=== 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: d7412ee8675cb76473b00ae3e51e7018021298a2 ################# 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/01 job id: 42785736 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/02 job id: 42785738 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/03 job id: 42785748 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/04 job id: 42785761 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/05 job id: 42785767 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/06 job id: 42785770 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/07 job id: 42785774 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/08 job id: 42785779 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/09 job id: 42785780 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/10 job id: 42785781 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/11 job id: 42785785 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/12 job id: 42785786 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/13 job id: 42785787 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/14 job id: 42785790 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/15 job id: 42785792 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/16 job id: 42785794 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/17 job id: 42785795 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/18 job id: 42785796 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/19 job id: 42785797 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/20 job id: 42785800 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/21 job id: 42785802 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/22 job id: 42785803 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/23 job id: 42785805 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/24 job id: 42785807 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/25 job id: 42785809 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/26 job id: 42785813 --- 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/d7412ee8675cb76473b00ae3e51e7018021298a2_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 0.148 0.203 134.977 134.978 farming_run 1 2.0 133.714 133.725 134.822 134.924 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.457865E+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 0.012 0.027 115.592 115.592 qs_energies 1 2.0 0.000 0.000 115.267 115.268 mp2_main 1 3.0 0.000 0.000 113.298 113.299 mp2_gpw_main 1 4.0 0.019 0.026 112.440 112.440 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 93.496 93.903 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.493 55.900 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.824 47.195 get_2c_integrals 1 6.0 0.000 0.000 37.150 37.832 integrate_v_rspace 273 8.0 0.435 0.450 25.139 30.319 pw_transfer 6555 10.6 0.370 0.400 27.404 27.994 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.071 26.701 grid_integrate_task_list 273 9.0 20.947 26.641 20.947 26.641 fft_wrap_pw1pw2_100 2178 12.4 1.180 1.336 23.634 24.267 compute_2c_integrals 1 7.0 0.002 0.002 19.611 19.612 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.904 19.322 mp2_eri_2c_integrate_gpw 1 9.0 2.373 2.429 18.901 19.319 rpa_ri_compute_en 1 5.0 0.000 0.000 18.836 19.022 cp_fm_cholesky_decompose 12 8.2 17.506 18.188 17.506 18.188 cholesky_decomp 1 7.0 0.000 0.000 16.387 17.060 fft3d_s 5443 13.4 16.178 16.766 16.200 16.788 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.606 10.850 15.606 calculate_wavefunction 272 8.0 5.474 5.624 12.602 13.211 rpa_num_int 1 6.0 0.000 0.000 10.603 10.603 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.557 