prepare_preconditioner 11 7.9 0.000 0.000 15.476 15.506 make_preconditioner 11 8.9 0.000 0.000 15.476 15.506 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.732 14.891 sum_up_and_integrate 128 10.3 0.089 0.108 14.097 14.114 integrate_v_rspace 128 11.3 0.004 0.005 14.008 14.027 make_m2s 5014 13.6 0.105 0.114 13.550 13.905 make_images 5014 14.6 0.400 0.419 13.368 13.733 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.518 13.640 calculate_rho_elec 128 8.7 0.046 0.065 13.517 13.639 init_scf_run 11 5.9 0.000 0.002 12.746 12.746 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.745 12.746 density_rs2pw 128 9.7 0.006 0.007 6.926 10.210 mp_sum_l 7950 12.9 8.666 9.881 8.666 9.881 wfi_extrapolate 11 7.9 0.001 0.001 9.075 9.075 rs_pw_transfer 1046 11.9 0.017 0.019 5.603 8.894 cp_fm_cholesky_invert 11 10.9 8.782 8.790 8.782 8.790 calculate_dm_sparse 128 9.5 0.001 0.001 8.399 8.471 dbcsr_mm_accdrv_process 124484 16.2 3.106 3.236 7.806 8.366 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 5.952 8.136 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.819 8.011 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.850 7.970 pw_transfer 1547 11.6 0.074 0.090 7.680 7.851 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.477 7.652 make_images_data 5014 15.6 0.067 0.073 6.674 7.569 grid_integrate_task_list 128 12.3 6.949 7.470 6.949 7.470 hybrid_alltoall_any 5200 16.5 0.291 2.259 5.867 7.110 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.869 6.880 fft_wrap_pw1pw2_140 523 13.2 1.273 1.319 6.599 6.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.002 6.120 mp_waitany 16020 13.9 2.701 5.946 2.701 5.946 grid_collocate_task_list 128 9.7 4.547 5.772 4.547 5.772 fft3d_ps 1291 14.7 2.140 2.791 5.245 5.578 mp_alltoall_d11v 2415 14.1 4.232 5.530 4.232 5.530 rs_pw_transfer_RS2PW_140 139 11.5 0.285 0.302 2.137 5.436 cp_fm_cholesky_decompose 22 10.9 4.542 4.557 4.542 4.557 potential_pw2rs 128 12.3 0.009 0.011 4.511 4.531 mp_sum_d 4479 12.1 3.526 4.384 3.526 4.384 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.212000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.272727, yerr=6.311475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.692800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58220. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 5381923. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.149 0.222 192.989 192.991 qs_mol_dyn_low 1 2.0 0.005 0.021 191.131 191.276 qs_forces 11 3.9 0.009 0.041 190.908 190.933 qs_energies 11 4.9 0.096 0.396 184.028 184.044 scf_env_do_scf 11 5.9 0.003 0.021 165.257 165.267 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 132.801 132.803 velocity_verlet 10 3.0 0.005 0.034 119.006 119.014 dbcsr_multiply_generic 2507 12.6 0.218 0.276 97.918 99.135 qs_scf_new_mos 117 7.6 0.001 0.001 92.808 93.239 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.807 93.238 ot_scf_mini 117 9.6 0.004 0.005 88.020 88.536 multiply_cannon 2507 13.6 0.476 0.530 78.100 82.443 multiply_cannon_loop 2507 14.6 1.249 1.285 74.925 77.411 ot_mini 117 10.6 0.001 0.002 49.342 49.830 mp_waitall_1 214728 16.6 24.103 37.986 24.103 37.986 multiply_cannon_multrec 30084 15.6 22.244 27.265 31.867 37.434 rebuild_ks_matrix 128 8.3 0.001 0.001 33.414 33.938 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.023 33.414 33.937 init_scf_loop 11 6.9 0.001 0.004 32.367 32.370 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.978 30.474 multiply_cannon_metrocomm3 30084 15.6 0.091 0.096 15.436 29.019 prepare_preconditioner 11 7.9 0.000 0.000 27.922 27.990 make_preconditioner 11 8.9 0.000 0.001 27.922 27.990 qs_ot_get_derivative 117 11.6 0.002 0.013 27.447 27.945 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.642 27.153 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.000 23.142 apply_single 128 13.6 0.001 0.001 22.000 23.141 ot_diis_step 117 11.6 0.016 0.029 21.725 21.732 multiply_cannon_sync_h2d 30084 15.6 19.527 21.648 19.527 21.648 qs_ot_get_p 128 10.4 0.001 0.001 20.392 21.068 cp_fm_cholesky_invert 11 10.9 16.290 16.302 16.290 16.302 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.714 15.748 make_m2s 5014 13.6 0.089 0.094 14.155 15.723 make_images 5014 14.6 1.153 1.337 13.948 15.521 sum_up_and_integrate 128 10.3 0.116 0.134 14.726 14.772 integrate_v_rspace 128 11.3 0.004 0.004 14.609 14.661 cp_dbcsr_syevd 83 12.4 0.006 0.007 14.589 14.590 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.015 14.058 calculate_rho_elec 128 8.7 0.088 0.104 14.015 14.057 init_scf_run 11 5.9 0.005 0.206 13.355 13.356 scf_env_initial_rho_setup 11 6.9 0.013 0.061 13.350 13.356 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.510 11.545 cp_fm_diag_elpa_base 83 14.4 11.257 11.343 11.503 11.534 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.753 11.081 make_images_data 5014 15.6 0.065 0.074 8.641 10.615 multiply_cannon_metrocomm4 27577 15.6 0.103 0.115 3.680 10.443 mp_irecv_dv 69486 16.3 3.482 10.051 3.482 10.051 density_rs2pw 128 9.7 0.007 0.008 7.375 9.758 hybrid_alltoall_any 5200 16.5 0.342 1.502 7.318 9.750 dbcsr_mm_accdrv_process 62242 16.2 4.599 5.309 9.076 9.661 pw_transfer 1547 11.6 0.086 0.096 8.796 8.859 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.571 8.639 wfi_extrapolate 11 7.9 0.002 0.006 8.368 8.371 rs_pw_transfer 1046 11.9 0.014 0.016 5.690 8.094 grid_integrate_task_list 128 12.3 7.154 7.700 7.154 7.700 fft_wrap_pw1pw2_140 523 13.2 1.340 1.374 7.611 7.691 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.134 6.909 calculate_dm_sparse 128 9.5 0.001 0.001 6.757 6.905 cp_fm_cholesky_decompose 22 10.9 6.821 6.902 6.821 6.