=== 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: 32b39844d4254f7d1efd5072dc95c1a6354e6919 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/01 job id: 42802082 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/02 job id: 42802083 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/03 job id: 42802084 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/04 job id: 42802085 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/05 job id: 42802086 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/06 job id: 42802087 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/07 job id: 42802088 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/08 job id: 42802090 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/09 job id: 42802091 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/10 job id: 42802093 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/11 job id: 42802094 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/12 job id: 42802095 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/13 job id: 42802097 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/14 job id: 42802098 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/15 job id: 42802100 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/16 job id: 42802101 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/17 job id: 42802104 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/18 job id: 42802105 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/19 job id: 42802107 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/20 job id: 42802108 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/21 job id: 42802112 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/22 job id: 42802113 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/23 job id: 42802114 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/24 job id: 42802116 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/25 job id: 42802119 --- 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/26 job id: 42802120 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.036 132.147 132.148 farming_run 1 2.0 131.681 131.687 132.121 132.124 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457017E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.025 114.757 114.757 qs_energies 1 2.0 0.000 0.000 114.566 114.569 mp2_main 1 3.0 0.000 0.000 112.730 112.733 mp2_gpw_main 1 4.0 0.019 0.025 111.916 111.918 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.084 93.560 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.540 56.016 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.166 41.872 47.324 get_2c_integrals 1 6.0 0.000 0.000 36.619 37.373 integrate_v_rspace 273 8.0 0.434 0.448 25.123 30.356 pw_transfer 6555 10.6 0.374 0.386 27.268 27.857 grid_integrate_task_list 273 9.0 20.944 26.657 20.944 26.657 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.927 26.494 fft_wrap_pw1pw2_100 2178 12.4 1.161 1.308 23.487 24.059 compute_2c_integrals 1 7.0 0.002 0.002 19.189 19.190 rpa_ri_compute_en 1 5.0 0.000 0.000 18.722 18.960 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.804 18.921 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.429 18.801 18.918 cp_fm_cholesky_decompose 12 8.2 17.381 18.119 17.381 18.119 cholesky_decomp 1 7.0 0.000 0.000 16.287 17.030 fft3d_s 5443 13.4 16.110 16.591 16.132 16.612 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.558 10.855 15.558 calculate_wavefunction 272 8.0 5.547 5.698 12.665 13.379 rpa_num_int 1 6.0 0.000 0.000 10.487 10.487 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.463 10.484 calc_mat_Q 8 8.0 0.000 0.000 9.320 9.419 contract_S_to_Q 8 9.0 0.000 0.000 8.741 8.838 calc_potential_gpw 544 9.5 0.006 0.007 8.219 8.588 parallel_gemm_fm 14 9.1 0.000 0.000 8.324 8.408 parallel_gemm_fm_cosma 14 10.1 8.324 8.408 8.324 8.408 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.174 8.390 potential_pw2rs 545 10.0 0.107 0.110 7.679 8.333 collocate_single_gaussian 272 10.0 0.041 0.043 7.421 7.670 create_integ_mat 1 6.0 0.022 0.028 7.665 7.665 array2fm 1 7.0 0.000 0.000 6.693 7.134 pw_scatter_s 2720 13.7 4.411 4.584 4.411 4.584 pw_gather_s 2722 13.2 3.874 4.217 3.874 4.217 array2fm_buffer_send 1 8.0 2.978 3.161 2.978 3.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.916914, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2730.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.041 396.645 396.646 farming_run 1 2.0 395.895 395.903 396.580 396.582 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.223950E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 26876955. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.019 0.057 209.798 209.798 qs_energies 1 2.0 0.000 0.000 209.510 209.519 scf_env_do_scf 1 3.0 0.000 0.000 106.020 106.020 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.935 104.946 rebuild_ks_matrix 4 6.0 0.000 0.000 104.934 104.944 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.069 104.934 104.944 hfx_ks_matrix 4 8.0 0.001 0.001 104.508 104.511 integrate_four_center 4 9.0 0.144 0.457 104.507 104.510 mp2_main 1 3.0 0.000 0.000 103.164 103.173 mp2_gpw_main 1 4.0 0.036 0.081 102.101 102.111 integrate_four_center_main 4 10.0 0.108 0.548 96.457 98.763 integrate_four_center_bin 265 11.0 96.349 98.749 96.349 98.749 init_scf_loop 1 4.0 0.000 0.000 91.774 91.774 mp2_ri_gpw_compute_in 1 5.0 0.073 0.151 75.007 76.077 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.031 54.522 55.593 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.266 47.413 integrate_v_rspace 95 8.0 0.398 0.562 28.647 33.622 pw_transfer 2240 10.6 0.143 0.163 29.979 30.455 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.980 29.451 grid_integrate_task_list 95 9.0 23.890 29.070 23.890 29.070 ao_to_mo_and_store_B_mult_1 91 7.0 10.571 28.833 10.571 28.833 mp2_ri_gpw_compute_en 1 5.0 0.056 0.065 26.919 28.652 fft_wrap_pw1pw2_100 730 12.4 1.306 1.499 26.667 27.121 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.884 25.202 25.212 get_2c_integrals 1 6.0 0.001 0.010 20.374 20.421 compute_2c_integrals 1 7.0 0.003 0.005 19.348 19.365 compute_2c_integrals_loop_lm 1 8.0 0.004 0.027 18.957 19.190 mp2_eri_2c_integrate_gpw 1 9.0 1.730 1.859 18.953 19.190 fft3d_s 1823 13.4 18.422 18.836 18.435 18.849 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.243 14.243 calculate_wavefunction 91 8.0 2.045 2.082 9.775 10.011 potential_pw2rs 186 10.0 0.034 0.035 8.716 9.366 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.593 8.710 9.032 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.002 8.306 8.674 mp2_ri_gpw_compute_en_comm 22 7.0 0.485 0.504 8.246 8.648 local_gemm 172 8.0 8.151 8.452 8.151 8.452 calc_potential_gpw 182 9.5 0.002 0.002 7.944 8.186 collocate_single_gaussian 91 10.0 0.020 0.078 7.909 8.147 mp_sendrecv_dm3 2068 8.0 6.319 6.694 6.319 6.694 mp2_ri_gpw_compute_en_ener 172 7.0 6.357 6.427 6.357 6.427 mp_sync 38 10.4 2.729 6.107 2.729 6.107 pw_gather_s 912 13.2 4.911 5.496 4.911 5.496 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.087049, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 450.871296E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.028 53.622 53.625 qs_mol_dyn_low 1 2.0 0.003 0.005 53.328 53.335 qs_forces 11 3.9 0.002 0.002 53.266 53.268 qs_energies 11 4.9 0.001 0.002 51.790 51.804 scf_env_do_scf 11 5.9 0.000 0.001 44.580 44.580 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.489 42.489 dbcsr_multiply_generic 2286 12.5 0.094 0.097 35.101 35.591 qs_scf_new_mos 108 7.5 0.000 0.001 32.547 32.838 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.547 32.837 ot_scf_mini 108 9.5 0.002 0.002 30.819 30.983 multiply_cannon 2286 13.5 0.181 0.189 27.366 29.024 multiply_cannon_loop 2286 14.5 1.519 1.593 26.673 28.318 velocity_verlet 10 3.0 0.001 0.001 26.166 26.167 ot_mini 108 10.5 0.001 0.001 19.914 20.154 qs_ot_get_derivative 108 11.5 0.001 0.001 16.936 17.118 mp_waitall_1 267858 16.1 9.809 16.424 9.809 16.424 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 6.383 13.648 multiply_cannon_multrec 54864 15.5 4.154 6.414 7.883 12.165 rebuild_ks_matrix 119 8.3 0.000 0.000 7.824 7.954 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.824 7.954 multiply_cannon_sync_h2d 54864 15.5 5.813 7.604 5.813 7.604 mp_sum_l 7207 12.9 5.784 7.508 5.784 7.508 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.904 7.023 qs_ot_get_p 119 10.4 0.001 0.001 6.239 6.580 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.789 6.257 init_scf_run 11 5.9 0.000 0.001 5.878 5.879 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.878 5.878 dbcsr_mm_accdrv_process 76910 16.1 1.201 1.853 3.651 5.814 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.655 5.767 sum_up_and_integrate 119 10.3 0.012 0.014 4.486 4.494 integrate_v_rspace 119 11.3 0.002 0.002 4.474 4.483 calculate_dm_sparse 119 9.5 0.000 0.000 4.081 4.242 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.040 4.194 calculate_rho_elec 119 8.7 0.011 0.016 4.039 4.193 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.566 3.994 jit_kernel_multiply 13 15.8 2.387 3.891 2.387 3.891 calculate_first_density_matrix 1 7.0 0.000 0.000 3.693 3.712 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.418 3.496 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.888 3.075 apply_single 119 13.6 0.000 0.000 2.888 3.075 rs_pw_transfer 974 11.9 0.011 0.012 2.805 2.892 ot_diis_step 108 11.5 0.006 0.006 2.695 2.696 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.604 2.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.518 2.561 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.494 2.554 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.396 2.397 cp_fm_redistribute_end 50 14.0 2.175 2.374 2.179 2.376 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.370 2.372 cp_fm_diag_elpa_base 50 14.0 0.195 2.338 0.195 2.347 acc_transpose_blocks 54864 15.5 0.235 0.250 1.805 2.241 density_rs2pw 119 9.7 0.004 0.004 2.133 2.226 grid_integrate_task_list 119 12.3 2.022 2.122 2.022 2.122 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.069 2.122 wfi_extrapolate 11 7.9 0.001 0.001 2.115 2.115 init_scf_loop 11 6.9 0.000 0.000 2.075 2.075 mp_sum_d 4125 12.0 1.383 2.018 1.383 2.018 potential_pw2rs 119 12.3 0.004 0.004 1.854 1.871 pw_transfer 1439 11.6 0.051 0.055 1.627 1.696 make_m2s 4572 13.5 0.054 0.056 1.641 1.686 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.552 1.623 make_images 4572 14.5 0.133 0.139 1.558 1.602 mp_alltoall_d11v 2130 13.8 1.332 1.571 1.332 1.571 mp_waitany 12084 13.8 1.304 1.522 1.304 1.522 fft3d_ps 1201 14.6 0.359 0.460 1.330 1.392 grid_collocate_task_list 119 9.7 1.289 1.361 1.289 1.361 fft_wrap_pw1pw2_140 487 13.2 0.080 0.092 1.196 1.269 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.751 1.155 acc_transpose_blocks_kernels 54864 16.5 0.256 0.393 0.799 1.076 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.625000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.571456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1243685. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.104 38.696 38.697 qs_mol_dyn_low 1 2.0 0.003 0.005 38.413 38.420 qs_forces 11 3.9 0.002 0.004 37.859 37.864 qs_energies 11 4.9 0.001 0.002 36.194 36.201 scf_env_do_scf 11 5.9 0.000 0.001 31.167 31.168 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.673 28.673 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.397 21.764 qs_scf_new_mos 108 7.5 0.001 0.001 19.775 20.031 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.774 20.030 ot_scf_mini 108 9.5 0.002 0.003 18.881 19.064 velocity_verlet 10 3.0 0.062 0.490 18.688 18.689 multiply_cannon 2286 13.5 0.209 0.220 16.323 17.911 multiply_cannon_loop 2286 14.5 0.900 0.973 15.178 16.622 ot_mini 108 10.5 0.001 0.003 11.823 12.064 mp_waitall_1 217478 16.2 5.998 11.134 5.998 11.134 qs_ot_get_derivative 108 11.5 0.001 0.002 9.345 9.529 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.148 9.514 multiply_cannon_multrec 27432 15.5 1.963 4.505 5.905 8.786 rebuild_ks_matrix 119 8.3 0.000 0.000 7.056 7.196 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.017 7.056 7.195 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.241 6.366 dbcsr_mm_accdrv_process 47894 16.0 3.036 5.214 3.873 5.719 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.599 4.449 qs_ot_get_p 119 10.4 0.001 0.001 4.116 4.347 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.056 4.142 apply_single 119 13.6 0.000 0.000 3.056 4.142 sum_up_and_integrate 119 10.3 0.024 0.029 4.112 4.119 integrate_v_rspace 119 11.3 0.002 0.002 4.088 4.096 mp_sum_l 7207 12.9 2.099 4.084 2.099 4.084 init_scf_run 11 5.9 0.000 0.001 3.828 3.828 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.828 3.828 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.580 3.619 calculate_rho_elec 119 8.7 0.021 0.024 3.579 3.619 multiply_cannon_sync_h2d 27432 15.5 2.172 2.859 2.172 2.859 make_m2s 4572 13.5 0.052 0.054 2.540 2.790 make_images 4572 14.5 0.201 0.238 2.451 2.697 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.676 2.695 rs_pw_transfer 974 11.9 0.010 0.011 2.492 2.578 init_scf_loop 11 6.9 0.000 0.000 2.473 2.474 ot_diis_step 108 11.5 0.011 0.013 2.426 2.428 calculate_first_density_matrix 1 7.0 0.000 0.000 2.328 2.329 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.157 2.250 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.245 2.245 calculate_dm_sparse 119 9.5 0.000 0.000 2.066 2.140 density_rs2pw 119 9.7 0.004 0.004 1.943 2.016 jit_kernel_multiply 10 16.0 0.784 1.985 0.784 1.985 grid_integrate_task_list 119 12.3 1.830 1.924 1.830 1.924 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.915 1.916 cp_fm_redistribute_end 50 14.0 1.579 1.893 1.582 1.895 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.884 1.887 cp_fm_diag_elpa_base 50 14.0 0.303 1.836 0.311 1.867 potential_pw2rs 119 12.3 0.006 0.006 1.854 1.864 pw_transfer 1439 11.6 0.063 0.066 1.744 1.774 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.687 1.724 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.654 1.686 prepare_preconditioner 11 7.9 0.000 0.000 1.583 1.610 make_preconditioner 11 8.9 0.000 0.000 1.583 1.610 make_images_data 4572 15.5 0.046 0.053 1.176 1.597 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.485 1.543 hybrid_alltoall_any 4725 16.4 0.051 0.113 1.018 1.490 acc_transpose_blocks 27432 15.5 0.109 0.115 1.185 1.480 wfi_extrapolate 11 7.9 0.001 0.003 1.450 1.450 fft3d_ps 1201 14.6 0.498 0.551 1.361 1.392 mp_allgather_i34 2286 14.5 0.586 1.382 0.586 1.382 grid_collocate_task_list 119 9.7 1.242 1.322 1.242 1.322 fft_wrap_pw1pw2_140 487 13.2 0.076 0.086 1.271 1.302 mp_alltoall_d11v 2130 13.8 1.197 1.291 1.197 1.291 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.281 1.287 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.201 1.250 mp_sum_d 4125 12.0 0.577 1.011 0.577 1.011 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.934 0.939 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.871 0.886 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.651 0.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.697000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.363636, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 526.508032E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.045 31.907 31.909 qs_mol_dyn_low 1 2.0 0.003 0.003 31.611 31.618 qs_forces 11 3.9 0.002 0.002 31.544 31.554 qs_energies 11 4.9 0.003 0.010 29.994 30.009 scf_env_do_scf 11 5.9 0.001 0.001 25.279 25.280 scf_env_do_scf_inner_loop 108 6.5 0.006 0.019 22.697 22.698 dbcsr_multiply_generic 2286 12.5 0.095 0.100 16.366 16.464 velocity_verlet 10 3.0 0.001 0.002 15.097 15.098 qs_scf_new_mos 108 7.5 0.001 0.001 14.818 14.833 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.817 14.833 ot_scf_mini 108 9.5 0.003 0.003 14.094 14.108 multiply_cannon 2286 13.5 0.198 0.204 13.025 13.837 multiply_cannon_loop 2286 14.5 0.636 0.668 12.226 13.064 ot_mini 108 10.5 0.001 0.001 8.796 8.817 qs_ot_get_derivative 108 11.5 0.001 0.001 7.293 7.307 multiply_cannon_multrec 18288 15.5 1.941 2.906 6.761 7.055 rebuild_ks_matrix 119 8.3 0.000 0.000 6.282 6.301 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.281 6.300 dbcsr_mm_accdrv_process 38222 16.0 4.077 5.324 4.735 5.621 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.552 5.569 sum_up_and_integrate 119 10.3 0.031 0.032 3.918 3.924 integrate_v_rspace 119 11.3 0.002 0.003 3.887 3.893 mp_waitall_1 169478 16.3 2.887 3.758 2.887 3.758 init_scf_run 11 5.9 0.000 0.001 3.503 3.504 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.503 3.504 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.719 3.349 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.208 3.221 calculate_rho_elec 119 8.7 0.030 0.031 3.207 3.220 qs_ot_get_p 119 10.4 0.001 0.002 3.146 3.169 init_scf_loop 11 6.9 0.001 0.005 2.566 2.568 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.469 2.372 rs_pw_transfer 974 11.9 0.009 0.009 2.252 2.365 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.056 2.361 apply_single 119 13.6 0.000 0.000 2.055 2.360 calculate_first_density_matrix 1 7.0 0.001 0.002 2.277 2.278 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.064 2.070 make_m2s 4572 13.5 0.045 0.046 1.810 1.962 grid_integrate_task_list 119 12.3 1.804 1.901 1.804 1.901 density_rs2pw 119 9.7 0.004 0.004 1.790 1.883 make_images 4572 14.5 0.192 0.209 1.724 1.874 jit_kernel_multiply 10 16.0 0.606 1.864 0.606 1.864 calculate_dm_sparse 119 9.5 0.000 0.000 1.808 1.822 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.798 1.798 prepare_preconditioner 11 7.9 0.000 0.000 1.762 1.765 make_preconditioner 11 8.9 0.000 0.001 1.762 1.765 potential_pw2rs 119 12.3 0.007 0.008 1.706 1.713 pw_transfer 1439 11.6 0.063 0.065 1.693 1.704 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.617 1.698 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.687 1.689 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.602 1.616 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.595 1.601 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.569 1.570 multiply_cannon_sync_h2d 18288 15.5 1.378 1.554 1.378 1.554 cp_fm_redistribute_end 50 14.0 1.162 1.546 1.163 1.547 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.523 1.532 cp_fm_diag_elpa_base 50 14.0 0.368 1.489 0.382 1.527 ot_diis_step 108 11.5 0.011 0.011 1.479 1.479 mp_sum_l 7207 12.9 1.143 1.423 1.143 1.423 grid_collocate_task_list 119 9.7 1.204 1.313 1.204 1.313 acc_transpose_blocks 18288 15.5 0.075 0.077 1.228 1.305 fft3d_ps 1201 14.6 0.504 0.520 1.288 1.302 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.254 1.266 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.178 1.182 wfi_extrapolate 11 7.9 0.001 0.001 1.172 1.172 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.355 1.161 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.973 0.983 make_images_data 4572 15.5 0.046 0.050 0.803 0.969 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.880 0.906 hybrid_alltoall_any 4725 16.4 0.056 0.115 0.685 0.879 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.786 0.857 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.801 0.803 mp_alltoall_d11v 2130 13.8 0.700 0.791 0.700 0.791 cp_fm_cholesky_invert 11 10.9 0.741 0.745 0.741 0.745 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.656 0.709 mp_waitany 9880 13.7 0.560 0.690 0.560 0.690 mp_alltoall_z22v 1201 16.6 0.626 0.687 0.626 0.687 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.124 0.546 0.660 jit_kernel_transpose 5 15.6 0.577 0.639 0.577 0.639 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.909000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=498.727273, yerr=4.091919 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 560.988160E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.030 34.474 34.475 qs_mol_dyn_low 1 2.0 0.003 0.003 34.283 34.291 qs_forces 11 3.9 0.002 0.002 34.214 34.224 qs_energies 11 4.9 0.003 0.007 32.515 32.529 scf_env_do_scf 11 5.9 0.001 0.001 27.525 27.526 scf_env_do_scf_inner_loop 108 6.5 0.007 0.021 24.084 24.085 dbcsr_multiply_generic 2286 12.5 0.101 0.106 18.091 18.263 velocity_verlet 10 3.0 0.001 0.002 17.409 17.411 qs_scf_new_mos 108 7.5 0.001 0.001 16.101 16.149 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.100 16.148 ot_scf_mini 108 9.5 0.003 0.003 15.168 15.214 multiply_cannon 2286 13.5 0.240 0.270 14.469 14.910 multiply_cannon_loop 2286 14.5 0.942 0.974 13.539 13.926 ot_mini 108 10.5 0.001 0.001 9.370 9.434 multiply_cannon_multrec 27432 15.5 2.334 2.993 8.605 8.932 qs_ot_get_derivative 108 11.5 0.001 0.001 7.559 7.611 dbcsr_mm_accdrv_process 47916 15.9 5.318 7.322 6.178 7.423 rebuild_ks_matrix 119 8.3 0.000 0.000 6.377 6.427 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.377 6.426 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.659 5.702 sum_up_and_integrate 119 10.3 0.035 0.038 3.679 3.689 integrate_v_rspace 119 11.3 0.002 0.002 3.644 3.654 init_scf_run 11 5.9 0.000 0.001 3.627 3.628 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.627 3.628 init_scf_loop 11 6.9 0.001 0.004 3.421 3.423 qs_ot_get_p 119 10.4 0.001 0.002 3.212 3.284 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.194 3.234 calculate_rho_elec 119 8.7 0.040 0.046 3.193 3.234 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.687 3.139 mp_waitall_1 145218 16.4 2.070 2.743 2.070 2.743 prepare_preconditioner 11 7.9 0.000 0.000 2.579 2.586 make_preconditioner 11 8.9 0.000 0.001 2.579 2.586 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.097 2.552 apply_single 119 13.6 0.000 0.000 2.097 2.552 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.174 2.510 make_m2s 4572 13.5 0.054 0.056 2.184 2.296 calculate_first_density_matrix 1 7.0 0.001 0.002 2.293 2.295 make_images 4572 14.5 0.273 0.337 2.076 2.187 calculate_dm_sparse 119 9.5 0.000 0.000 2.051 2.109 rs_pw_transfer 974 11.9 0.008 0.009 1.915 2.015 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.939 1.948 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.909 1.935 grid_integrate_task_list 119 12.3 1.818 1.907 1.818 1.907 jit_kernel_multiply 10 15.5 0.799 1.836 0.799 1.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.820 1.821 density_rs2pw 119 9.7 0.004 0.004 1.692 1.819 ot_diis_step 108 11.5 0.012 0.012 1.766 1.766 pw_transfer 1439 11.6 0.062 0.066 1.709 1.742 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.620 1.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.623 1.637 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.625 1.625 mp_sum_l 7207 12.9 1.012 1.575 1.012 1.575 potential_pw2rs 119 12.3 0.008 0.009 1.476 1.480 acc_transpose_blocks 27432 15.5 0.111 0.112 1.449 1.467 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.818 1.434 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.393 1.394 cp_fm_redistribute_end 50 14.0 0.917 1.371 0.919 1.372 fft_wrap_pw1pw2_140 487 13.2 0.084 0.094 1.328 1.367 cp_fm_diag_elpa_base 50 14.0 0.431 1.313 0.451 1.354 fft3d_ps 1201 14.6 0.532 0.582 1.301 1.326 grid_collocate_task_list 119 9.7 1.221 1.319 1.221 1.319 wfi_extrapolate 11 7.9 0.001 0.001 1.287 1.287 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.247 cp_fm_upper_to_full 72 13.5 0.823 1.145 0.823 1.145 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.103 1.113 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.095 1.110 dbcsr_complete_redistribute 329 12.2 0.124 0.178 0.801 1.082 multiply_cannon_sync_h2d 27432 15.5 0.988 1.052 0.988 1.052 make_images_data 4572 15.5 0.046 0.049 0.850 0.965 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.739 0.902 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.604 0.879 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.874 mp_alltoall_d11v 2130 13.8 0.723 0.868 0.723 0.868 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.823 0.829 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.813 mp_alltoall_i22 627 13.8 0.441 0.750 0.441 0.750 cp_fm_cholesky_invert 11 10.9 0.744 0.747 0.744 0.