10.564 calc_mat_Q 8 8.0 0.000 0.000 9.384 9.488 contract_S_to_Q 8 9.0 0.000 0.000 8.804 8.905 calc_potential_gpw 544 9.5 0.005 0.006 8.272 8.704 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.222 8.518 parallel_gemm_fm 14 9.1 0.000 0.000 8.393 8.465 parallel_gemm_fm_cosma 14 10.1 8.393 8.465 8.393 8.465 potential_pw2rs 545 10.0 0.107 0.109 7.685 8.278 create_integ_mat 1 6.0 0.006 0.008 7.733 7.733 collocate_single_gaussian 272 10.0 0.040 0.043 7.486 7.720 array2fm 1 7.0 0.000 0.000 6.710 7.187 pw_scatter_s 2720 13.7 4.454 4.828 4.454 4.828 pw_gather_s 2722 13.2 3.888 4.309 3.888 4.309 array2fm_buffer_send 1 8.0 3.032 3.115 3.032 3.115 pw_poisson_solve 545 10.5 1.122 1.173 2.164 2.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.440066, 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/d7412ee8675cb76473b00ae3e51e7018021298a2_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 0.029 0.041 396.513 396.514 farming_run 1 2.0 395.579 395.607 396.474 396.478 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223700E+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 11472073. 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 0.013 0.068 209.266 209.266 qs_energies 1 2.0 0.000 0.000 209.027 209.034 scf_env_do_scf 1 3.0 0.000 0.000 105.969 105.969 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.128 105.136 rebuild_ks_matrix 4 6.0 0.000 0.000 105.127 105.135 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.127 105.135 hfx_ks_matrix 4 8.0 0.001 0.001 104.755 104.758 integrate_four_center 4 9.0 0.144 0.460 104.754 104.757 mp2_main 1 3.0 0.002 0.018 102.774 102.781 mp2_gpw_main 1 4.0 0.040 0.067 101.914 101.925 integrate_four_center_main 4 10.0 0.104 0.573 96.948 98.914 integrate_four_center_bin 269 11.0 96.844 98.853 96.844 98.853 init_scf_loop 1 4.0 0.000 0.000 91.827 91.827 mp2_ri_gpw_compute_in 1 5.0 0.080 0.114 74.825 75.885 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.005 54.499 55.572 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.084 47.262 integrate_v_rspace 95 8.0 0.398 0.573 28.423 33.404 pw_transfer 2240 10.6 0.145 0.158 29.834 30.215 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.865 29.237 grid_integrate_task_list 95 9.0 23.718 28.898 23.718 28.898 mp2_ri_gpw_compute_en 1 5.0 0.066 0.081 26.888 28.607 ao_to_mo_and_store_B_mult_1 91 7.0 10.734 27.113 10.734 27.113 fft_wrap_pw1pw2_100 730 12.4 1.238 1.417 26.601 26.966 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.902 25.168 25.177 get_2c_integrals 1 6.0 0.002 0.011 20.219 20.242 compute_2c_integrals 1 7.0 0.022 0.043 19.201 19.203 compute_2c_integrals_loop_lm 1 8.0 0.003 0.005 18.745 19.052 mp2_eri_2c_integrate_gpw 1 9.0 1.750 1.827 18.742 19.051 fft3d_s 1823 13.4 18.411 18.662 18.425 18.676 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.140 14.140 calculate_wavefunction 91 8.0 2.039 2.067 9.766 10.020 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.579 8.813 9.246 potential_pw2rs 186 10.0 0.033 0.035 8.584 9.137 local_gemm 172 8.0 8.258 8.701 8.258 8.701 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.509 8.127 8.600 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.164 8.564 calc_potential_gpw 182 9.5 0.002 0.002 7.889 8.056 collocate_single_gaussian 91 10.0 0.018 0.037 7.822 8.049 mp_sendrecv_dm3 2068 8.0 6.166 6.645 6.166 6.645 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.420 6.340 6.420 pw_gather_s 912 13.2 4.927 5.392 4.927 5.392 mp_sync 38 10.4 2.404 4.530 2.404 4.530 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.906736, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 451.502080E+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 1339086. 