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.869 6.878 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.877 6.013 fft3d_ps 1291 14.7 2.818 3.004 5.874 5.937 grid_collocate_task_list 128 9.7 4.712 5.833 4.712 5.833 mp_sum_l 7950 12.9 3.870 5.572 3.870 5.572 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.283 5.343 potential_pw2rs 128 12.3 0.015 0.018 4.976 5.010 mp_waitany 11748 13.9 2.540 5.001 2.540 5.001 mp_allgather_i34 2507 14.6 1.749 4.894 1.749 4.894 calculate_first_density_matrix 1 7.0 0.055 0.243 4.784 4.802 mp_alltoall_d11v 2415 14.1 4.033 4.709 4.033 4.709 mp_sum_d 4470 12.1 3.072 4.541 3.072 4.541 rs_pw_transfer_RS2PW_140 139 11.5 0.362 0.397 2.128 4.514 dbcsr_complete_redistribute 395 12.7 0.859 0.990 3.185 4.075 dbcsr_dot_sd 1318 12.0 0.755 0.833 2.875 4.055 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.829 3.862 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.991000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.363636, yerr=5.531278 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 952.770560E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+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 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.030 176.106 176.107 qs_mol_dyn_low 1 2.0 0.003 0.003 175.582 175.596 qs_forces 11 3.9 0.004 0.005 175.396 175.401 qs_energies 11 4.9 0.001 0.002 168.806 168.815 scf_env_do_scf 11 5.9 0.001 0.001 151.784 151.784 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.949 116.950 velocity_verlet 10 3.0 0.001 0.001 112.309 112.314 dbcsr_multiply_generic 2507 12.6 0.181 0.188 81.399 82.613 qs_scf_new_mos 117 7.6 0.001 0.001 78.994 79.314 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.993 79.313 ot_scf_mini 117 9.6 0.003 0.004 74.858 75.247 multiply_cannon 2507 13.6 0.510 0.546 61.928 66.407 multiply_cannon_loop 2507 14.6 0.861 0.884 58.565 61.287 ot_mini 117 10.6 0.001 0.001 42.278 42.670 init_scf_loop 11 6.9 0.000 0.000 34.738 34.740 mp_waitall_1 170520 16.6 24.441 33.951 24.441 33.951 rebuild_ks_matrix 128 8.3 0.001 0.001 30.617 31.069 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.616 31.068 prepare_preconditioner 11 7.9 0.000 0.000 30.720 30.774 make_preconditioner 11 8.9 0.000 0.000 30.720 30.774 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.325 29.799 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.585 28.005 multiply_cannon_multrec 20056 15.6 13.516 16.869 22.267 25.710 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 14.952 24.659 qs_ot_get_derivative 117 11.6 0.001 0.002 22.631 23.021 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.744 20.647 apply_single 128 13.6 0.001 0.001 19.744 20.646 ot_diis_step 117 11.6 0.018 0.021 19.555 19.556 qs_ot_get_p 128 10.4 0.001 0.001 17.974 18.425 multiply_cannon_sync_h2d 20056 15.6 14.434 15.693 14.434 15.693 make_m2s 5014 13.6 0.080 0.085 14.511 15.663 make_images 5014 14.6 1.169 1.245 14.280 15.427 sum_up_and_integrate 128 10.3 0.133 0.145 14.484 14.511 integrate_v_rspace 128 11.3 0.004 0.005 14.350 14.381 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.166 14.212 calculate_rho_elec 128 8.7 0.130 0.146 14.165 14.211 cp_fm_cholesky_invert 11 10.9 14.194 14.202 14.194 14.202 qs_ot_p2m_diag 83 11.4 0.266 0.273 13.777 13.784 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.789 12.789 init_scf_run 11 5.9 0.000 0.001 12.021 12.021 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.021 12.021 make_images_data 5014 15.6 0.060 0.066 8.900 10.357 cp_fm_diag_elpa 83 13.4 0.000 0.001 9.684 9.701 cp_fm_diag_elpa_base 83 14.4 9.279 9.429 9.681 9.699 hybrid_alltoall_any 5200 16.5 0.435 1.965 7.915 9.675 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.405 9.370 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.864 9.139 mp_irecv_dv 50230 16.2 3.284 9.131 3.284 9.131 pw_transfer 1547 11.6 0.086 0.106 8.804 8.904 density_rs2pw 128 9.7 0.006 0.007 7.062 8.874 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.580 8.684 dbcsr_mm_accdrv_process 41502 16.2 4.415 5.001 8.194 8.323 grid_integrate_task_list 128 12.3 7.319 7.775 7.319 7.775 fft_wrap_pw1pw2_140 523 13.2 1.418 1.451 7.638 7.757 wfi_extrapolate 11 7.9 0.001 0.001 7.408 7.408 cp_fm_upper_to_full 105 14.8 5.833 7.372 5.833 7.372 cp_fm_cholesky_decompose 22 10.9 7.249 7.276 7.249 7.276 dbcsr_complete_redistribute 395 12.7 1.565 1.750 5.124 6.904 rs_pw_transfer 1046 11.9 0.014 0.014 4.972 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.116 6.121 calculate_dm_sparse 128 9.5 0.001 0.001 5.726 5.839 fft3d_ps 1291 14.7 2.743 2.968 5.695 5.774 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.978 5.757 grid_collocate_task_list 128 9.7 4.922 5.663 4.922 5.663 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.558 5.213 mp_allgather_i34 2507 14.6 1.860 4.847 1.860 4.847 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.693 4.808 mp_alltoall_d11v 2415 14.1 4.182 4.757 4.182 4.757 potential_pw2rs 128 12.3 0.021 0.023 4.609 4.618 mp_sum_l 7950 12.9 3.137 4.541 3.137 4.541 calculate_first_density_matrix 1 7.0 0.000 0.000 4.431 4.433 transfer_fm_to_dbcsr 11 9.9 0.020 0.030 2.376 4.128 mp_waitany 11748 13.9 2.288 4.105 2.288 4.105 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.991 4.009 mp_alltoall_i22 716 14.1 1.965 3.918 1.965 3.918 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.722 3.722 rs_pw_transfer_RS2PW_140 139 11.5 0.330 0.348 1.883 3.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.529 3.