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.475000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.181818, yerr=3.761484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 599.887872E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.028 27.352 27.352 qs_mol_dyn_low 1 2.0 0.003 0.003 27.189 27.196 qs_forces 11 3.9 0.002 0.002 27.128 27.132 qs_energies 11 4.9 0.001 0.001 25.447 25.452 scf_env_do_scf 11 5.9 0.000 0.001 20.722 20.722 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.250 18.251 velocity_verlet 10 3.0 0.002 0.002 13.898 13.901 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.157 12.233 qs_scf_new_mos 108 7.5 0.001 0.001 10.843 10.867 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.843 10.866 ot_scf_mini 108 9.5 0.002 0.002 10.179 10.206 multiply_cannon 2286 13.5 0.231 0.238 9.664 10.071 multiply_cannon_loop 2286 14.5 0.331 0.342 8.757 8.950 multiply_cannon_multrec 9144 15.5 1.789 2.025 5.945 6.144 ot_mini 108 10.5 0.001 0.001 5.802 5.835 rebuild_ks_matrix 119 8.3 0.000 0.000 5.751 5.773 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.750 5.773 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.127 5.146 qs_ot_get_derivative 108 11.5 0.001 0.001 4.522 4.549 dbcsr_mm_accdrv_process 12550 15.8 3.082 4.085 4.056 4.135 sum_up_and_integrate 119 10.3 0.037 0.041 3.488 3.493 integrate_v_rspace 119 11.3 0.002 0.003 3.451 3.456 init_scf_run 11 5.9 0.000 0.001 3.308 3.308 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.308 3.308 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.128 3.134 calculate_rho_elec 119 8.7 0.059 0.061 3.127 3.134 qs_ot_get_p 119 10.4 0.001 0.001 2.634 2.674 init_scf_loop 11 6.9 0.000 0.000 2.452 2.454 calculate_first_density_matrix 1 7.0 0.000 0.000 2.211 2.211 mp_waitall_1 121218 16.5 1.512 2.046 1.512 2.046 grid_integrate_task_list 119 12.3 1.838 1.951 1.838 1.951 make_m2s 4572 13.5 0.034 0.036 1.713 1.885 make_images 4572 14.5 0.268 0.303 1.623 1.794 jit_kernel_multiply 10 15.6 0.935 1.751 0.935 1.751 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.726 1.729 calculate_dm_sparse 119 9.5 0.000 0.000 1.702 1.721 prepare_preconditioner 11 7.9 0.000 0.000 1.687 1.691 make_preconditioner 11 8.9 0.000 0.000 1.687 1.691 pw_transfer 1439 11.6 0.063 0.065 1.677 1.684 rs_pw_transfer 974 11.9 0.008 0.008 1.568 1.641 density_rs2pw 119 9.7 0.003 0.004 1.551 1.615 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.578 1.606 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.586 1.594 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.533 1.534 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.520 1.521 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.488 1.505 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.363 1.372 grid_collocate_task_list 119 9.7 1.271 1.354 1.271 1.354 fft_wrap_pw1pw2_140 487 13.2 0.082 0.087 1.299 1.308 potential_pw2rs 119 12.3 0.010 0.010 1.280 1.283 ot_diis_step 108 11.5 0.012 0.013 1.266 1.266 fft3d_ps 1201 14.6 0.540 0.550 1.253 1.263 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.257 1.258 cp_fm_redistribute_end 50 14.0 0.624 1.239 0.625 1.240 cp_fm_diag_elpa_base 50 14.0 0.571 1.170 0.614 1.228 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.212 1.215 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.143 1.169 apply_single 119 13.6 0.000 0.000 1.143 1.169 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.130 1.135 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.108 1.121 wfi_extrapolate 11 7.9 0.001 0.001 1.052 1.052 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.768 1.021 make_images_data 4572 15.5 0.040 0.043 0.791 0.995 mp_alltoall_d11v 2130 13.8 0.775 0.932 0.775 0.932 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.918 acc_transpose_blocks 9144 15.5 0.038 0.039 0.896 0.903 cp_fm_cholesky_invert 11 10.9 0.843 0.846 0.843 0.846 multiply_cannon_sync_h2d 9144 15.5 0.705 0.794 0.705 0.794 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.745 0.748 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.669 0.721 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.704 0.713 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.349 0.681 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.655 0.658 mp_allgather_i34 2286 14.5 0.238 0.644 0.238 0.644 mp_alltoall_z22v 1201 16.6 0.586 0.620 0.586 0.620 qs_create_task_list 11 7.9 0.000 0.000 0.543 0.568 generate_qs_task_list 11 8.9 0.190 0.212 0.543 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.352000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.909091, yerr=5.160242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 751.448064E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.030 0.034 40.000 40.001 qs_mol_dyn_low 1 2.0 0.003 0.003 39.793 39.800 qs_forces 11 3.9 0.001 0.002 39.734 39.736 qs_energies 11 4.9 0.001 0.001 37.803 37.807 scf_env_do_scf 11 5.9 0.000 0.001 32.050 32.051 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.272 24.274 velocity_verlet 10 3.0 0.002 0.002 22.615 22.621 dbcsr_multiply_generic 2286 12.5 0.102 0.104 17.355 17.555 qs_scf_new_mos 108 7.5 0.001 0.001 15.552 15.655 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.551 15.654 ot_scf_mini 108 9.5 0.002 0.002 14.473 14.582 multiply_cannon 2286 13.5 0.307 0.316 13.568 14.445 multiply_cannon_loop 2286 14.5 0.344 0.349 12.313 13.224 ot_mini 108 10.5 0.001 0.001 8.670 8.794 multiply_cannon_multrec 9144 15.5 3.383 4.858 8.565 8.683 init_scf_loop 11 6.9 0.000 0.000 7.752 7.756 rebuild_ks_matrix 119 8.3 0.000 0.001 6.728 6.872 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.727 6.872 prepare_preconditioner 11 7.9 0.000 0.000 6.837 6.850 make_preconditioner 11 8.9 0.000 0.000 6.837 6.850 qs_ot_get_derivative 108 11.5 0.001 0.001 6.667 6.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.397 6.729 dbcsr_mm_accdrv_process 12550 15.8 4.103 5.834 5.059 6.408 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.090 6.221 cp_fm_upper_to_full 72 14.2 3.199 4.581 3.199 4.581 init_scf_run 11 5.9 0.000 0.001 3.759 3.759 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.758 3.759 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.630 3.680 calculate_rho_elec 119 8.7 0.118 0.121 3.630 3.680 sum_up_and_integrate 119 10.3 0.064 0.066 3.588 3.595 integrate_v_rspace 119 11.3 0.003 0.003 3.524 3.532 mp_waitall_1 97218 16.6 2.455 3.415 2.455 3.415 qs_ot_get_p 119 10.4 0.001 0.001 3.088 3.232 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.469 2.859 dbcsr_complete_redistribute 329 12.2 0.289 0.293 1.967 2.770 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.666 2.478 make_m2s 4572 13.5 0.038 0.038 2.270 2.441 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.136 2.424 apply_single 119 13.6 0.000 0.000 2.136 2.424 calculate_first_density_matrix 1 7.0 0.000 0.000 2.317 2.371 make_images 4572 14.5 0.351 0.384 2.149 2.319 calculate_dm_sparse 119 9.5 0.000 0.000 2.222 2.269 mp_alltoall_i22 627 13.8 1.415 2.261 1.415 2.261 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.238 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.294 2.165 grid_integrate_task_list 119 12.3 2.000 2.015 2.000 2.015 ot_diis_step 108 11.5 0.014 0.014 1.980 1.980 pw_transfer 1439 11.6 0.066 0.066 1.957 1.959 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.861 1.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.857 1.858 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.851 1.853 jit_kernel_multiply 10 15.7 0.928 1.793 0.928 1.793 mp_sum_l 7207 12.9 0.971 1.738 0.971 1.738 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.736 1.737 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.655 1.709 density_rs2pw 119 9.7 0.004 0.004 1.694 1.701 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.604 1.646 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.580 1.581 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.562 1.566 fft3d_ps 1201 14.6 0.569 0.581 1.494 1.498 grid_collocate_task_list 119 9.7 1.443 1.455 1.443 1.455 cp_fm_cholesky_invert 11 10.9 1.445 1.448 1.445 1.448 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.135 1.359 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.289 1.338 wfi_extrapolate 11 7.9 0.001 0.001 1.328 1.328 rs_pw_transfer 974 11.9 0.009 0.009 1.290 1.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.300 1.300 cp_fm_diag_elpa_base 50 14.0 1.155 1.208 1.298 1.299 make_images_data 4572 15.5 0.043 0.046 1.072 1.287 mp_alltoall_d11v 2130 13.8 1.169 1.230 1.169 1.230 potential_pw2rs 119 12.3 0.014 0.014 1.149 1.152 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.115 1.134 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.078 1.094 multiply_cannon_sync_h2d 9144 15.5 1.041 1.044 1.041 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.985 1.040 qs_create_task_list 11 7.9 0.000 0.000 0.935 0.945 generate_qs_task_list 11 8.9 0.372 0.391 0.935 0.945 acc_transpose_blocks 9144 15.5 0.038 0.038 0.914 0.918 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.849 0.862 mp_alltoall_z22v 1201 16.6 0.790 0.813 0.790 0.813 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.001000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=702.363636, yerr=15.801454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.501952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 406977. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 84.286 84.287 qs_mol_dyn_low 1 2.0 0.003 0.003 83.858 83.867 qs_forces 11 3.9 0.002 0.002 83.779 83.788 qs_energies 11 4.9 0.002 0.008 80.871 80.892 scf_env_do_scf 11 5.9 0.000 0.001 71.832 71.834 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.683 65.684 dbcsr_multiply_generic 2055 12.4 0.117 0.149 51.673 51.946 qs_scf_new_mos 99 7.5 0.000 0.001 47.757 47.883 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.756 47.882 ot_scf_mini 99 9.5 0.002 0.003 45.417 45.504 multiply_cannon 2055 13.4 0.185 0.200 42.151 43.041 velocity_verlet 10 3.0 0.001 0.002 42.480 42.481 multiply_cannon_loop 2055 14.4 1.548 1.598 41.048 41.957 ot_mini 99 10.5 0.001 0.001 27.412 27.493 qs_ot_get_derivative 99 11.5 0.001 0.001 20.581 20.691 multiply_cannon_multrec 49320 15.4 12.312 13.063 17.271 18.043 rebuild_ks_matrix 110 8.3 0.000 0.001 15.066 15.176 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 15.066 15.175 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.266 13.363 mp_waitall_1 241148 16.1 11.738 12.714 11.738 12.714 multiply_cannon_sync_h2d 49320 15.4 10.118 10.814 10.118 10.814 qs_ot_get_p 110 10.4 0.002 0.013 9.562 9.702 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.210 7.714 apply_single 110 13.6 0.000 0.001 7.210 7.714 sum_up_and_integrate 110 10.3 0.036 0.043 7.519 7.534 integrate_v_rspace 110 11.3 0.002 0.003 7.482 7.506 multiply_cannon_metrocomm3 49320 15.4 0.082 0.088 6.398 7.377 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.638 7.193 init_scf_run 11 5.9 0.000 0.001 6.927 6.927 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.927 6.927 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.830 6.892 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.763 6.881 calculate_rho_elec 110 8.6 0.020 0.024 6.763 6.880 ot_diis_step 99 11.5 0.005 0.006 6.579 6.580 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.147 6.174 init_scf_loop 11 6.9 0.001 0.004 6.118 6.121 mp_sum_l 6514 12.8 4.787 5.658 4.787 5.658 dbcsr_mm_accdrv_process 87628 16.1 2.002 2.197 4.837 5.192 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.162 5.163 rs_pw_transfer 902 11.9 0.011 0.013 4.395 4.572 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.543 4.545 cp_fm_redistribute_end 48 14.0 3.942 4.504 3.946 4.505 cp_fm_diag_elpa_base 48 14.0 0.552 4.419 0.555 4.445 make_m2s 4110 13.4 0.061 0.065 4.283 4.395 make_images 4110 14.4 0.178 0.191 4.186 4.302 wfi_extrapolate 11 7.9 0.001 0.001 4.146 4.146 density_rs2pw 110 9.6 0.004 0.004 3.906 4.078 prepare_preconditioner 11 7.9 0.000 0.000 3.856 3.872 make_preconditioner 11 8.9 0.000 0.003 3.856 3.872 calculate_dm_sparse 110 9.5 0.001 0.001 3.709 3.798 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.519 3.709 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.614 3.670 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.594 3.598 pw_transfer 1331 11.6 0.053 0.067 3.491 3.575 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.404 3.492 grid_integrate_task_list 110 12.3 3.224 3.402 3.224 3.402 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.235 3.262 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.202 3.253 potential_pw2rs 110 12.3 0.006 0.007 3.095 3.141 fft3d_ps 1111 14.6 0.751 0.846 2.946 3.028 fft_wrap_pw1pw2_140 451 13.1 0.166 0.185 2.784 2.875 calculate_first_density_matrix 1 7.0 0.000 0.003 2.665 2.676 mp_alltoall_d11v 2046 13.8 2.192 2.636 2.192 2.636 jit_kernel_multiply 13 15.9 2.554 2.567 2.554 2.567 acc_transpose_blocks 49320 15.4 0.215 0.230 2.111 2.230 grid_collocate_task_list 110 9.6 2.082 2.183 2.082 2.183 cp_fm_cholesky_invert 11 10.9 2.102 2.107 2.102 2.107 mp_waitany 14300 13.8 1.799 1.994 1.799 1.994 mp_alltoall_z22v 1111 16.6 1.832 1.990 1.832 1.990 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.944 1.958 make_images_data 4110 15.4 0.043 0.047 1.819 1.953 mp_sum_d 3879 11.9 1.507 1.934 1.507 1.934 hybrid_alltoall_any 4261 16.3 0.083 0.480 1.575 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.287000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.863276 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 586.080256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2220252. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.201 70.157 70.159 qs_mol_dyn_low 1 2.0 0.003 0.003 69.556 69.566 qs_forces 11 3.9 0.002 0.002 69.489 69.491 qs_energies 11 4.9 0.001 0.001 66.175 66.180 scf_env_do_scf 11 5.9 0.000 0.001 57.170 57.174 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.013 49.013 dbcsr_multiply_generic 2055 12.4 0.115 0.120 38.007 38.231 velocity_verlet 10 3.0 0.002 0.002 36.724 36.725 qs_scf_new_mos 99 7.5 0.001 0.001 33.162 33.298 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.161 33.298 multiply_cannon 2055 13.4 0.224 0.246 31.131 32.177 ot_scf_mini 99 9.5 0.003 0.003 31.482 31.607 multiply_cannon_loop 2055 14.4 0.926 0.943 29.733 30.552 ot_mini 99 10.5 0.001 0.001 18.544 18.680 multiply_cannon_multrec 24660 15.4 7.618 8.896 13.881 15.140 rebuild_ks_matrix 110 8.3 0.000 0.001 13.638 13.794 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.638 13.794 qs_ot_get_derivative 99 11.5 0.001 0.001 12.694 12.818 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.030 12.170 mp_waitall_1 186928 16.3 8.003 10.759 8.003 10.759 init_scf_loop 11 6.9 0.000 0.000 8.117 8.118 multiply_cannon_sync_h2d 24660 15.4 6.982 8.043 6.982 8.043 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.188 7.622 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.546 7.293 apply_single 110 13.6 0.000 0.001 6.545 7.293 init_scf_run 11 5.9 0.000 0.001 6.592 6.593 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.592 6.593 sum_up_and_integrate 110 10.3 0.052 0.059 6.559 6.570 integrate_v_rspace 110 11.3 0.002 0.002 6.507 6.521 qs_ot_get_p 110 10.4 0.001 0.001 6.282 6.472 dbcsr_mm_accdrv_process 52282 16.1 4.732 5.530 6.104 6.419 prepare_preconditioner 11 7.9 0.000 0.000 6.042 6.062 make_preconditioner 11 8.9 0.000 0.000 6.042 6.061 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.823 5.834 calculate_rho_elec 110 8.6 0.039 0.047 5.823 5.834 ot_diis_step 99 11.5 0.010 0.010 5.795 5.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.613 5.769 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.735 5.462 make_m2s 4110 13.4 0.057 0.060 4.318 4.807 make_images 4110 14.4 0.400 0.444 4.209 4.693 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.225 4.246 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.775 3.776 wfi_extrapolate 11 7.9 0.001 0.001 3.604 3.604 pw_transfer 1331 11.6 0.065 0.069 3.325 3.463 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.321 3.394 density_rs2pw 110 9.6 0.004 0.004 3.143 3.361 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.219 3.356 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.344 3.346 grid_integrate_task_list 110 12.3 3.147 3.300 3.147 3.300 rs_pw_transfer 902 11.9 0.012 0.013 3.065 3.244 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.151 3.154 cp_fm_redistribute_end 48 14.0 2.342 3.118 2.344 3.120 cp_fm_diag_elpa_base 48 14.0 0.739 2.998 0.771 3.085 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.970 3.033 calculate_dm_sparse 110 9.5 0.001 0.001 2.987 3.017 calculate_first_density_matrix 1 7.0 0.000 0.001 2.897 2.900 cp_fm_cholesky_invert 11 10.9 2.877 2.885 2.877 2.885 make_images_data 4110 15.4 0.048 0.052 2.349 2.845 hybrid_alltoall_any 4261 16.3 0.103 0.447 2.075 2.756 fft_wrap_pw1pw2_140 451 13.1 0.199 0.216 2.609 2.751 fft3d_ps 1111 14.6 1.074 1.269 2.577 2.697 mp_sum_l 6514 12.8 1.937 2.592 1.937 2.592 potential_pw2rs 110 12.3 0.008 0.009 2.455 2.474 grid_collocate_task_list 110 9.6 2.119 2.256 2.119 2.256 mp_alltoall_d11v 2046 13.8 1.787 2.020 1.787 2.020 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.997 2.016 jit_kernel_multiply 10 16.3 1.018 1.930 1.018 1.930 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.888 1.891 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.774 1.788 mp_allgather_i34 2055 14.4 0.724 1.716 0.724 1.716 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.794 1.703 mp_irecv_dv 57340 16.2 0.669 1.583 0.669 1.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.556 acc_transpose_blocks 24660 15.4 0.110 0.112 1.514 1.530 dbcsr_complete_redistribute 325 12.2 0.234 0.287 1.255 1.527 cp_fm_cholesky_decompose 22 10.9 1.496 1.503 1.496 1.503 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.159000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.636364, yerr=6.944883 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 659.628032E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.042 60.856 60.857 qs_mol_dyn_low 1 2.0 0.003 0.003 60.513 60.524 qs_forces 11 3.9 0.002 0.002 60.448 60.449 qs_energies 11 4.9 0.003 0.010 57.256 57.261 scf_env_do_scf 11 5.9 0.001 0.002 48.874 48.874 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.106 40.107 velocity_verlet 10 3.0 0.001 0.002 32.911 32.913 dbcsr_multiply_generic 2055 12.4 0.109 0.113 28.780 29.078 qs_scf_new_mos 99 7.5 0.001 0.001 25.416 25.531 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.415 25.530 ot_scf_mini 99 9.5 0.003 0.003 24.178 24.307 multiply_cannon 2055 13.4 0.212 0.221 22.189 23.466 multiply_cannon_loop 2055 14.4 0.615 0.634 20.951 22.060 ot_mini 99 10.5 0.001 0.001 13.901 14.036 rebuild_ks_matrix 110 8.3 0.000 0.000 12.319 12.470 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.319 12.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.853 10.985 mp_waitall_1 146766 16.3 7.594 10.699 7.594 10.699 multiply_cannon_multrec 16440 15.4 3.787 4.709 9.518 10.307 qs_ot_get_derivative 99 11.5 0.001 0.001 9.339 9.472 init_scf_loop 11 6.9 0.001 0.005 8.730 8.731 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.575 7.490 prepare_preconditioner 11 7.9 0.000 0.000 6.941 6.957 make_preconditioner 11 8.9 0.000 0.002 6.940 6.957 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.259 6.616 sum_up_and_integrate 110 10.3 0.060 0.061 6.533 6.549 integrate_v_rspace 110 11.3 0.002 0.003 6.473 6.488 init_scf_run 11 5.9 0.000 0.001 5.979 5.979 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.979 5.979 dbcsr_mm_accdrv_process 34862 16.1 4.661 5.188 5.589 5.739 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.641 5.654 calculate_rho_elec 110 8.6 0.058 0.058 5.641 5.654 qs_ot_get_p 110 10.4 0.001 0.002 5.425 5.591 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.064 5.524 apply_single 110 13.6 0.000 0.000 5.064 5.524 make_m2s 4110 13.4 0.050 0.051 4.347 4.688 make_images 4110 14.4 0.391 0.508 4.231 4.572 ot_diis_step 99 11.5 0.010 0.011 4.519 4.519 multiply_cannon_sync_h2d 16440 15.4 3.650 4.208 3.650 4.208 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.182 3.827 qs_ot_p2m_diag 48 11.0 0.042 0.045 3.697 3.702 grid_integrate_task_list 110 12.3 3.180 3.389 3.180 3.389 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.339 3.340 pw_transfer 1331 11.6 0.064 0.070 3.296 3.304 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.190 3.202 density_rs2pw 110 9.6 0.004 0.004 2.910 3.090 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 make_images_data 4110 15.4 0.044 0.048 2.523 2.978 rs_pw_transfer 902 11.9 0.010 0.011 2.759 2.936 calculate_first_density_matrix 1 7.0 0.001 0.004 2.890 2.891 cp_fm_cholesky_invert 11 10.9 2.866 2.872 2.866 2.872 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.797 2.799 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.770 2.772 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.171 2.766 cp_fm_redistribute_end 48 14.0 1.720 2.742 1.722 2.743 cp_fm_diag_elpa_base 48 14.0 0.959 2.605 1.015 2.711 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.685 2.696 calculate_dm_sparse 110 9.5 0.001 0.001 2.511 2.544 fft3d_ps 1111 14.6 1.065 1.074 2.505 2.517 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.433 2.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.492 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.904 2.485 mp_irecv_dv 48980 15.7 0.834 2.357 0.834 2.357 potential_pw2rs 110 12.3 0.010 0.010 2.329 2.340 grid_collocate_task_list 110 9.6 2.175 2.323 2.175 2.323 mp_sum_l 6514 12.8 1.631 2.304 1.631 2.304 mp_alltoall_d11v 2046 13.8 1.833 2.211 1.833 2.211 dbcsr_complete_redistribute 325 12.2 0.326 0.357 1.481 1.955 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.945 1.946 cp_fm_upper_to_full 70 13.6 1.381 1.834 1.381 1.834 mp_allgather_i34 2055 14.4 0.561 1.713 0.561 1.713 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.667 1.686 cp_fm_cholesky_decompose 22 10.9 1.642 1.660 1.642 1.660 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.460 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.012 1.464 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.364 1.447 mp_waitany 17072 13.8 1.145 1.346 1.145 1.346 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.296 1.303 make_basis_sm 11 9.8 0.001 0.001 1.268 1.270 rs_gather_matrices 110 12.3 0.138 0.153 0.910 1.270 parallel_gemm_fm 81 9.0 0.000 0.000 1.263 1.266 parallel_gemm_fm_cosma 81 10.0 1.263 1.266 1.263 1.266 mp_alltoall_z22v 1111 16.6 1.226 1.249 1.226 1.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.857000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.727273, yerr=9.342872 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 728.322048E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.041 65.073 65.074 qs_mol_dyn_low 1 2.0 0.003 0.003 64.664 64.673 qs_forces 11 3.9 0.002 0.005 64.599 64.599 qs_energies 11 4.9 0.002 0.009 61.224 61.231 scf_env_do_scf 11 5.9 0.001 0.001 52.837 52.840 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.935 40.937 velocity_verlet 10 3.0 0.002 0.003 36.923 36.925 dbcsr_multiply_generic 2055 12.4 0.116 0.120 29.549 29.776 qs_scf_new_mos 99 7.5 0.001 0.001 26.480 26.584 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.480 26.584 ot_scf_mini 99 9.5 0.003 0.003 24.855 24.966 multiply_cannon 2055 13.4 0.246 0.267 22.438 23.488 multiply_cannon_loop 2055 14.4 0.890 0.907 20.975 21.587 ot_mini 99 10.5 0.001 0.001 14.126 14.240 multiply_cannon_multrec 24660 15.4 4.183 6.729 12.603 13.746 rebuild_ks_matrix 110 8.3 0.000 0.000 11.989 12.104 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.988 12.104 init_scf_loop 11 6.9 0.002 0.018 11.860 11.863 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.609 10.709 prepare_preconditioner 11 7.9 0.000 0.000 10.086 10.105 make_preconditioner 11 8.9 0.000 0.002 10.086 10.105 qs_ot_get_derivative 99 11.5 0.001 0.001 9.938 10.048 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.287 9.764 dbcsr_mm_accdrv_process 52304 16.0 6.860 8.135 8.273 9.210 mp_waitall_1 126806 16.4 4.584 6.479 4.584 6.479 sum_up_and_integrate 110 10.3 0.068 0.071 6.378 6.393 integrate_v_rspace 110 11.3 0.002 0.003 6.310 6.324 qs_ot_get_p 110 10.4 0.001 0.001 5.719 5.863 make_m2s 4110 13.4 0.059 0.061 5.489 5.826 make_images 4110 14.4 0.577 0.699 5.347 5.681 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.635 5.647 calculate_rho_elec 110 8.6 0.077 0.081 5.634 5.646 init_scf_run 11 5.9 0.000 0.001 5.535 5.535 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.534 5.535 cp_fm_upper_to_full 70 13.8 3.350 4.753 3.350 4.753 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.046 4.188 apply_single 110 13.6 0.000 0.000 4.046 4.188 ot_diis_step 99 11.5 0.011 0.011 4.147 4.147 dbcsr_complete_redistribute 325 12.2 0.420 0.472 2.760 3.916 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.874 3.888 grid_integrate_task_list 110 12.3 3.288 3.413 3.288 3.413 multiply_cannon_sync_h2d 24660 15.4 3.174 3.387 3.174 3.387 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.382 3.382 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.317 3.366 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.433 3.352 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.182 3.316 pw_transfer 1331 11.6 0.064 0.071 3.274 3.299 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.169 3.197 make_images_data 4110 15.4 0.047 0.051 2.794 3.109 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.439 3.108 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.031 3.092 wfi_extrapolate 11 7.9 0.001 0.001 2.975 2.975 calculate_dm_sparse 110 9.5 0.001 0.001 2.912 2.945 density_rs2pw 110 9.6 0.004 0.004 2.777 2.912 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.789 2.907 mp_alltoall_i22 605 13.7 1.679 2.876 1.679 2.