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 0.036 0.079 53.784 53.789 qs_mol_dyn_low 1 2.0 0.003 0.003 53.120 53.136 qs_forces 11 3.9 0.002 0.002 53.040 53.041 qs_energies 11 4.9 0.009 0.045 51.583 51.598 scf_env_do_scf 11 5.9 0.001 0.001 44.353 44.353 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.268 42.269 dbcsr_multiply_generic 2286 12.5 0.094 0.102 34.106 34.594 qs_scf_new_mos 108 7.5 0.000 0.001 32.343 32.640 qs_scf_loop_do_ot 108 8.5 0.001 0.001 32.342 32.640 ot_scf_mini 108 9.5 0.002 0.002 30.663 30.848 multiply_cannon 2286 13.5 0.179 0.186 26.253 27.959 multiply_cannon_loop 2286 14.5 1.501 1.569 25.553 27.303 velocity_verlet 10 3.0 0.001 0.001 25.482 25.483 ot_mini 108 10.5 0.001 0.001 19.611 19.870 qs_ot_get_derivative 108 11.5 0.001 0.001 16.609 16.799 mp_waitall_1 267858 16.1 9.051 15.358 9.051 15.358 multiply_cannon_metrocomm3 54864 15.5 0.067 0.071 6.233 13.648 multiply_cannon_multrec 54864 15.5 4.179 6.432 7.447 10.868 rebuild_ks_matrix 119 8.3 0.000 0.000 7.848 8.000 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.847 7.999 multiply_cannon_sync_h2d 54864 15.5 5.852 7.772 5.852 7.772 mp_sum_l 7207 12.9 5.912 7.555 5.912 7.555 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.934 7.077 qs_ot_get_p 119 10.4 0.001 0.001 6.352 6.697 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.571 5.992 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.721 5.851 init_scf_run 11 5.9 0.000 0.001 5.706 5.707 scf_env_initial_rho_setup 11 6.9 0.001 0.005 5.706 5.707 sum_up_and_integrate 119 10.3 0.012 0.015 4.438 4.445 integrate_v_rspace 119 11.3 0.002 0.003 4.425 4.433 dbcsr_mm_accdrv_process 76910 16.1 1.187 1.821 3.191 4.426 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.082 4.180 calculate_rho_elec 119 8.7 0.011 0.016 4.082 4.179 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.943 4.038 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.503 3.599 calculate_first_density_matrix 1 7.0 0.005 0.021 3.533 3.542 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.895 3.114 apply_single 119 13.6 0.000 0.000 2.895 3.114 rs_pw_transfer 974 11.9 0.012 0.013 2.873 3.044 calculate_dm_sparse 119 9.5 0.007 0.016 2.906 3.029 ot_diis_step 108 11.5 0.006 0.006 2.724 2.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.681 2.686 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.620 2.620 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.517 2.606 jit_kernel_multiply 13 15.8 1.941 2.519 1.941 2.519 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.420 2.421 acc_transpose_blocks 54864 15.5 0.232 0.249 1.872 2.409 cp_fm_redistribute_end 50 14.0 2.198 2.400 2.203 2.402 cp_fm_diag_elpa_base 50 14.0 0.197 2.365 0.198 2.375 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.303 2.365 density_rs2pw 119 9.7 0.004 0.004 2.181 2.305 grid_integrate_task_list 119 12.3 1.997 2.107 1.997 2.107 mp_sum_d 4125 12.0 1.470 2.076 1.470 2.076 wfi_extrapolate 11 7.9 0.001 0.001 2.068 2.068 init_scf_loop 11 6.9 0.001 0.002 2.065 2.065 potential_pw2rs 119 12.3 0.004 0.004 1.824 1.838 make_m2s 4572 13.5 0.054 0.057 1.644 1.696 pw_transfer 1439 11.6 0.051 0.054 1.616 1.690 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.542 1.617 make_images 4572 14.5 0.134 0.142 1.561 1.613 mp_waitany 12084 13.8 1.396 1.589 1.396 1.589 parallel_gemm_fm 81 9.0 0.000 0.000 1.433 1.438 parallel_gemm_fm_cosma 81 10.0 1.433 1.438 1.433 1.438 mp_alltoall_d11v 2130 13.8 1.308 1.433 1.308 1.433 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.378 1.403 fft3d_ps 1201 14.6 0.356 0.459 1.320 1.389 grid_collocate_task_list 119 9.7 1.286 1.350 1.286 1.350 acc_transpose_blocks_kernels 54864 16.5 0.254 0.384 0.868 1.247 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.160 1.238 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.791 1.185 make_basis_sm 11 9.8 0.006 0.047 1.079 1.081 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.789000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 487.878656E+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 263016. 