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.107000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=902.818182, yerr=8.397166 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243750E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806546176 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.139003E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922098320 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57632. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509574. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.178 189.224 189.230 qs_mol_dyn_low 1 2.0 0.003 0.004 188.373 188.386 qs_forces 11 3.9 0.012 0.067 188.266 188.271 qs_energies 11 4.9 0.007 0.047 181.058 181.068 scf_env_do_scf 11 5.9 0.004 0.029 163.223 163.230 velocity_verlet 10 3.0 0.001 0.001 125.068 125.071 scf_env_do_scf_inner_loop 118 6.6 0.004 0.013 116.948 116.953 dbcsr_multiply_generic 2534 12.6 0.187 0.194 79.953 80.585 qs_scf_new_mos 118 7.6 0.001 0.001 79.261 79.500 qs_scf_loop_do_ot 118 8.6 0.002 0.007 79.260 79.499 ot_scf_mini 118 9.6 0.004 0.011 74.720 74.967 multiply_cannon 2534 13.6 0.566 0.603 55.222 58.358 multiply_cannon_loop 2534 14.6 1.194 1.222 51.652 53.074 init_scf_loop 11 6.9 0.002 0.015 46.153 46.154 ot_mini 118 10.6 0.001 0.002 42.285 42.520 prepare_preconditioner 11 7.9 0.000 0.000 41.992 42.012 make_preconditioner 11 8.9 0.000 0.002 41.992 42.012 make_full_inverse_cholesky 11 9.9 0.013 0.023 35.637 40.674 multiply_cannon_multrec 30408 15.6 14.447 19.562 26.778 31.543 rebuild_ks_matrix 129 8.3 0.001 0.001 30.078 30.279 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 30.077 30.278 mp_waitall_1 149452 16.7 17.274 27.585 17.274 27.585 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.131 27.321 qs_ot_get_derivative 118 11.6 0.002 0.003 22.534 22.779 make_m2s 5068 13.6 0.096 0.100 20.426 21.314 make_images 5068 14.6 1.966 2.320 20.119 21.011 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.121 19.631 apply_single 129 13.6 0.001 0.001 19.121 19.630 ot_diis_step 118 11.6 0.019 0.030 19.618 19.620 qs_ot_get_p 129 10.4 0.001 0.001 18.613 18.874 cp_fm_upper_to_full 106 14.8 11.452 16.877 11.452 16.877 cp_fm_cholesky_invert 11 10.9 15.977 15.986 15.977 15.986 multiply_cannon_metrocomm3 30408 15.6 0.048 0.050 6.242 15.432 sum_up_and_integrate 129 10.3 0.141 0.152 14.767 14.791 qs_ot_p2m_diag 84 11.4 0.347 0.395 14.616 14.669 integrate_v_rspace 129 11.3 0.005 0.032 14.625 14.653 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.567 14.613 calculate_rho_elec 129 8.7 0.175 0.191 14.567 14.612 cp_dbcsr_syevd 84 12.4 0.005 0.006 13.337 13.340 multiply_cannon_sync_h2d 30408 15.6 11.868 13.024 11.868 13.024 dbcsr_complete_redistribute 397 12.7 1.505 1.617 8.899 12.691 make_images_data 5068 15.6 0.064 0.069 11.095 12.670 dbcsr_mm_accdrv_process 62840 16.2 7.479 8.437 11.898 12.492 init_scf_run 11 5.9 0.000 0.001 11.901 11.902 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.900 11.902 hybrid_alltoall_any 5255 16.5 0.526 2.184 9.824 11.660 copy_fm_to_dbcsr 210 11.7 0.003 0.011 7.571 11.328 cp_fm_diag_elpa 84 13.4 0.000 0.001 10.261 10.273 cp_fm_diag_elpa_base 84 14.4 9.301 9.628 10.254 10.265 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.331 9.983 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.364 9.562 mp_alltoall_i22 720 14.1 5.583 9.373 5.583 9.373 pw_transfer 1559 11.6 0.087 0.101 9.183 9.267 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.954 9.045 density_rs2pw 129 9.7 0.006 0.007 6.960 8.243 fft_wrap_pw1pw2_140 527 13.2 1.571 1.606 7.994 8.096 grid_integrate_task_list 129 12.3 7.513 8.004 7.513 8.004 wfi_extrapolate 11 7.9 0.001 0.001 7.714 7.714 cp_fm_cholesky_decompose 22 10.9 7.490 7.604 7.490 7.604 multiply_cannon_metrocomm4 25340 15.6 0.079 0.090 2.735 7.120 mp_irecv_dv 76866 16.2 2.588 6.853 2.588 6.853 calculate_dm_sparse 129 9.5 0.001 0.001 6.236 6.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.903 5.991 fft3d_ps 1301 14.7 2.854 2.956 5.854 5.924 rs_pw_transfer 1054 12.0 0.013 0.014 4.595 5.919 grid_collocate_task_list 129 9.7 5.118 5.832 5.118 5.832 mp_alltoall_d11v 2429 14.1 4.824 5.824 4.824 5.824 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.845 4.919 potential_pw2rs 129 12.3 0.024 0.024 4.500 4.511 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.455 4.456 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.330 4.415 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.170 4.243 calculate_first_density_matrix 1 7.0 0.001 0.006 3.988 3.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.230000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.818182, yerr=17.872367 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 5.865088E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.524330E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350154512 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101675. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58352. MP_Allreduce 11057 1167. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.041 173.333 173.333 qs_mol_dyn_low 1 2.0 0.003 0.003 172.486 172.498 qs_forces 11 3.9 0.058 0.069 172.379 172.391 qs_energies 11 4.9 0.001 0.002 164.729 164.780 scf_env_do_scf 11 5.9 0.001 0.001 147.385 147.389 velocity_verlet 10 3.0 0.001 0.001 113.540 113.544 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 111.718 111.720 dbcsr_multiply_generic 2507 12.6 0.180 0.192 74.073 74.575 qs_scf_new_mos 117 7.6 0.001 0.001 73.372 73.519 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.371 73.519 ot_scf_mini 117 9.6 0.003 0.004 68.926 69.010 multiply_cannon 2507 13.6 0.597 0.634 54.258 58.888 multiply_cannon_loop 2507 14.6 0.444 0.454 49.481 50.285 ot_mini 117 10.6 0.001 0.001 39.177 39.263 init_scf_loop 11 6.9 0.000 0.000 35.522 35.524 mp_waitall_1 125778 16.7 25.683 32.754 25.683 32.754 prepare_preconditioner 11 7.9 