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.840 2.841 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.810 2.812 cp_fm_redistribute_end 48 14.0 1.398 2.781 1.400 2.782 cp_fm_cholesky_invert 11 10.9 2.773 2.781 2.773 2.781 cp_fm_diag_elpa_base 48 14.0 1.298 2.647 1.379 2.760 rs_pw_transfer 902 11.9 0.010 0.011 2.519 2.711 fft_wrap_pw1pw2_140 451 13.1 0.202 0.215 2.679 2.707 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.470 2.510 fft3d_ps 1111 14.6 1.065 1.101 2.471 2.488 calculate_first_density_matrix 1 7.0 0.000 0.003 2.469 2.474 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.348 2.349 grid_collocate_task_list 110 9.6 2.223 2.337 2.223 2.337 potential_pw2rs 110 12.3 0.012 0.013 2.142 2.155 mp_alltoall_d11v 2046 13.8 1.858 1.996 1.858 1.996 jit_kernel_multiply 10 15.7 1.080 1.818 1.080 1.818 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.711 1.746 cp_fm_cholesky_decompose 22 10.9 1.663 1.716 1.663 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.601 1.700 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.627 1.639 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.539 acc_transpose_blocks 24660 15.4 0.103 0.105 1.506 1.534 mp_sum_l 6514 12.8 0.939 1.493 0.939 1.493 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.846 1.474 mp_waitany 13376 13.8 1.131 1.400 1.131 1.400 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.287 1.395 mp_irecv_dv 62702 16.1 0.746 1.393 0.746 1.393 mp_allgather_i34 2055 14.4 0.558 1.369 0.558 1.369 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.074000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.000000, yerr=9.351714 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 844.410880E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 55.052 55.053 qs_mol_dyn_low 1 2.0 0.003 0.003 54.778 54.787 qs_forces 11 3.9 0.002 0.002 54.715 54.715 qs_energies 11 4.9 0.001 0.001 51.101 51.106 scf_env_do_scf 11 5.9 0.000 0.001 42.587 42.587 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.629 34.630 velocity_verlet 10 3.0 0.002 0.002 30.764 30.766 dbcsr_multiply_generic 2055 12.4 0.106 0.108 22.943 23.057 qs_scf_new_mos 99 7.5 0.001 0.001 20.556 20.626 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.555 20.626 ot_scf_mini 99 9.5 0.002 0.002 19.324 19.386 multiply_cannon 2055 13.4 0.251 0.268 17.428 18.728 multiply_cannon_loop 2055 14.4 0.322 0.335 16.087 16.297 rebuild_ks_matrix 110 8.3 0.000 0.000 11.518 11.556 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.517 11.556 ot_mini 99 10.5 0.001 0.001 10.433 10.489 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.261 10.298 multiply_cannon_multrec 8220 15.4 3.204 4.445 7.491 8.555 mp_waitall_1 106626 16.5 6.282 7.942 6.282 7.942 init_scf_loop 11 6.9 0.000 0.000 7.908 7.908 qs_ot_get_derivative 99 11.5 0.001 0.001 6.598 6.659 prepare_preconditioner 11 7.9 0.000 0.000 6.213 6.223 make_preconditioner 11 8.9 0.000 0.000 6.213 6.223 sum_up_and_integrate 110 10.3 0.079 0.081 6.157 6.169 integrate_v_rspace 110 11.3 0.003 0.003 6.078 6.090 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.789 5.862 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.588 5.600 calculate_rho_elec 110 8.6 0.115 0.115 5.588 5.599 init_scf_run 11 5.9 0.000 0.001 5.412 5.412 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.411 5.412 dbcsr_mm_accdrv_process 17442 15.9 2.996 3.994 4.156 5.053 qs_ot_get_p 110 10.4 0.001 0.001 4.926 4.996 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.120 4.511 make_m2s 4110 13.4 0.039 0.040 4.223 4.477 make_images 4110 14.4 0.637 0.695 4.093 4.345 ot_diis_step 99 11.5 0.012 0.012 3.813 3.814 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.737 3.796 apply_single 110 13.6 0.000 0.000 3.737 3.796 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.569 3.573 grid_integrate_task_list 110 12.3 3.352 3.528 3.352 3.528 pw_transfer 1331 11.6 0.064 0.069 3.255 3.275 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.263 3.264 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.148 3.172 cp_fm_cholesky_invert 11 10.9 2.999 3.003 2.999 3.003 multiply_cannon_sync_h2d 8220 15.4 2.904 3.000 2.904 3.000 make_images_data 4110 15.4 0.039 0.044 2.411 2.827 density_rs2pw 110 9.6 0.004 0.004 2.639 2.769 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.734 2.735 fft_wrap_pw1pw2_140 451 13.1 0.213 0.214 2.696 2.732 hybrid_alltoall_any 4261 16.3 0.201 0.865 2.360 2.732 wfi_extrapolate 11 7.9 0.001 0.001 2.706 2.706 cp_fm_redistribute_end 48 14.0 0.690 2.703 0.694 2.703 cp_fm_diag_elpa_base 48 14.0 1.826 2.511 2.001 2.672 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.646 2.647 calculate_first_density_matrix 1 7.0 0.000 0.000 2.619 2.620 grid_collocate_task_list 110 9.6 2.330 2.549 2.330 2.549 calculate_dm_sparse 110 9.5 0.001 0.001 2.502 2.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.517 2.518 fft3d_ps 1111 14.6 1.120 1.173 2.406 2.422 rs_pw_transfer 902 11.9 0.010 0.010 2.149 2.305 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.152 2.166 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.988 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.936 1.969 potential_pw2rs 110 12.3 0.015 0.015 1.914 1.918 mp_alltoall_d11v 2046 13.8 1.686 1.884 1.686 1.884 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.812 1.847 cp_fm_cholesky_decompose 22 10.9 1.720 1.741 1.720 1.741 mp_allgather_i34 2055 14.4 0.521 1.618 0.521 1.618 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.493 1.613 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.589 1.594 dbcsr_complete_redistribute 325 12.2 0.569 0.604 1.491 1.580 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.474 1.486 qs_create_task_list 11 7.9 0.000 0.000 1.222 1.321 generate_qs_task_list 11 8.9 0.378 0.446 1.222 1.321 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.477 1.285 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.805 1.252 mp_irecv_dv 24056 15.7 0.453 1.245 0.453 1.245 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.188 1.212 mp_waitany 9240 13.8 1.044 1.197 1.044 1.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.053000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=788.000000, yerr=17.252141 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 1.333334E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.038 86.105 86.106 qs_mol_dyn_low 1 2.0 0.003 0.003 85.403 85.411 qs_forces 11 3.9 0.022 0.028 85.332 85.338 qs_energies 11 4.9 0.001 0.001 81.290 81.293 scf_env_do_scf 11 5.9 0.000 0.001 71.252 71.253 velocity_verlet 10 3.0 0.002 0.002 54.629 54.635 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.998 43.000 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.941 29.064 init_scf_loop 11 6.9 0.000 0.000 28.179 28.183 qs_scf_new_mos 99 7.5 0.001 0.001 26.645 26.693 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.645 26.692 prepare_preconditioner 11 7.9 0.000 0.000 26.248 26.257 make_preconditioner 11 8.9 0.000 0.000 26.248 26.257 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.553 25.714 ot_scf_mini 99 9.5 0.002 0.002 24.870 24.902 multiply_cannon 2055 13.4 0.352 0.379 21.870 22.566 multiply_cannon_loop 2055 14.4 0.341 0.344 20.060 20.349 cp_fm_upper_to_full 70 14.2 12.662 18.066 12.662 18.066 ot_mini 99 10.5 0.001 0.001 13.893 13.934 rebuild_ks_matrix 110 8.3 0.001 0.001 13.086 13.152 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.085 13.151 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.902 11.961 dbcsr_complete_redistribute 325 12.2 1.029 1.052 7.292 10.431 multiply_cannon_multrec 8220 15.4 4.640 4.793 9.941 10.072 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.262 9.398 qs_ot_get_derivative 99 11.5 0.001 0.001 9.324 9.357 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.680 8.783 mp_waitall_1 87304 16.6 7.736 8.664 7.736 8.664 mp_alltoall_i22 605 13.7 5.322 8.483 5.322 8.483 sum_up_and_integrate 110 10.3 0.150 0.151 6.617 6.631 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.475 6.507 calculate_rho_elec 110 8.6 0.227 0.227 6.475 6.507 integrate_v_rspace 110 11.3 0.004 0.004 6.467 6.481 qs_ot_get_p 110 10.4 0.001 0.001 5.786 5.840 init_scf_run 11 5.9 0.000 0.001 5.800 5.800 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.800 5.800 make_m2s 4110 13.4 0.043 0.043 5.236 5.761 make_images 4110 14.4 0.882 0.934 5.048 5.573 dbcsr_mm_accdrv_process 11614 15.7 3.221 3.655 5.160 5.424 cp_fm_cholesky_invert 11 10.9 5.335 5.339 5.335 5.339 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.643 5.155 apply_single 110 13.6 0.000 0.000 4.643 5.155 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.749 5.003 ot_diis_step 99 11.5 0.015 0.016 4.540 4.540 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.218 4.226 multiply_cannon_sync_h2d 8220 15.4 3.951 3.954 3.951 3.954 pw_transfer 1331 11.6 0.073 0.074 3.775 3.780 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.773 3.774 grid_integrate_task_list 110 12.3 3.656 3.705 3.656 3.705 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.659 3.665 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.642 3.643 hybrid_alltoall_any 4261 16.3 0.257 0.556 2.892 3.583 make_images_data 4110 15.4 0.042 0.046 2.831 3.509 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.910 3.363 calculate_dm_sparse 110 9.5 0.001 0.001 3.283 3.305 wfi_extrapolate 11 7.9 0.001 0.001 3.214 3.215 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.168 3.179 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.173 3.174 cp_fm_diag_elpa_base 48 14.0 2.638 2.838 3.171 3.171 density_rs2pw 110 9.6 0.004 0.004 2.972 2.991 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.897 2.901 fft3d_ps 1111 14.6 1.272 1.281 2.867 2.876 grid_collocate_task_list 110 9.6 2.625 2.646 2.625 2.646 calculate_first_density_matrix 1 7.0 0.000 0.000 2.478 2.482 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.401 2.423 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.377 2.402 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.174 2.229 rs_pw_transfer 902 11.9 0.010 0.011 2.123 2.170 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.079 2.168 potential_pw2rs 110 12.3 0.021 0.022 2.069 2.076 mp_alltoall_d11v 2046 13.8 2.029 2.071 2.029 2.071 cp_fm_cholesky_decompose 22 10.9 2.038 2.057 2.038 2.057 qs_create_task_list 11 7.9 0.000 0.000 1.901 1.947 generate_qs_task_list 11 8.9 0.739 0.796 1.901 1.947 jit_kernel_multiply 10 15.4 1.735 1.910 1.735 1.910 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.805 1.819 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.767 1.815 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.789 1.796 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.106000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1197.636364, yerr=56.383867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 630.972416E+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 4103 56513. MP_Allreduce 11298 784. MP_Sync 170 MP_Alltoall 2226 2794565. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 202.384 202.385 qs_mol_dyn_low 1 2.0 0.003 0.003 202.033 202.046 qs_forces 11 3.9 0.003 0.004 201.937 201.948 qs_energies 11 4.9 0.001 0.002 196.447 196.466 scf_env_do_scf 11 5.9 0.001 0.001 180.194 180.198 scf_env_do_scf_inner_loop 117 6.6 0.002 0.008 159.585 159.588 dbcsr_multiply_generic 2507 12.6 0.178 0.183 124.071 125.033 velocity_verlet 10 3.0 0.002 0.002 122.421 122.423 qs_scf_new_mos 117 7.6 0.001 0.001 120.999 121.245 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.999 121.245 ot_scf_mini 117 9.6 0.003 0.003 114.404 114.698 multiply_cannon 2507 13.6 0.236 0.246 100.829 103.739 multiply_cannon_loop 2507 14.6 2.106 2.208 98.589 101.215 ot_mini 117 10.6 0.001 0.001 65.655 65.918 multiply_cannon_multrec 60168 15.6 33.339 35.097 41.609 43.667 qs_ot_get_derivative 117 11.6 0.001 0.001 40.807 41.067 rebuild_ks_matrix 128 8.3 0.001 0.001 33.165 33.732 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.164 33.731 mp_waitall_1 291448 16.2 28.530 31.241 28.530 31.241 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.846 30.359 multiply_cannon_sync_h2d 60168 15.6 27.401 29.387 27.401 29.387 qs_ot_get_p 128 10.4 0.001 0.001 26.974 27.245 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.374 25.486 apply_single 128 13.6 0.001 0.001 24.374 25.486 ot_diis_step 117 11.6 0.007 0.008 24.545 24.546 init_scf_loop 11 6.9 0.000 0.001 20.531 20.532 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.137 20.219 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.920 19.129 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.575 17.577 multiply_cannon_metrocomm3 60168 15.6 0.119 0.124 15.621 17.435 prepare_preconditioner 11 7.9 0.000 0.000 15.926 15.971 make_preconditioner 11 8.9 0.000 0.000 15.926 15.971 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.148 15.307 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.756 14.761 cp_fm_redistribute_end 83 14.4 11.629 14.684 11.641 14.688 cp_fm_diag_elpa_base 83 14.4 3.004 14.459 3.035 14.578 make_m2s 5014 13.6 0.103 0.110 13.829 14.155 make_images 5014 14.6 0.401 0.419 13.649 13.987 sum_up_and_integrate 128 10.3 0.089 0.109 13.880 13.893 integrate_v_rspace 128 11.3 0.003 0.004 13.790 13.809 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.574 12.641 calculate_rho_elec 128 8.7 0.045 0.063 12.574 12.641 init_scf_run 11 5.9 0.000 0.001 12.132 12.133 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.132 12.133 mp_sum_l 7870 13.0 8.137 9.198 8.137 9.198 cp_fm_cholesky_invert 11 10.9 9.059 9.067 9.059 9.067 wfi_extrapolate 11 7.9 0.001 0.001 8.951 8.951 multiply_cannon_metrocomm1 60168 15.6 0.086 0.090 6.007 8.722 calculate_dm_sparse 128 9.5 0.001 0.001 8.403 8.513 dbcsr_mm_accdrv_process 124484 16.2 3.202 3.430 7.825 8.373 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.047 8.187 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.750 7.816 make_images_data 5014 15.6 0.068 0.073 6.833 7.812 grid_integrate_task_list 128 12.3 7.087 7.494 7.087 7.494 hybrid_alltoall_any 5200 16.5 0.292 2.252 5.991 7.398 pw_transfer 1547 11.6 0.075 0.103 6.811 7.076 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.608 6.846 density_rs2pw 128 9.7 0.006 0.007 6.385 6.783 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.590 6.597 rs_pw_transfer 1046 11.9 0.016 0.019 5.510 5.