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 0.074 0.148 40.983 40.984 qs_mol_dyn_low 1 2.0 0.003 0.003 40.317 40.325 qs_forces 11 3.9 0.002 0.002 40.255 40.256 qs_energies 11 4.9 0.001 0.004 38.513 38.517 scf_env_do_scf 11 5.9 0.001 0.001 32.571 32.571 scf_env_do_scf_inner_loop 108 6.5 0.007 0.041 29.862 29.865 dbcsr_multiply_generic 2286 12.5 0.102 0.126 21.891 22.265 qs_scf_new_mos 108 7.5 0.001 0.001 20.436 20.689 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.435 20.688 ot_scf_mini 108 9.5 0.002 0.003 19.529 19.711 velocity_verlet 10 3.0 0.001 0.001 18.935 18.936 multiply_cannon 2286 13.5 0.211 0.220 16.422 17.942 multiply_cannon_loop 2286 14.5 0.900 0.974 15.217 16.559 ot_mini 108 10.5 0.001 0.001 12.148 12.389 mp_waitall_1 217478 16.2 6.459 11.508 6.459 11.508 qs_ot_get_derivative 108 11.5 0.001 0.001 9.669 9.852 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.150 9.548 multiply_cannon_multrec 27432 15.5 1.971 4.556 5.875 8.685 rebuild_ks_matrix 119 8.3 0.000 0.000 7.477 7.619 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.476 7.619 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.590 6.718 dbcsr_mm_accdrv_process 47894 16.0 2.891 5.191 3.835 5.679 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.793 4.659 qs_ot_get_p 119 10.4 0.001 0.001 4.419 4.644 init_scf_run 11 5.9 0.001 0.022 4.597 4.597 scf_env_initial_rho_setup 11 6.9 0.003 0.021 4.596 4.597 sum_up_and_integrate 119 10.3 0.024 0.027 4.431 4.442 integrate_v_rspace 119 11.3 0.002 0.002 4.407 4.419 mp_sum_l 7207 12.9 2.223 4.191 2.223 4.191 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.077 4.159 apply_single 119 13.6 0.000 0.000 3.077 4.159 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.856 3.898 calculate_rho_elec 119 8.7 0.021 0.024 3.855 3.897 make_m2s 4572 13.5 0.053 0.055 2.830 3.057 rs_pw_transfer 974 11.9 0.010 0.011 2.949 3.049 calculate_first_density_matrix 1 7.0 0.001 0.002 3.029 3.034 make_images 4572 14.5 0.205 0.265 2.742 2.968 multiply_cannon_sync_h2d 27432 15.5 2.201 2.952 2.201 2.952 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.800 2.818 init_scf_loop 11 6.9 0.000 0.001 2.684 2.707 ot_diis_step 108 11.5 0.011 0.011 2.429 2.430 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.353 2.354 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.223 2.316 density_rs2pw 119 9.7 0.004 0.004 2.196 2.283 potential_pw2rs 119 12.3 0.006 0.006 2.153 2.170 calculate_dm_sparse 119 9.5 0.000 0.001 2.052 2.125 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.058 2.060 pw_transfer 1439 11.6 0.063 0.066 1.961 1.997 jit_kernel_multiply 10 16.1 0.891 1.971 0.891 1.971 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.960 1.961 grid_integrate_task_list 119 12.3 1.836 1.936 1.836 1.936 cp_fm_redistribute_end 50 14.0 1.608 1.928 1.612 1.930 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.871 1.910 cp_fm_diag_elpa_base 50 14.0 0.307 1.859 0.315 1.892 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.002 1.809 1.852 prepare_preconditioner 11 7.9 0.000 0.000 1.711 1.737 make_preconditioner 11 8.9 0.000 0.001 1.711 1.737 make_images_data 4572 15.5 0.045 0.051 1.250 1.