0.000 0.000 31.565 31.598 make_preconditioner 11 8.9 0.000 0.000 31.565 31.598 rebuild_ks_matrix 128 8.3 0.001 0.001 29.845 29.977 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.844 29.976 make_full_inverse_cholesky 11 9.9 0.017 0.027 29.446 29.700 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.088 27.211 multiply_cannon_multrec 10028 15.6 10.392 14.234 17.810 20.740 ot_diis_step 117 11.6 0.020 0.021 19.809 19.809 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.414 19.781 apply_single 128 13.6 0.001 0.001 19.414 19.780 qs_ot_get_derivative 117 11.6 0.002 0.002 19.300 19.382 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.115 19.239 make_m2s 5014 13.6 0.068 0.072 16.296 18.766 make_images 5014 14.6 2.300 2.784 15.989 18.463 cp_fm_cholesky_invert 11 10.9 18.040 18.046 18.040 18.046 qs_ot_get_p 128 10.4 0.001 0.001 16.769 16.864 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.151 15.189 calculate_rho_elec 128 8.7 0.256 0.267 15.150 15.188 sum_up_and_integrate 128 10.3 0.181 0.190 14.923 14.975 integrate_v_rspace 128 11.3 0.004 0.004 14.742 14.803 qs_ot_p2m_diag 83 11.4 0.496 0.502 13.078 13.096 make_images_data 5014 15.6 0.053 0.061 10.023 12.742 multiply_cannon_sync_h2d 10028 15.6 11.612 12.562 11.612 12.562 hybrid_alltoall_any 5200 16.5 0.822 3.616 9.905 12.110 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.959 11.961 init_scf_run 11 5.9 0.000 0.001 10.434 10.434 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.434 10.434 pw_transfer 1547 11.6 0.086 0.096 9.381 9.409 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.156 9.192 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.990 9.000 cp_fm_diag_elpa_base 83 14.4 8.754 8.831 8.986 8.997 grid_integrate_task_list 128 12.3 7.751 8.285 7.751 8.285 fft_wrap_pw1pw2_140 523 13.2 1.915 1.950 8.122 8.158 cp_fm_cholesky_decompose 22 10.9 7.865 7.971 7.865 7.971 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.740 7.796 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.313 7.727 dbcsr_mm_accdrv_process 20762 16.1 2.606 3.405 7.047 7.671 density_rs2pw 128 9.7 0.006 0.006 6.769 7.649 mp_allgather_i34 2507 14.6 2.973 7.605 2.973 7.605 wfi_extrapolate 11 7.9 0.001 0.001 7.412 7.412 calculate_dm_sparse 128 9.5 0.001 0.001 6.182 6.258 grid_collocate_task_list 128 9.7 5.393 6.069 5.393 6.069 mp_alltoall_d11v 2415 14.1 4.860 5.992 4.860 5.992 fft3d_ps 1291 14.7 2.757 2.838 5.632 5.664 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.366 5.420 dbcsr_complete_redistribute 395 12.7 2.117 2.179 5.021 5.262 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.146 5.153 rs_pw_transfer 1046 11.9 0.013 0.013 4.141 5.053 potential_pw2rs 128 12.3 0.027 0.028 4.380 4.392 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.118 4.150 multiply_cannon_metrocomm4 7521 15.6 0.023 0.027 1.864 4.084 mp_irecv_dv 28860 15.9 1.828 4.015 1.828 4.015 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.566 3.891 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.668 3.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.660 3.670 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.313 3.610 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.449 3.543 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.333000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1406.363636, yerr=42.268133 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/20/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 11.785520E+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 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 3.131965E+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 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 58831. MP_Allreduce 11137 1501. MP_Sync 87 MP_Alltoall 1724 36993750. 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 0.053 0.132 297.389 297.389 qs_mol_dyn_low 1 2.0 0.003 0.003 296.642 296.653 qs_forces 11 3.9 0.005 0.006 296.552 296.555 qs_energies 11 4.9 0.002 0.002 287.241 287.250 scf_env_do_scf 11 5.9 0.001 0.001 264.638 264.649 velocity_verlet 10 3.0 0.001 0.001 214.059 214.066 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 140.935 140.938 init_scf_loop 11 6.9 0.000 0.000 123.451 123.454 prepare_preconditioner 11 7.9 0.000 0.000 118.415 118.442 make_preconditioner 11 8.9 0.000 0.000 118.415 118.442 make_full_inverse_cholesky 11 9.9 0.034 0.039 94.927 115.714 qs_scf_new_mos 118 7.6 0.001 0.001 90.027 90.154 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.026 90.154 ot_scf_mini 118 9.6 0.004 0.004 85.203 85.253 dbcsr_multiply_generic 2527 12.6 0.212 0.224 83.725 84.352 cp_fm_upper_to_full 105 14.8 52.937 75.578 52.937 75.578 multiply_cannon 2527 13.6 0.710 0.744 59.201 59.649 multiply_cannon_loop 2527 14.6 0.477 0.485 55.485 56.893 ot_mini 118 10.6 0.001 0.001 44.466 44.515 dbcsr_complete_redistribute 395 12.7 3.995 4.036 29.071 41.783 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.731 38.508 rebuild_ks_matrix 129 8.3 0.001 0.001 38.396 38.451 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 38.396 38.450 transfer_fm_to_dbcsr 11 9.9 0.030 0.033 23.440 36.043 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.486 35.538 mp_alltoall_i22 718 14.1 21.281 34.222 21.281 34.222 cp_fm_cholesky_invert 11 10.9 33.017 33.023 33.017 33.023 mp_waitall_1 104500 16.8 28.365 31.929 28.365 31.929 qs_ot_get_p 129 10.4 0.001 0.001 25.099 25.136 qs_ot_get_derivative 118 11.6 0.002 0.002 24.042 24.090 qs_ot_p2m_diag 83 11.4 0.879 0.884 21.118 21.147 make_m2s 5054 13.6 0.077 0.079 19.895 20.881 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.470 20.490 calculate_rho_elec 129 8.7 0.485 0.485 20.470 20.490 