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.751 5.902 fft_wrap_pw1pw2_140 523 13.2 0.445 0.510 5.677 5.854 fft3d_ps 1291 14.7 2.094 2.564 5.400 5.589 mp_alltoall_d11v 2415 14.1 4.218 5.487 4.218 5.487 grid_collocate_task_list 128 9.7 4.724 5.156 4.724 5.156 cp_fm_cholesky_decompose 22 10.9 4.676 4.692 4.676 4.692 potential_pw2rs 128 12.3 0.009 0.010 4.369 4.398 mp_sum_d 4469 12.2 3.533 4.375 3.533 4.375 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.385000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.545455, yerr=7.253383 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 826.777600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5618882. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 189.339 189.340 qs_mol_dyn_low 1 2.0 0.003 0.003 189.017 189.031 qs_forces 11 3.9 0.003 0.003 188.923 188.925 qs_energies 11 4.9 0.002 0.002 182.219 182.234 scf_env_do_scf 11 5.9 0.001 0.001 166.115 166.124 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.296 133.299 velocity_verlet 10 3.0 0.003 0.004 119.165 119.167 dbcsr_multiply_generic 2507 12.6 0.186 0.194 97.427 98.705 qs_scf_new_mos 117 7.6 0.001 0.001 94.591 95.033 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.590 95.032 ot_scf_mini 117 9.6 0.004 0.004 89.783 90.280 multiply_cannon 2507 13.6 0.478 0.528 77.444 81.443 multiply_cannon_loop 2507 14.6 1.251 1.297 74.147 76.511 ot_mini 117 10.6 0.001 0.001 49.620 50.114 mp_waitall_1 226760 16.4 24.711 38.286 24.711 38.286 multiply_cannon_multrec 30084 15.6 22.099 26.182 31.742 36.004 rebuild_ks_matrix 128 8.3 0.001 0.001 32.835 33.372 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.834 33.371 init_scf_loop 11 6.9 0.000 0.000 32.727 32.728 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.590 30.069 multiply_cannon_metrocomm3 30084 15.6 0.094 0.099 15.486 28.954 prepare_preconditioner 11 7.9 0.000 0.000 28.361 28.415 make_preconditioner 11 8.9 0.000 0.000 28.361 28.415 qs_ot_get_derivative 117 11.6 0.001 0.002 27.708 28.213 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.048 27.590 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.999 23.038 apply_single 128 13.6 0.001 0.001 21.999 23.037 qs_ot_get_p 128 10.4 0.001 0.001 21.842 22.432 multiply_cannon_sync_h2d 30084 15.6 19.520 22.105 19.520 22.105 ot_diis_step 117 11.6 0.014 0.015 21.738 21.740 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.039 17.072 cp_fm_cholesky_invert 11 10.9 16.590 16.602 16.590 16.602 make_m2s 5014 13.6 0.090 0.097 14.300 15.928 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.896 15.897 make_images 5014 14.6 1.157 1.367 14.090 15.722 sum_up_and_integrate 128 10.3 0.116 0.132 14.341 14.377 integrate_v_rspace 128 11.3 0.003 0.003 14.224 14.265 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.960 12.995 calculate_rho_elec 128 8.7 0.087 0.105 12.959 12.995 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.803 12.810 cp_fm_redistribute_end 83 14.4 7.444 12.741 7.456 12.742 cp_fm_diag_elpa_base 83 14.4 5.051 12.289 5.270 12.658 init_scf_run 11 5.9 0.000 0.001 11.371 11.373 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.371 11.373 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.904 11.242 make_images_data 5014 15.6 0.066 0.073 8.721 10.615 multiply_cannon_metrocomm4 27577 15.6 0.098 0.113 3.708 10.602 mp_irecv_dv 69486 16.3 3.513 10.212 3.513 10.212 hybrid_alltoall_any 5200 16.5 0.344 1.510 7.399 9.942 dbcsr_mm_accdrv_process 62242 16.2 4.397 5.087 9.102 9.685 wfi_extrapolate 11 7.9 0.001 0.001 8.314 8.314 pw_transfer 1547 11.6 0.084 0.096 7.777 7.843 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.553 7.609 grid_integrate_task_list 128 12.3 7.168 7.586 7.168 7.586 density_rs2pw 128 9.7 0.006 0.006 6.730 7.210 cp_fm_cholesky_decompose 22 10.9 6.899 6.985 6.899 6.985 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.157 6.916 fft_wrap_pw1pw2_140 523 13.2 0.469 0.518 6.600 6.647 calculate_dm_sparse 128 9.5 0.001 0.001 6.469 6.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.116 6.126 rs_pw_transfer 1046 11.9 0.014 0.016 5.563 6.081 fft3d_ps 1291 14.7 2.777 2.923 5.923 5.969 mp_sum_l 7870 13.0 3.894 5.879 3.894 5.879 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.267 5.392 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.309 5.369 grid_collocate_task_list 128 9.7 4.912 5.308 4.912 5.308 mp_allgather_i34 2507 14.6 1.867 5.136 1.867 5.136 potential_pw2rs 128 12.3 0.014 0.016 4.804 4.826 mp_alltoall_d11v 2415 14.1 4.126 4.536 4.126 4.536 mp_sum_d 4455 12.2 2.603 3.989 2.603 3.989 dbcsr_complete_redistribute 395 12.7 0.770 0.847 3.119 3.978 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.340000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.363636, yerr=3.796475 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 972.034048E+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 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 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 0.026 0.047 175.746 175.747 qs_mol_dyn_low 1 2.0 0.003 0.003 175.325 175.338 qs_forces 11 3.9 0.003 0.003 175.199 175.202 qs_energies 11 4.9 0.004 0.024 168.683 168.692 scf_env_do_scf 11 5.9 0.001 0.001 153.303 153.307 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.689 117.691 velocity_verlet 10 3.0 0.002 0.002 112.373 112.376 dbcsr_multiply_generic 2507 12.6 0.182 0.185 81.452 82.636 qs_scf_new_mos 117 7.6 0.001 0.001 81.260 81.539 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.259 81.538 ot_scf_mini 117 9.6 0.004 0.004 77.124 77.474 multiply_cannon 2507 13.6 0.499 0.521 61.545 66.035 multiply_cannon_loop 2507 14.6 0.859 0.889 58.406 61.142 ot_mini 117 10.6 0.001 0.001 42.414 42.770 init_scf_loop 11 6.9 0.001 0.004 35.516 35.518 mp_waitall_1 178456 16.5 25.171 34.746 25.171 34.746 rebuild_ks_matrix 128 8.3 0.001 0.001 30.585 31.067 qs_ks_build_kohn_sham_matrix 128 9.3 0.069 0.442 30.585 31.066 prepare_preconditioner 11 7.9 0.000 0.000 30.951 31.001 make_preconditioner 11 8.9 0.006 0.047 30.951 31.000 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.556 29.936 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.644 28.085 multiply_cannon_multrec 20056 15.6 13.575 16.397 22.256 25.107 multiply_cannon_metrocomm3 20056 15.6 0.061 0.065 14.996 24.507 qs_ot_get_derivative 117 11.6 0.001 0.002 22.706 23.071 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.813 20.795 apply_single 128 13.6 0.001 0.001 19.813 20.795 qs_ot_get_p 128 10.4 0.001 0.002 20.113 20.522 ot_diis_step 117 11.6 0.017 0.018 19.603 19.603 make_m2s 5014 13.6 0.081 0.085 14.888 15.899 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.804 15.810 multiply_cannon_sync_h2d 20056 15.6 14.295 15.810 14.295 15.810 make_images 5014 14.6 1.189 1.292 14.654 15.665 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.802 14.803 cp_fm_cholesky_invert 11 10.9 14.386 14.395 14.386 14.395 sum_up_and_integrate 128 10.3 0.135 0.146 14.114 14.142 integrate_v_rspace 128 11.3 0.003 0.004 13.980 14.010 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.990 13.031 calculate_rho_elec 128 8.7 0.132 0.146 12.989 13.031 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.768 11.770 cp_fm_redistribute_end 83 14.4 4.418 11.706 4.431 11.708 cp_fm_diag_elpa_base 83 14.4 6.847 11.141 7.255 11.623 make_images_data 5014 15.6 0.061 0.068 9.229 10.746 init_scf_run 11 5.9 0.000 0.001 10.355 10.355 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.354 10.355 hybrid_alltoall_any 5200 16.5 0.429 1.972 7.983 9.819 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.897 9.181 multiply_cannon_metrocomm4 17549 15.6 0.060 0.070 3.384 9.062 mp_irecv_dv 50230 16.2 3.262 8.823 3.262 8.823 dbcsr_mm_accdrv_process 41502 16.2 4.352 5.155 8.129 8.240 pw_transfer 1547 11.6 0.084 0.104 7.772 7.873 grid_integrate_task_list 128 12.3 7.308 7.847 7.308 7.847 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.549 7.655 cp_fm_cholesky_decompose 22 10.9 7.357 7.415 7.357 7.415 cp_fm_upper_to_full 105 14.5 5.719 7.394 5.719 7.394 wfi_extrapolate 11 7.9 0.001 0.001 7.369 7.370 fft_wrap_pw1pw2_140 523 13.2 0.477 0.532 6.643 6.758 density_rs2pw 128 9.7 0.006 0.006 6.445 6.752 dbcsr_complete_redistribute 395 12.7 1.173 1.203 4.518 6.299 fft3d_ps 1291 14.7 2.690 2.893 5.820 5.882 calculate_dm_sparse 128 9.5 0.001 0.001 5.728 5.862 grid_collocate_task_list 128 9.7 5.074 5.551 5.074 5.551 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.440 5.449 rs_pw_transfer 1046 11.9 0.014 0.014 5.007 5.350 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.562 5.264 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.336 5.110 mp_allgather_i34 2507 14.6 1.658 4.952 1.658 4.952 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.541 4.696 mp_alltoall_d11v 2415 14.1 4.204 4.632 4.204 4.632 potential_pw2rs 128 12.3 0.020 0.023 4.493 4.512 mp_sum_l 7870 13.0 3.202 4.490 3.202 4.490 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.370 4.081 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.997 4.035 mp_alltoall_i22 716 14.1 1.957 3.849 1.957 3.849 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.738 3.739 mp_sum_d 4467 12.2 2.465 3.577 2.465 3.577 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.747000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=913.272727, yerr=19.074452 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.143067E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837542000 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4063 57495. MP_Allreduce 11196 1069. MP_Sync 168 MP_Alltoall 1700 12496361. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.103 0.179 186.440 186.442 qs_mol_dyn_low 1 2.0 0.003 0.003 185.814 185.827 qs_forces 11 3.9 0.005 0.007 185.694 185.700 qs_energies 11 4.9 0.004 0.009 178.667 178.678 scf_env_do_scf 11 5.9 0.001 0.001 161.702 161.725 velocity_verlet 10 3.0 0.001 0.002 122.866 122.869 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.119 115.121 qs_scf_new_mos 116 7.6 0.001 0.001 79.923 80.235 qs_scf_loop_do_ot 116 8.6 0.001 0.001 79.922 80.234 dbcsr_multiply_generic 2485 12.5 0.188 0.192 78.962 79.568 ot_scf_mini 116 9.6 0.004 0.004 75.466 75.746 multiply_cannon 2485 13.5 0.555 0.585 54.241 57.114 multiply_cannon_loop 2485 14.5 1.179 1.202 50.538 52.121 init_scf_loop 11 6.9 0.001 0.006 46.457 46.458 prepare_preconditioner 11 7.9 0.000 0.000 42.339 42.365 make_preconditioner 11 8.9 0.000 0.002 42.339 42.365 ot_mini 116 10.6 0.001 0.001 41.965 42.243 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.865 40.954 multiply_cannon_multrec 29820 15.5 14.060 19.151 25.960 30.569 rebuild_ks_matrix 127 8.3 0.001 0.001 28.871 29.133 qs_ks_build_kohn_sham_matrix 127 9.3 0.020 0.032 28.871 29.132 mp_waitall_1 152434 16.5 17.523 27.296 17.523 27.296 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.090 26.322 qs_ot_get_derivative 116 11.6 0.001 0.002 22.531 22.822 make_m2s 4970 13.5 0.094 0.100 20.415 21.457 make_images 4970 14.5 1.949 2.260 20.112 21.154 qs_ot_get_p 127 10.4 0.001 0.001 19.828 20.130 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.883 19.402 apply_single 127 13.6 0.001 0.001 18.883 19.402 ot_diis_step 116 11.6 0.017 0.018 19.308 19.310 cp_fm_upper_to_full 104 14.7 11.247 16.676 11.247 16.676 cp_fm_cholesky_invert 11 10.9 16.031 16.040 16.031 16.040 qs_ot_p2m_diag 82 11.4 0.339 0.385 15.735 15.785 multiply_cannon_metrocomm3 29820 15.5 0.048 0.050 6.191 14.724 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.470 14.471 sum_up_and_integrate 127 10.3 0.140 0.152 13.916 13.942 integrate_v_rspace 127 11.3 0.003 0.004 13.776 13.807 make_images_data 4970 15.5 0.063 0.067 11.112 13.113 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.991 13.028 calculate_rho_elec 127 8.7 0.173 0.189 12.990 13.028 dbcsr_complete_redistribute 393 12.7 1.514 1.644 9.049 12.877 multiply_cannon_sync_h2d 29820 15.5 11.641 12.279 11.641 12.279 hybrid_alltoall_any 5155 16.4 0.521 2.182 9.817 12.007 dbcsr_mm_accdrv_process 61748 16.2 7.309 8.347 11.479 11.995 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.687 11.508 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.430 11.432 cp_fm_redistribute_end 82 14.4 1.943 11.356 1.957 11.362 cp_fm_diag_elpa_base 82 14.4 8.788 10.769 9.372 11.255 init_scf_run 11 5.9 0.000 0.001 10.689 10.691 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.689 10.690 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.451 10.167 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.210 9.421 mp_alltoall_i22 712 14.1 5.622 9.357 5.622 9.357 pw_transfer 1535 11.6 0.084 0.098 7.828 7.893 grid_integrate_task_list 127 12.3 7.478 7.848 7.478 7.848 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.605 7.677 cp_fm_cholesky_decompose 22 10.9 7.519 7.675 7.519 7.675 wfi_extrapolate 11 7.9 0.001 0.001 7.558 7.558 multiply_cannon_metrocomm4 24850 15.5 0.075 0.086 2.748 6.931 fft_wrap_pw1pw2_140 519 13.2 0.477 0.486 6.735 6.818 density_rs2pw 127 9.7 0.005 0.006 6.194 6.752 mp_irecv_dv 75445 16.2 2.604 6.665 2.604 6.665 calculate_dm_sparse 127 9.5 0.001 0.001 6.149 6.242 fft3d_ps 1281 14.7 2.771 2.827 5.846 5.899 grid_collocate_task_list 127 9.7 5.169 5.552 5.169 5.