make_images 5054 14.6 3.774 3.905 19.414 20.406 ot_diis_step 118 11.6 0.022 0.023 20.396 20.396 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 19.055 20.300 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.582 19.831 apply_single 129 13.6 0.001 0.001 19.581 19.831 sum_up_and_integrate 129 10.3 0.324 0.326 19.644 19.724 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.404 19.406 integrate_v_rspace 129 11.3 0.004 0.004 19.320 19.400 multiply_cannon_multrec 10108 15.6 10.548 12.279 18.116 18.181 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.257 16.258 cp_fm_diag_elpa_base 83 14.4 11.841 13.446 16.254 16.254 multiply_cannon_sync_h2d 10108 15.6 15.772 15.786 15.772 15.786 init_scf_run 11 5.9 0.000 0.001 12.508 12.509 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.507 12.509 hybrid_alltoall_any 5240 16.5 1.310 3.058 10.703 12.261 make_images_data 5054 15.6 0.061 0.065 10.576 12.197 pw_transfer 1559 11.6 0.094 0.095 11.373 11.383 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 11.134 11.140 mp_alltoall_d11v 2423 14.1 8.453 10.241 8.453 10.241 fft_wrap_pw1pw2_140 527 13.2 3.099 3.125 9.912 9.920 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.299 9.341 dbcsr_mm_accdrv_process 20926 16.1 3.929 5.641 7.328 9.109 wfi_extrapolate 11 7.9 0.001 0.001 9.094 9.094 cp_fm_cholesky_decompose 22 10.9 8.786 8.895 8.786 8.895 grid_integrate_task_list 129 12.3 8.655 8.806 8.655 8.806 density_rs2pw 129 9.7 0.006 0.006 7.932 8.073 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.018 8.019 calculate_dm_sparse 129 9.5 0.001 0.001 6.674 6.793 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.464 6.532 grid_collocate_task_list 129 9.7 6.416 6.454 6.416 6.454 rs_scatter_matrices 140 9.7 3.624 4.575 6.075 6.305 fft3d_ps 1301 14.7 2.866 2.890 6.168 6.197 copy_dbcsr_to_fm 186 11.8 0.004 0.004 5.998 6.046 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.389000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2844.545455, yerr=160.546999 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 1.255993E+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 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.053 84.911 84.913 qs_energies 1 2.0 0.000 0.000 84.352 84.366 ls_scf 1 3.0 0.000 0.000 83.442 83.456 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.458 72.627 multiply_cannon 111 7.7 0.017 0.020 55.822 57.043 multiply_cannon_loop 111 8.7 0.208 0.220 52.412 53.844 ls_scf_main 1 4.0 0.000 0.000 52.187 52.187 density_matrix_trs4 2 5.0 0.002 0.003 46.710 46.802 ls_scf_init_scf 1 4.0 0.000 0.000 28.195 28.196 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.108 27.163 mp_waitall_1 11031 10.9 22.406 25.719 22.406 25.719 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.973 24.991 multiply_cannon_multrec 2664 9.7 8.166 8.778 15.558 17.206 multiply_cannon_sync_h2d 2664 9.7 13.484 15.393 13.484 15.393 make_m2s 222 7.7 0.008 0.011 12.903 13.401 make_images 222 8.7 0.100 0.109 12.881 13.381 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.734 12.526 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.473 8.295 make_images_data 222 9.7 0.004 0.005 7.531 8.052 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.644 7.013 7.993 hybrid_alltoall_any 227 10.6 0.215 1.841 6.481 7.805 dbcsr_mm_accdrv_process_sort 4760 11.4 6.304 7.251 6.304 7.251 calculate_norms 4752 9.8 5.516 6.148 5.516 6.148 mp_sum_l 887 5.1 3.211 5.367 3.211 5.367 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.174 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.403 4.253 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.058 3.544 mp_irecv_dv 6231 10.9 2.040 3.521 2.040 3.521 make_images_sizes 222 9.7 0.000 0.000 0.675 3.376 mp_alltoall_i44 222 10.7 0.675 3.376 0.675 3.376 arnoldi_extremal 4 6.8 0.000 0.001 3.237 3.267 arnoldi_normal_ev 4 7.8 0.001 0.003 3.237 3.267 build_subspace 16 8.4 0.009 0.012 3.147 3.149 ls_scf_post 1 4.0 0.000 0.000 3.061 3.074 ls_scf_store_result 1 5.0 0.000 0.000 2.864 2.917 dbcsr_special_finalize 555 9.7 0.005 0.006 2.342 2.682 dbcsr_merge_single_wm 555 10.7 0.460 0.582 2.333 2.674 make_images_pack 222 9.7 2.208 2.624 2.210 2.625 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.313 2.568 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.479 2.070 2.481 dbcsr_sort_data 658 11.4 2.131 2.451 2.131 2.451 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.247 2.343 buffer_matrices_ensure_size 222 8.7 1.751 2.028 1.751 2.028 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.760 1.761 rebuild_ks_matrix 3 7.3 0.000 0.000 1.751 1.752 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.751 1.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.913000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1128.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 2.099094E+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 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.192 0.212 92.067 92.068 qs_energies 1 2.0 0.007 0.030 91.188 91.212 ls_scf 1 3.0 0.001 0.003 89.863 89.870 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.259 75.594 multiply_cannon 111 7.7 0.029 0.045 53.358 56.859 ls_scf_main 1 4.0 0.001 0.017 54.809 54.822 multiply_cannon_loop 111 8.7 0.116 0.122 50.061 53.237 density_matrix_trs4 2 5.0 0.004 0.030 49.120 49.289 ls_scf_init_scf 1 4.0 0.000 0.001 31.199 31.217 mp_waitall_1 9105 10.9 21.108 30.107 21.108 30.107 ls_scf_init_matrix_S 1 5.0 0.001 0.005 29.988 30.094 matrix_sqrt_Newton_Schulz 2 6.5 0.028 0.217 27.580 27.593 multiply_cannon_multrec 1332 9.7 13.134 17.026 22.415 27.466 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.833 20.914 make_m2s 222 