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.467 5.511 rs_pw_transfer 1038 11.9 0.013 0.014 4.628 5.273 mp_alltoall_d11v 2401 14.1 4.733 5.219 4.733 5.219 qs_energies_init_hamiltonians 11 5.9 0.025 0.046 4.750 4.754 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.470 4.556 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.405 4.482 potential_pw2rs 127 12.3 0.023 0.023 4.202 4.215 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.114 4.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.442000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1078.545455, yerr=20.128513 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 1.504059E+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 10977 1175. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 169.406 169.407 qs_mol_dyn_low 1 2.0 0.003 0.003 169.000 169.012 qs_forces 11 3.9 0.003 0.003 168.899 168.910 qs_energies 11 4.9 0.001 0.002 161.601 161.614 scf_env_do_scf 11 5.9 0.001 0.001 144.679 144.683 velocity_verlet 10 3.0 0.001 0.002 110.884 110.888 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 109.119 109.120 qs_scf_new_mos 117 7.6 0.001 0.001 74.071 74.143 qs_scf_loop_do_ot 117 8.6 0.001 0.001 74.070 74.143 dbcsr_multiply_generic 2507 12.6 0.183 0.188 73.617 73.950 ot_scf_mini 117 9.6 0.004 0.004 69.647 69.697 multiply_cannon 2507 13.6 0.589 0.626 54.112 58.464 multiply_cannon_loop 2507 14.6 0.446 0.457 49.483 50.122 ot_mini 117 10.6 0.001 0.001 39.196 39.244 init_scf_loop 11 6.9 0.000 0.000 35.408 35.409 mp_waitall_1 129618 16.6 25.575 32.736 25.575 32.736 prepare_preconditioner 11 7.9 0.000 0.000 31.627 31.659 make_preconditioner 11 8.9 0.000 0.000 31.627 31.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.530 29.791 rebuild_ks_matrix 128 8.3 0.001 0.001 28.207 28.262 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.206 28.261 qs_ks_update_qs_env 128 7.6 0.001 0.001 25.757 25.806 multiply_cannon_multrec 10028 15.6 10.425 14.524 17.798 20.565 ot_diis_step 117 11.6 0.020 0.021 19.675 19.675 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.263 19.507 apply_single 128 13.6 0.001 0.001 19.263 19.506 qs_ot_get_derivative 117 11.6 0.002 0.002 19.448 19.503 multiply_cannon_metrocomm3 10028 15.6 0.024 0.025 12.125 19.143 make_m2s 5014 13.6 0.066 0.071 15.969 18.125 cp_fm_cholesky_invert 11 10.9 18.044 18.050 18.044 18.050 make_images 5014 14.6 2.315 2.861 15.663 17.810 qs_ot_get_p 128 10.4 0.001 0.001 17.620 17.696 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.938 13.961 sum_up_and_integrate 128 10.3 0.180 0.190 13.835 13.883 integrate_v_rspace 128 11.3 0.004 0.004 13.655 13.712 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.866 12.895 calculate_rho_elec 128 8.7 0.258 0.269 12.865 12.894 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.809 12.811 make_images_data 5014 15.6 0.054 0.061 9.668 12.231 multiply_cannon_sync_h2d 10028 15.6 11.656 12.057 11.656 12.057 hybrid_alltoall_any 5200 16.5 0.793 3.591 9.383 11.908 init_scf_run 11 5.9 0.000 0.001 10.219 10.219 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.219 10.219 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.806 9.807 cp_fm_diag_elpa_base 83 14.4 9.563 9.635 9.797 9.797 cp_fm_cholesky_decompose 22 10.9 7.967 8.098 7.967 8.098 grid_integrate_task_list 128 12.3 7.810 8.054 7.810 8.054 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.828 7.853 dbcsr_mm_accdrv_process 20762 16.1 2.581 3.475 7.003 7.682 pw_transfer 1547 11.6 0.082 0.090 7.437 7.456 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.216 7.231 wfi_extrapolate 11 7.9 0.001 0.001 7.182 7.182 mp_allgather_i34 2507 14.6 2.826 7.030 2.826 7.030 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.314 7.013 fft_wrap_pw1pw2_140 523 13.2 0.503 0.527 6.360 6.374 calculate_dm_sparse 128 9.5 0.001 0.001 6.110 6.192 density_rs2pw 128 9.7 0.005 0.006 5.746 6.089 grid_collocate_task_list 128 9.7 5.488 5.758 5.488 5.758 dbcsr_complete_redistribute 395 12.7 2.138 2.203 5.158 5.534 fft3d_ps 1291 14.7 2.723 2.788 5.373 5.411 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.203 5.208 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.192 5.200 mp_alltoall_d11v 2415 14.1 4.404 4.758 4.404 4.758 rs_pw_transfer 1046 11.9 0.013 0.013 4.038 4.407 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.111 4.141 potential_pw2rs 128 12.3 0.027 0.027 3.950 3.963 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.570 3.893 multiply_cannon_metrocomm4 7521 15.6 0.023 0.025 1.841 3.875 mp_irecv_dv 28860 15.9 1.804 3.808 1.804 3.808 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.681 3.718 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.368 3.680 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.496 3.571 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.426 3.447 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.407000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1381.727273, yerr=51.610589 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.993684E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 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 0.044 0.077 299.180 299.181 qs_mol_dyn_low 1 2.0 0.003 0.003 298.560 298.572 qs_forces 11 3.9 0.003 0.003 298.466 298.469 qs_energies 11 4.9 0.002 0.002 289.836 289.854 scf_env_do_scf 11 5.9 0.001 0.001 267.706 267.719 velocity_verlet 10 3.0 0.002 0.002 214.469 214.477 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 140.909 140.912 init_scf_loop 11 6.9 0.000 0.000 126.530 126.536 prepare_preconditioner 11 7.9 0.000 0.000 121.765 121.790 make_preconditioner 11 8.9 0.000 0.000 121.765 121.790 make_full_inverse_cholesky 11 9.9 0.000 0.000 97.648 118.869 qs_scf_new_mos 118 7.6 0.001 0.001 97.989 98.093 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.988 98.092 ot_scf_mini 118 9.6 0.004 0.004 93.111 93.129 dbcsr_multiply_generic 2535 12.6 0.217 0.224 85.650 86.148 cp_fm_upper_to_full 106 14.8 53.169 76.230 53.169 76.230 multiply_cannon 2535 13.6 0.712 0.752 60.139 61.507 multiply_cannon_loop 2535 14.6 0.479 0.492 56.118 57.365 ot_mini 118 10.6 0.001 0.001 46.090 46.128 dbcsr_complete_redistribute 397 12.7 3.983 4.029 29.956 42.795 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.470 39.347 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.071 36.867 mp_alltoall_i22 720 14.1 21.896 34.745 21.896 34.745 cp_fm_cholesky_invert 11 10.9 34.635 34.641 34.635 34.641 rebuild_ks_matrix 129 8.3 0.001 0.001 34.135 34.166 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.134 34.166 mp_waitall_1 106626 16.7 29.619 33.233 29.619 33.233 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.865 31.899 qs_ot_get_p 129 10.4 0.001 0.001 31.044 31.148 qs_ot_p2m_diag 84 11.4 0.889 0.893 26.594 26.623 qs_ot_get_derivative 118 11.6 0.002 0.002 25.253 25.271 cp_dbcsr_syevd 84 12.4 0.006 0.006 24.783 24.786 make_m2s 5070 13.6 0.077 0.079 20.420 21.528 cp_fm_diag_elpa 84 13.4 0.000 0.000 21.335 21.336 cp_fm_diag_elpa_base 84 14.4 16.872 18.497 21.330 21.330 make_images 5070 14.6 3.809 3.890 19.938 21.050 ot_diis_step 118 11.6 0.022 0.023 20.781 20.781 multiply_cannon_metrocomm3 10140 15.6 0.025 0.025 19.673 20.599 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.911 20.152 apply_single 129 13.6 0.001 0.001 19.910 20.152 multiply_cannon_multrec 10140 15.6 10.568 12.359 18.056 18.124 sum_up_and_integrate 129 10.3 0.324 0.326 16.058 16.146 integrate_v_rspace 129 11.3 0.004 0.004 15.734 15.821 multiply_cannon_sync_h2d 10140 15.6 15.781 15.793 15.781 15.793 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.334 15.366 calculate_rho_elec 129 8.7 0.486 0.487 15.334 15.365 hybrid_alltoall_any 5257 16.5 1.311 3.061 10.952 12.817 make_images_data 5070 15.6 0.061 0.065 10.886 12.794 init_scf_run 11 5.9 0.000 0.001 12.179 12.180 scf_env_initial_rho_setup 11 6.9 0.053 0.054 12.179 12.180 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.041 10.075 cp_fm_cholesky_decompose 22 10.9 9.653 9.666 9.653 9.666 dbcsr_mm_accdrv_process 20958 16.1 3.857 5.816 7.251 9.108 wfi_extrapolate 11 7.9 0.001 0.001 8.973 8.973 grid_integrate_task_list 129 12.3 8.600 8.781 8.600 8.781 pw_transfer 1559 11.6 0.091 0.091 8.697 8.704 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.462 8.469 qs_energies_init_hamiltonians 11 5.9 0.041 0.054 7.929 7.929 fft_wrap_pw1pw2_140 527 13.2 0.543 0.545 7.424 7.437 mp_alltoall_d11v 2429 14.1 7.213 7.383 7.213 7.383 calculate_dm_sparse 129 9.5 0.001 0.001 6.737 6.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.542 6.628 fft3d_ps 1301 14.7 2.776 2.787 6.493 6.509 grid_collocate_task_list 129 9.7 6.386 6.425 6.386 6.425 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.246 6.352 density_rs2pw 129 9.7 0.005 0.005 6.305 6.336 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.181000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2635.272727, yerr=189.283861 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 1.261457E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.032 0.077 85.370 85.372 qs_energies 1 2.0 0.000 0.001 84.811 84.819 ls_scf 1 3.0 0.000 0.000 83.864 83.872 dbcsr_multiply_generic 111 6.7 0.015 0.015 72.911 73.085 multiply_cannon 111 7.7 0.018 0.021 55.990 57.150 multiply_cannon_loop 111 8.7 0.206 0.221 52.554 53.990 ls_scf_main 1 4.0 0.000 0.000 52.402 52.403 density_matrix_trs4 2 5.0 0.002 0.003 46.916 46.997 ls_scf_init_scf 1 4.0 0.000 0.001 28.413 28.414 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.356 27.402 mp_waitall_1 11316 10.9 22.703 25.617 22.703 25.617 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.189 25.205 multiply_cannon_multrec 2664 9.7 8.154 8.952 15.550 17.411 multiply_cannon_sync_h2d 2664 9.7 13.552 15.505 13.552 15.505 make_m2s 222 7.7 0.008 0.011 13.191 13.643 make_images 222 8.7 0.099 0.110 13.169 13.622 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.820 12.800 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.471 8.473 make_images_data 222 9.7 0.004 0.005 7.722 8.261 hybrid_alltoall_any 227 10.6 0.216 1.837 6.697 8.041 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.631 7.016 8.016 dbcsr_mm_accdrv_process_sort 4760 11.4 6.304 7.231 6.304 7.231 calculate_norms 4752 9.8 5.503 6.220 5.503 6.220 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.034 5.185 mp_sum_l 807 5.4 3.210 4.568 3.210 4.568 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.075 3.705 mp_irecv_dv 6231 10.9 2.057 3.676 2.057 3.676 make_images_sizes 222 9.7 0.000 0.000 0.760 3.630 mp_alltoall_i44 222 10.7 0.760 3.630 0.760 3.630 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.399 3.328 arnoldi_extremal 4 6.8 0.000 0.000 3.269 3.297 arnoldi_normal_ev 4 7.8 0.001 0.002 3.269 3.297 build_subspace 16 8.4 0.009 0.012 3.174 3.176 ls_scf_post 1 4.0 0.000 0.000 3.049 3.056 ls_scf_store_result 1 5.0 0.000 0.000 2.869 2.908 dbcsr_special_finalize 555 9.7 0.005 0.006 2.351 2.708 dbcsr_merge_single_wm 555 10.7 0.469 0.621 2.343 2.701 make_images_pack 222 9.7 2.211 2.637 2.213 2.638 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.329 2.588 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.470 2.066 2.471 dbcsr_sort_data 658 11.4 2.131 2.441 2.131 2.441 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.241 2.331 buffer_matrices_ensure_size 222 8.7 1.788 2.186 1.788 2.186 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.745 1.747 rebuild_ks_matrix 3 7.3 0.000 0.000 1.736 1.737 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.736 1.737 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.372000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 2.104259E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.042 0.062 91.988 91.990 qs_energies 1 2.0 0.002 0.011 91.428 91.462 ls_scf 1 3.0 0.000 0.001 90.093 90.096 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.093 76.455 multiply_cannon 111 7.7 0.029 0.037 53.424 57.770 ls_scf_main 1 4.0 0.000 0.003 55.567 55.572 multiply_cannon_loop 111 8.7 0.116 0.123 50.154 53.702 density_matrix_trs4 2 5.0 0.002 0.005 49.885 50.072 ls_scf_init_scf 1 4.0 0.000 0.001 30.996 30.997 mp_waitall_1 9246 10.9 21.705 30.709 21.705 30.709 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.779 29.885 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.374 27.390 multiply_cannon_multrec 1332 9.7 13.047 16.470 22.339 26.859 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.075 21.146 make_m2s 222 7.7 0.006 0.007 15.598 16.291 make_images 222 8.7 1.574 1.918 15.568 16.263 dbcsr_mm_accdrv_process 4041 10.4 0.291 0.451 8.893 10.495 dbcsr_mm_accdrv_process_sort 4041 11.4 8.486 10.044 8.486 10.044 make_images_data 222 9.7 0.004 0.004 9.049 9.992 hybrid_alltoall_any 227 10.6 0.520 2.450 8.453 9.351 mp_sum_l 807 5.4 5.776 9.046 5.776 9.046 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.244 7.955 mp_irecv_dv 3311 11.0 3.225 7.901 3.225 7.901 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.506 7.085 calculate_norms 2376 9.8 6.000 6.634 6.000 6.634 multiply_cannon_sync_h2d 1332 9.7 4.731 5.933 4.731 5.933 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.061 5.308 arnoldi_extremal 4 6.8 0.000 0.000 4.692 4.724 arnoldi_normal_ev 4 7.8 0.001 0.005 4.692 4.724 build_subspace 16 8.4 0.014 0.021 4.434 4.437 ls_scf_post 1 4.0 0.000 0.002 3.530 3.534 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.164 3.371 ls_scf_store_result 1 5.0 0.000 0.000 3.236 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.219 2.743 3.221 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.198 2.741 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.547 2.646 mp_allgather_i34 111 8.7 0.967 2.487 0.967 2.487 make_images_pack 222 9.7 2.027 2.431 2.030 2.433 dbcsr_sort_data 436 11.2 1.814 2.083 1.814 2.083 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.897 1.899 rebuild_ks_matrix 3 7.3 0.000 0.000 1.884 1.886 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.884 1.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.990000, 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/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 2.696462E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.040 0.098 94.229 94.230 qs_energies 1 2.0 0.000 0.001 93.550 93.564 ls_scf 1 3.0 0.001 0.011 92.126 92.130 dbcsr_multiply_generic 111 6.7 0.015 0.017 76.949 77.187 ls_scf_main 1 4.0 0.000 0.003 57.825 57.831 multiply_cannon 111 7.7 0.041 0.084 52.993 57.639 multiply_cannon_loop 111 8.7 0.100 0.109 49.387 52.771 density_matrix_trs4 2 5.0 0.002 0.004 51.885 52.067 mp_waitall_1 7374 11.0 24.246 33.792 24.246 33.792 ls_scf_init_scf 1 4.0 0.000 0.002 30.742 30.746 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.611 29.672 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.233 27.247 multiply_cannon_multrec 888 9.7 12.467 15.199 21.069 24.393 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.278 22.646 make_m2s 222 7.7 0.006 0.007 17.253 18.496 make_images 222 8.7 1.978 2.303 17.215 18.459 make_images_data 222 9.7 0.003 0.004 9.948 10.946 hybrid_alltoall_any 227 10.6 0.618 2.868 9.576 10.877 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.417 8.128 9.370 dbcsr_mm_accdrv_process_sort 3754 11.4 7.758 8.953 7.758 8.953 mp_sum_l 807 5.4 5.159 8.736 5.159 8.736 multiply_cannon_sync_h2d 888 9.7 5.970 7.195 5.970 7.195 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.009 6.967 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.493 6.724 mp_irecv_dv 2335 11.1 2.478 6.669 2.478 6.669 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.635 6.649 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.965 5.210 arnoldi_extremal 4 6.8 0.000 0.000 5.153 5.171 arnoldi_normal_ev 4 7.8 0.001 0.005 5.153 5.171 build_subspace 16 8.4 0.014 0.020 4.848 4.855 calculate_norms 1584 9.8 4.328 4.723 4.328 4.723 mp_allgather_i34 111 8.7 1.444 3.927 1.444 3.927 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.485 3.821 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.594 3.030 3.597 ls_scf_post 1 4.0 0.002 0.018 3.559 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.311 3.382 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.761 2.849 dbcsr_sort_data 325 11.1 1.901 2.162 1.901 2.162 make_images_pack 222 9.7 1.832 2.151 1.835 2.154 make_images_sizes 222 9.7 0.000 0.000 0.946 2.072 mp_alltoall_i44 222 10.7 0.946 2.072 0.946 2.072 dbcsr_data_release 9322 10.9 1.303 1.977 1.303 1.977 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.230000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 3.332444E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.060 97.388 97.390 qs_energies 1 2.0 0.000 0.001 96.672 96.680 ls_scf 1 3.0 0.000 0.001 95.027 95.035 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.819 79.048 ls_scf_main 1 4.0 0.000 0.002 59.112 59.115 multiply_cannon 111 7.7 0.060 0.180 51.696 56.423 density_matrix_trs4 2 5.0 0.003 0.012 53.027 53.160 multiply_cannon_loop 111 8.7 0.115 0.126 46.637 50.007 ls_scf_init_scf 1 4.0 0.001 0.005 32.619 32.625 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.447 31.522 mp_waitall_1 6438 11.0 22.878 30.808 22.878 30.808 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.961 28.972 multiply_cannon_multrec 1332 9.7 14.218 17.223 22.155 24.696 make_m2s 222 7.7 0.006 0.007 21.232 22.612 make_images 222 8.7 3.141 3.592 21.182 22.564 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.213 17.508 make_images_data 222 9.7 0.004 0.004 11.875 13.559 hybrid_alltoall_any 227 10.6 0.797 3.811 11.311 13.202 dbcsr_mm_accdrv_process 3641 10.4 0.204 0.401 7.562 9.099 dbcsr_mm_accdrv_process_sort 3641 11.4 7.189 8.676 7.189 8.676 mp_sum_l 807 5.4 4.323 7.757 4.323 7.757 multiply_cannon_sync_h2d 1332 9.7 5.489 6.252 5.489 6.252 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.291 6.041 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.063 6.027 mp_irecv_dv 3229 10.9 2.041 5.945 2.041 5.945 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.612 5.460 arnoldi_extremal 4 6.8 0.000 0.003 5.252 5.266 arnoldi_normal_ev 4 7.8 0.001 0.007 5.252 5.266 build_subspace 16 8.4 0.015 0.024 4.899 4.910 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.579 4.758 mp_allgather_i34 111 8.7 2.182 4.591 2.182 4.591 calculate_norms 2376 9.8 4.171 4.540 4.171 4.540 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.628 3.896 dbcsr_matrix_vector_mult_local 304 10.0 3.203 3.706 3.205 3.708 dbcsr_sort_data 658 11.4 3.116 3.508 3.116 3.508 ls_scf_post 1 4.0 0.000 0.002 3.296 3.305 dbcsr_special_finalize 555 9.7 0.006 0.007 2.874 3.220 dbcsr_merge_single_wm 555 10.7 0.538 0.669 2.866 3.212 ls_scf_store_result 1 5.0 0.000 0.000 3.009 3.070 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.914 2.977 dbcsr_data_release 10477 10.7 1.587 2.443 1.587 2.443 dbcsr_finalize 304 7.8 0.049 0.061 1.794 2.036 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.390000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2749.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 4.629635E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.048 0.059 99.813 99.814 qs_energies 1 2.0 0.000 0.000 99.052 99.059 ls_scf 1 3.0 0.000 0.000 97.126 97.132 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.737 78.919 ls_scf_main 1 4.0 0.000 0.000 62.667 62.667 multiply_cannon 111 7.7 0.091 0.142 55.840 61.362 density_matrix_trs4 2 5.0 0.002 0.003 55.619 55.715 multiply_cannon_loop 111 8.7 0.069 0.078 51.280 52.934 mp_waitall_1 5481 11.0 27.243 33.281 27.243 33.281 ls_scf_init_scf 1 4.0 0.000 0.000 30.820 30.825 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.653 29.687 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.478 27.490 multiply_cannon_multrec 444 9.7 14.068 16.577 21.175 24.706 make_m2s 222 7.7 0.004 0.005 17.952 20.487 make_images 222 8.7 3.724 4.438 17.890 20.427 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.484 16.630 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.316 14.413 make_images_data 222 9.7 0.003 0.004 10.243 12.621 hybrid_alltoall_any 227 10.6 0.790 3.776 9.976 12.378 multiply_cannon_sync_h2d 444 9.7 6.521 9.022 6.521 9.022 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.342 6.814 7.970 dbcsr_mm_accdrv_process_sort 3003 11.4 6.499 7.631 6.499 7.631 mp_allgather_i34 111 8.7 2.788 7.062 2.788 7.062 arnoldi_extremal 4 6.8 0.000 0.000 5.926 5.940 arnoldi_normal_ev 4 7.8 0.001 0.004 5.926 5.940 mp_sum_l 807 5.4 2.981 5.581 2.981 5.581 build_subspace 16 8.4 0.015 0.019 5.536 5.549 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.640 4.815 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.627 4.509 mp_irecv_dv 1241 11.2 1.608 4.485 1.608 4.485 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.222 4.457 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.055 4.235 dbcsr_matrix_vector_mult_local 304 10.0 3.718 4.203 3.721 4.205 calculate_norms 792 9.8 3.523 3.673 3.523 3.673 make_images_sizes 222 9.7 0.000 0.000 1.047 3.663 mp_alltoall_i44 222 10.7 1.046 3.663 1.046 3.663 ls_scf_post 1 4.0 0.000 0.000 3.639 3.647 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.405 3.513 ls_scf_store_result 1 5.0 0.000 0.000 3.422 3.470 dbcsr_finalize 304 7.8 0.062 0.078 2.204 2.326 dbcsr_merge_all 275 8.9 0.475 0.533 2.048 2.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.814000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3609.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/32b39844d4254f7d1efd5072dc95c1a6354e6919_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 8.719921E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 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 0.079 0.098 107.490 107.490 qs_energies 1 2.0 0.000 0.000 106.083 106.097 ls_scf 1 3.0 0.000 0.000 103.194 103.209 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.323 77.479 ls_scf_main 1 4.0 0.000 0.000 65.568 65.569 density_matrix_trs4 2 5.0 0.002 0.003 56.714 56.791 multiply_cannon 111 7.7 0.162 0.263 49.812 52.125 multiply_cannon_loop 111 8.7 0.067 0.070 46.417 47.104 ls_scf_init_scf 1 4.0 0.000 0.000 33.948 33.949 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.626 32.636 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.786 29.794 mp_waitall_1 4569 11.1 21.972 25.172 21.972 25.172 make_m2s 222 7.7 0.005 0.005 23.790 24.703 make_images 222 8.7 4.593 5.018 23.684 24.595 multiply_cannon_multrec 444 9.7 17.909 18.560 22.563 23.094 hybrid_alltoall_any 227 10.6 1.660 3.612 12.706 15.667 make_images_data 222 9.7 0.003 0.003 12.973 15.375 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.392 11.209 multiply_cannon_sync_h2d 444 9.7 8.848 8.897 8.848 8.897 arnoldi_extremal 4 6.8 0.000 0.000 7.452 7.469 arnoldi_normal_ev 4 7.8 0.002 0.008 7.452 7.469 build_subspace 16 8.4 0.026 0.037 6.912 6.920 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.511 5.679 dbcsr_matrix_vector_mult_local 304 10.0 5.061 5.382 5.063 5.385 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.029 5.247 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.819 4.912 dbcsr_mm_accdrv_process 1814 10.4 0.237 0.338 4.485 4.607 dbcsr_mm_accdrv_process_sort 1814 11.4 4.181 4.314 4.181 4.314 ls_scf_post 1 4.0 0.000 0.000 3.678 3.692 make_images_sizes 222 9.7 0.000 0.000 1.470 3.647 mp_alltoall_i44 222 10.7 1.470 3.647 1.470 3.647 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.435 calculate_norms 792 9.8 3.241 3.281 3.241 3.281 dbcsr_finalize 304 7.8 0.082 0.090 3.085 3.153 mp_allgather_i34 111 8.7 1.027 3.150 1.027 3.150 dbcsr_merge_all 275 8.9 0.894 0.922 2.870 2.932 dbcsr_complete_redistribute 5 7.6 1.442 1.477 2.775 2.866 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.858 2.858 dbcsr_data_release 12724 10.6 2.325 2.852 2.325 2.852 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.543 dbcsr_sort_data 325 11.1 2.440 2.518 2.440 2.518 dbcsr_new_transposed 4 7.5 0.245 0.256 2.292 2.311 dbcsr_frobenius_norm 74 6.6 2.057 2.130 2.194 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.211 dbcsr_add_anytype 103 7.2 0.859 0.892 2.134 2.211 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.198 2.199 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=107.490000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6811.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 32b39844d4254f7d1efd5072dc95c1a6354e6919 Summary: empty Status: OK