7.7 0.006 0.007 15.265 16.001 make_images 222 8.7 1.582 1.962 15.235 15.972 dbcsr_mm_accdrv_process 4041 10.4 0.294 0.449 8.880 10.467 dbcsr_mm_accdrv_process_sort 4041 11.4 8.475 10.017 8.475 10.017 make_images_data 222 9.7 0.004 0.004 8.748 9.671 hybrid_alltoall_any 227 10.6 0.523 2.473 8.164 9.191 mp_sum_l 887 5.1 5.335 8.127 5.335 8.127 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.219 7.483 mp_irecv_dv 3311 11.0 3.199 7.426 3.199 7.426 calculate_norms 2376 9.8 5.988 6.789 5.988 6.789 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.120 6.436 multiply_cannon_sync_h2d 1332 9.7 4.876 6.317 4.876 6.317 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.015 5.247 arnoldi_extremal 4 6.8 0.000 0.001 4.585 4.605 arnoldi_normal_ev 4 7.8 0.001 0.005 4.584 4.604 build_subspace 16 8.4 0.014 0.021 4.332 4.334 ls_scf_post 1 4.0 0.001 0.003 3.854 3.862 ls_scf_store_result 1 5.0 0.000 0.000 3.267 3.368 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.109 3.343 dbcsr_matrix_vector_mult_local 304 10.0 2.732 3.213 2.734 3.215 ls_scf_dm_to_ks 2 5.0 0.000 0.010 2.573 2.680 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.161 2.642 make_images_pack 222 9.7 2.036 2.417 2.038 2.420 mp_allgather_i34 111 8.7 0.982 2.407 0.982 2.407 dbcsr_sort_data 436 11.2 1.804 2.015 1.804 2.015 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.867 1.887 rebuild_ks_matrix 3 7.3 0.000 0.000 1.854 1.873 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.026 1.854 1.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.068000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 2.712228E+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 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.123 0.742 95.625 95.626 qs_energies 1 2.0 0.000 0.000 94.133 94.136 ls_scf 1 3.0 0.000 0.001 92.698 92.703 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.246 77.550 ls_scf_main 1 4.0 0.000 0.001 58.053 58.059 multiply_cannon 111 7.7 0.039 0.095 53.240 57.285 multiply_cannon_loop 111 8.7 0.101 0.111 49.713 53.920 density_matrix_trs4 2 5.0 0.002 0.005 52.172 52.353 mp_waitall_1 7281 11.0 24.322 35.070 24.322 35.070 ls_scf_init_scf 1 4.0 0.000 0.002 31.049 31.051 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.764 29.846 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.354 27.369 multiply_cannon_multrec 888 9.7 12.551 15.309 21.135 24.503 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.266 24.279 make_m2s 222 7.7 0.006 0.007 17.090 18.324 make_images 222 8.7 1.972 2.304 17.051 18.282 hybrid_alltoall_any 227 10.6 0.619 2.852 9.443 10.825 make_images_data 222 9.7 0.003 0.004 9.801 10.803 mp_sum_l 887 5.1 5.377 9.833 5.377 9.833 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.413 8.116 9.346 dbcsr_mm_accdrv_process_sort 3754 11.4 7.747 8.932 7.747 8.932 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.209 8.012 multiply_cannon_sync_h2d 888 9.7 6.064 7.936 6.064 7.936 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.503 7.383 mp_irecv_dv 2335 11.1 2.488 7.340 2.488 7.340 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.877 7.057 arnoldi_extremal 4 6.8 0.000 0.002 5.138 5.157 arnoldi_normal_ev 4 7.8 0.002 0.014 5.138 5.157 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.869 5.087 build_subspace 16 8.4 0.014 0.021 4.801 4.821 calculate_norms 1584 9.8 4.251 4.620 4.251 4.620 mp_allgather_i34 111 8.7 1.373 3.846 1.373 3.846 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.410 3.748 ls_scf_post 1 4.0 0.000 0.001 3.596 3.601 dbcsr_matrix_vector_mult_local 304 10.0 3.007 3.584 3.009 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.280 3.375 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.813 2.915 make_images_sizes 222 9.7 0.000 0.000 0.956 2.175 dbcsr_sort_data 325 11.1 1.890 2.175 1.890 2.175 mp_alltoall_i44 222 10.7 0.956 2.175 0.956 2.175 make_images_pack 222 9.7 1.815 2.088 1.817 2.091 dbcsr_data_release 9322 10.9 1.297 1.959 1.297 1.959 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.932 1.934 rebuild_ks_matrix 3 7.3 0.000 0.000 1.913 1.915 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.011 1.913 1.915 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.626000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2190.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 3.368141E+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 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.047 97.960 97.961 qs_energies 1 2.0 0.000 0.000 97.372 97.377 ls_scf 1 3.0 0.000 0.000 95.253 95.258 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.872 79.108 ls_scf_main 1 4.0 0.000 0.000 59.162 59.163 multiply_cannon 111 7.7 0.051 0.158 51.852 56.445 density_matrix_trs4 2 5.0 0.002 0.003 53.078 53.226 multiply_cannon_loop 111 8.7 0.115 0.126 46.836 50.243 ls_scf_init_scf 1 4.0 0.000 0.000 32.580 32.583 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.361 31.446 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.908 28.921 mp_waitall_1 6369 11.0 23.035 28.791 23.035 28.791 multiply_cannon_multrec 1332 9.7 14.132 17.921 22.043 24.858 make_m2s 222 7.7 0.007 0.007 21.131 22.533 make_images 222 8.7 3.138 3.600 21.081 22.485 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.510 17.483 make_images_data 222 9.7 0.004 0.004 11.828 13.384 hybrid_alltoall_any 227 10.6 0.793 3.733 11.186 12.898 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.397 7.550 9.088 dbcsr_mm_accdrv_process_sort 3641 11.4 7.177 8.673 7.177 8.673 mp_sum_l 887 5.1 4.306 8.314 4.306 8.314 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.341 6.650 multiply_cannon_sync_h2d 1332 9.7 5.512 6.289 5.512 6.289 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.074 6.134 mp_irecv_dv 3229 10.9 2.050 6.050 2.050 6.050 arnoldi_extremal 4 6.8 0.000 0.001 5.164 5.180 arnoldi_normal_ev 4 7.8 0.001 0.005 5.164 5.179 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.525 5.102 build_subspace 16 8.4 0.014 0.021 4.822 4.831 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.544 4.722 calculate_norms 2376 9.8 4.229 4.615 4.229 4.615 mp_allgather_i34 111 8.7 2.139 4.451 2.139 4.451 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.558 3.854 dbcsr_matrix_vector_mult_local 304 10.0 3.172 3.672 3.174 3.674 ls_scf_post 1 4.0 0.000 0.000 3.511 3.515 dbcsr_sort_data 658 11.4 3.077 3.338 3.077 3.338 dbcsr_special_finalize 555 9.7 0.006 0.007 2.830 3.143 dbcsr_merge_single_wm 555 10.7 0.539 0.650 2.822 3.135 ls_scf_store_result 1 5.0 0.000 0.000 2.989 3.045 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.978 3.040 dbcsr_data_release 10477 10.7 1.577 2.425 1.577 2.425 qs_energies_init_hamiltonians 1 3.0 0.029 0.043 2.112 2.112 dbcsr_finalize 304 7.8 0.049 0.061 1.799 1.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.961000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2762.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 4.676514E+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 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.069 100.647 100.648 qs_energies 1 2.0 0.000 0.000 99.190 99.194 ls_scf 1 3.0 0.000 0.000 97.268 97.278 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.586 78.797 ls_scf_main 1 4.0 0.000 0.000 62.447 62.448 multiply_cannon 111 7.7 0.082 0.181 55.654 60.324 density_matrix_trs4 2 5.0 0.002 0.003 55.277 55.357 multiply_cannon_loop 111 8.7 0.070 0.081 51.181 53.553 mp_waitall_1 5436 11.0 27.016 32.678 27.016 32.678 ls_scf_init_scf 1 4.0 0.000 0.000 31.135 31.140 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.684 29.726 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.514 27.528 multiply_cannon_multrec 444 9.7 14.064 16.360 21.157 23.614 make_m2s 222 7.7 0.004 0.005 17.855 20.301 make_images 222 8.7 3.731 4.434 17.794 20.241 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.413 16.406 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.234 15.168 make_images_data 222 9.7 0.003 0.004 10.169 12.450 hybrid_alltoall_any 227 10.6 0.788 3.765 9.905 12.281 multiply_cannon_sync_h2d 444 9.7 6.569 7.999 6.569 7.999 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.339 6.795 7.934 dbcsr_mm_accdrv_process_sort 3003 11.4 6.482 7.588 6.482 7.588 mp_allgather_i34 111 8.7 2.707 7.060 2.707 7.060 arnoldi_extremal 4 6.8 0.018 0.031 5.840 5.856 arnoldi_normal_ev 4 7.8 0.003 0.013 5.822 5.855 build_subspace 16 8.4 0.015 0.020 5.424 5.432 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.615 4.780 mp_sum_l 887 5.1 3.113 4.740 3.113 4.740 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.143 4.356 dbcsr_matrix_vector_mult_local 304 10.0 3.672 4.156 3.674 4.158 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.644 3.866 mp_irecv_dv 1241 11.2 1.625 3.818 1.625 3.818 ls_scf_post 1 4.0 0.000 0.000 3.685 3.690 calculate_norms 792 9.8 3.528 3.664 3.528 3.664 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.128 3.657 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.543 3.647 make_images_sizes 222 9.7 0.000 0.000 1.017 3.545 mp_alltoall_i44 222 10.7 1.017 3.545 1.017 3.545 ls_scf_store_result 1 5.0 0.000 0.000 3.473 3.509 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.288 dbcsr_merge_all 275 8.9 0.475 0.522 2.050 2.126 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.056 2.057 rebuild_ks_matrix 3 7.3 0.000 0.000 2.022 2.024 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.022 2.024 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.648000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3688.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_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 8.725729E+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 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.097 109.014 109.015 qs_energies 1 2.0 0.000 0.000 107.603 107.617 ls_scf 1 3.0 0.000 0.000 104.643 104.657 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.732 77.889 ls_scf_main 1 4.0 0.000 0.000 66.101 66.103 density_matrix_trs4 2 5.0 0.002 0.003 56.896 56.962 multiply_cannon 111 7.7 0.139 0.261 49.719 51.454 multiply_cannon_loop 111 8.7 0.067 0.070 46.187 47.222 ls_scf_init_scf 1 4.0 0.000 0.000 34.880 34.881 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.701 32.722 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.878 29.887 mp_waitall_1 4527 11.1 21.959 25.969 21.959 25.969 make_m2s 222 7.7 0.005 0.005 23.981 25.007 make_images 222 8.7 4.610 5.015 23.870 24.880 multiply_cannon_multrec 444 9.7 17.877 18.759 22.516 23.183 hybrid_alltoall_any 227 10.6 1.667 3.649 12.929 15.714 make_images_data 222 9.7 0.003 0.003 13.181 15.656 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.217 10.813 multiply_cannon_sync_h2d 444 9.7 8.845 8.891 8.845 8.891 arnoldi_extremal 4 6.8 0.000 0.001 7.304 7.315 arnoldi_normal_ev 4 7.8 0.003 0.009 7.303 7.314 build_subspace 16 8.4 0.026 0.036 6.755 6.763 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.389 5.552 dbcsr_matrix_vector_mult_local 304 10.0 4.974 5.293 4.977 5.296 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.292 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.147 5.239 dbcsr_mm_accdrv_process 1814 10.4 0.183 0.310 4.467 4.597 dbcsr_mm_accdrv_process_sort 1814 11.4 4.170 4.301 4.170 4.301 ls_scf_post 1 4.0 0.000 0.000 3.662 3.676 mp_allgather_i34 111 8.7 1.175 3.608 1.175 3.608 make_images_sizes 222 9.7 0.000 0.000 1.429 3.563 mp_alltoall_i44 222 10.7 1.428 3.563 1.428 3.563 ls_scf_store_result 1 5.0 0.000 0.000 3.409 3.422 calculate_norms 792 9.8 3.243 3.276 3.243 3.276 dbcsr_finalize 304 7.8 0.082 0.090 3.089 3.175 qs_ks_update_qs_env 3 6.3 0.000 0.000 3.092 3.094 rebuild_ks_matrix 3 7.3 0.000 0.000 3.025 3.027 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.013 3.025 3.027 dbcsr_merge_all 275 8.9 0.890 0.913 2.874 2.955 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.929 2.929 dbcsr_complete_redistribute 5 7.6 1.431 1.460 2.759 2.878 dbcsr_data_release 12724 10.6 2.331 2.840 2.331 2.840 matrix_ls_to_qs 2 6.0 0.000 0.000 2.424 2.557 dbcsr_sort_data 325 11.1 2.447 2.511 2.447 2.511 mp_sum_l 887 5.1 1.695 2.470 1.695 2.470 dbcsr_new_transposed 4 7.5 0.243 0.252 2.302 2.320 dbcsr_frobenius_norm 74 6.6 2.056 2.126 2.186 2.228 ls_scf_initial_guess 1 5.0 0.000 0.000 2.178 2.208 ls_scf_qs_atomic_guess 1 6.0 0.000 0.000 2.178 2.208 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.201 dbcsr_add_anytype 103 7.2 0.859 0.892 2.130 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.015000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6856.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d168495ed9e6cddd42101d7df39d7697855199e2_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 582.070272E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 161826732. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.596 227.680 227.681 qs_mol_dyn_low 1 2.0 0.005 0.041 225.816 225.868 qs_forces 5 3.8 0.006 0.098 225.636 225.648 qs_energies 5 4.8 0.002 0.083 222.545 222.566 scf_env_do_scf 5 5.8 0.008 0.414 204.932 204.935 scf_env_do_scf_inner_loop 105 6.6 0.002 0.015 178.969 178.982 qs_scf_new_mos 105 7.6 0.000 0.001 138.451 138.622 qs_scf_loop_do_ot 105 8.6 0.001 0.001 138.451 138.621 dbcsr_multiply_generic 1445 12.2 0.123 0.132 133.619 134.111 ot_scf_mini 105 9.6 0.003 0.005 128.816 129.016 multiply_cannon 1445 13.2 0.274 0.288 115.044 117.566 multiply_cannon_loop 1445 14.2 2.390 2.522 113.311 114.607 velocity_verlet 4 3.0 0.001 0.025 105.048 105.049 ot_mini 105 10.6 0.001 0.002 58.940 59.108 multiply_cannon_multrec 69360 15.2 31.861 37.116 39.512 44.491 qs_ot_get_p 112 10.4 0.001 0.001 40.277 40.787 mp_waitall_1 488190 16.1 33.114 40.282 33.114 40.282 qs_ot_get_derivative 55 11.6 0.001 0.002 37.286 37.442 multiply_cannon_sync_h2d 69360 15.2 30.883 35.369 30.883 35.369 multiply_cannon_metrocomm3 69360 15.2 0.194 0.206 24.375 31.851 qs_ot_p2m_diag 40 11.0 0.020 0.030 29.519 29.606 rebuild_ks_matrix 110 8.4 0.000 0.000 29.134 29.329 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.016 29.133 29.329 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.704 26.886 cp_dbcsr_syevd 40 12.0 0.002 0.002 26.420 26.421 init_scf_loop 7 6.6 0.000 0.005 25.924 26.251 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.943 23.175 apply_single 62 13.6 0.000 0.000 22.943 23.175 cp_fm_syevd 40 13.0 0.000 0.000 21.109 21.265 prepare_preconditioner 7 7.6 0.000 0.001 21.076 21.111 make_preconditioner 7 8.6 0.000 0.002 21.076 21.111 ot_new_cg_direction 55 11.6 0.001 0.004 20.953 20.953 qs_rho_update_rho_low 110 7.6 0.001 0.001 17.230 17.532 calculate_rho_elec 110 8.6 0.030 0.033 17.229 17.531 cp_fm_redistribute_end 40 14.0 8.231 16.421 8.235 16.422 cp_fm_syevd_base 40 14.0 8.180 16.372 8.180 16.372 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.421 14.600 init_scf_run 5 5.8 0.000 0.000 14.311 14.312 scf_env_initial_rho_setup 5 6.8 0.000 0.002 14.311 14.312 make_full_inverse_cholesky 7 9.6 0.000 0.000 13.945 14.002 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.756 13.879 mp_sum_l 4764 12.2 12.221 12.951 12.221 12.951 rs_pw_transfer 690 11.5 0.011 0.013 11.479 12.645 density_rs2pw 110 9.6 0.006 0.007 11.319 12.545 pw_transfer 1645 12.4 0.084 0.109 12.106 12.299 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 11.963 12.161 calculate_dm_sparse 110 9.5 0.000 0.001 11.554 11.702 qs_vxc_create 110 10.4 0.002 0.004 10.433 10.479 fft_wrap_pw1pw2_240 915 15.0 1.189 1.283 10.208 10.378 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 9.892 9.968 cp_fm_cholesky_invert 7 10.6 9.744 9.751 9.744 9.751 acc_transpose_blocks 69360 15.2 0.357 0.377 9.151 9.694 check_diag 80 13.5 8.592 8.892 9.500 9.671 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 9.558 9.573 calculate_first_density_matrix 1 7.0 0.000 0.003 9.494 9.511 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.119 9.183 sum_up_and_integrate 60 10.3 0.028 0.030 8.735 8.747 integrate_v_rspace 60 11.3 0.002 0.014 8.707 8.719 fft3d_pb 915 16.0 2.395 2.632 8.254 8.483 dbcsr_mm_accdrv_process 154766 15.8 3.993 4.204 7.520 8.382 multiply_cannon_metrocomm1 69360 15.2 0.090 0.094 4.226 7.740 acc_transpose_blocks_kernels 69360 16.2 0.855 0.909 7.286 7.708 xc_rho_set_and_dset_create 110 12.4 0.075 0.096 7.285 7.530 xc_vxc_pw_create 60 11.3 0.039 0.050 7.012 7.057 make_full_single_inverse 7 9.6 0.005 0.155 6.873 6.907 make_m2s 2890 13.2 0.077 0.085 6.305 6.878 jit_kernel_transpose 5 15.0 6.431 6.816 6.431 6.816 make_images 2890 14.2 0.241 0.259 6.201 6.774 xc_pw_derive 510 13.4 0.005 0.006 6.222 6.283 mp_waitany 7680 13.5 4.454 5.864 4.454 5.864 mp_alltoall_z22v 2340 17.7 5.310 5.607 5.310 5.607 potential_pw2rs 60 12.3 0.003 0.003 4.909 4.935 multiply_cannon_metrocomm4 67915 15.2 0.180 0.199 2.046 4.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=227.681000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=554.400000, yerr=2.059126 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d168495ed9e6cddd42101d7df39d7697855199e2 Summary: empty Status: OK