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.606 1.665 fft3d_ps 1201 14.6 0.499 0.553 1.580 1.614 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.079 1.604 wfi_extrapolate 11 7.9 0.001 0.001 1.507 1.507 acc_transpose_blocks 27432 15.5 0.108 0.113 1.183 1.476 fft_wrap_pw1pw2_140 487 13.2 0.076 0.084 1.431 1.471 mp_alltoall_d11v 2130 13.8 1.315 1.427 1.315 1.427 mp_allgather_i34 2286 14.5 0.641 1.411 0.641 1.411 grid_collocate_task_list 119 9.7 1.242 1.332 1.242 1.332 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.258 1.264 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.216 1.260 mp_sum_d 4125 12.0 0.670 1.122 0.670 1.122 make_images_sizes 4572 15.5 0.005 0.005 0.759 1.077 mp_alltoall_i44 4572 16.5 0.755 1.073 0.755 1.073 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.034 1.035 parallel_gemm_fm 81 9.0 0.000 0.000 1.000 1.003 parallel_gemm_fm_cosma 81 10.0 1.000 1.003 1.000 1.003 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.945 0.959 mp_alltoall_z22v 1201 16.6 0.837 0.918 0.837 0.918 make_basis_sm 11 9.8 0.001 0.001 0.882 0.884 rs_pw_transfer_PW2RS_50 119 14.3 0.592 0.611 0.828 0.874 acc_transpose_blocks_kernels 27432 16.5 0.179 0.268 0.650 0.854 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.984000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.285632E+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 0.045 0.058 34.083 34.084 qs_mol_dyn_low 1 2.0 0.003 0.003 33.570 33.580 qs_forces 11 3.9 0.002 0.003 33.507 33.508 qs_energies 11 4.9 0.001 0.001 31.832 31.835 scf_env_do_scf 11 5.9 0.000 0.001 26.112 26.113 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.368 23.368 dbcsr_multiply_generic 2286 12.5 0.096 0.100 16.638 16.706 velocity_verlet 10 3.0 0.001 0.001 15.883 15.885 qs_scf_new_mos 108 7.5 0.001 0.001 15.118 15.129 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.118 15.129 ot_scf_mini 108 9.5 0.002 0.002 14.389 14.405 multiply_cannon 2286 13.5 0.196 0.202 13.135 13.866 multiply_cannon_loop 2286 14.5 0.633 0.668 12.304 13.076 ot_mini 108 10.5 0.001 0.001 8.843 8.858 qs_ot_get_derivative 108 11.5 0.001 0.001 7.336 7.351 multiply_cannon_multrec 18288 15.5 1.914 2.789 6.805 7.073 rebuild_ks_matrix 119 8.3 0.000 0.000 6.607 6.619 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.029 6.607 6.619 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.791 5.802 dbcsr_mm_accdrv_process 38222 16.0 3.984 5.383 4.807 5.546 init_scf_run 11 5.9 0.000 0.001 4.339 4.339 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.339 4.339 mp_waitall_1 169478 16.3 3.219 4.301 3.219 4.301 sum_up_and_integrate 119 10.3 0.030 0.030 4.169 4.174 integrate_v_rspace 119 11.3 0.002 0.002 4.140 4.147 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.385 3.394 calculate_rho_elec 119 8.7 0.030 0.031 3.385 3.394 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.752 3.363 qs_ot_get_p 119 10.4 0.001 0.001 3.304 3.340 calculate_first_density_matrix 1 7.0 0.000 0.001 3.059 3.059 init_scf_loop 11 6.9 0.000 0.000 2.726 2.728 rs_pw_transfer 974 11.9 0.009 0.010 2.509 2.611 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.345 apply_single 119 13.6 0.000 0.000 2.034 2.345 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.366 2.324 make_m2s 4572 13.5 0.045 0.045 1.974 2.136 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.124 2.130 density_rs2pw 119 9.7 0.004 0.004 1.967 2.076 make_images 4572 14.5 0.191 0.202 1.890 2.051 jit_kernel_multiply 10 16.1 0.772 1.964 0.772 1.964 potential_pw2rs 119 12.3 0.007 0.008 1.928 1.936 grid_integrate_task_list 119 12.3 1.813 1.908 1.813 1.908 prepare_preconditioner 11 7.9 0.000 0.000 1.905 1.907 make_preconditioner 11 8.9 0.000 0.000 1.904 1.907 pw_transfer 1439 11.6 0.063 0.066 1.851 1.862 calculate_dm_sparse 119 9.5 0.000 0.000 1.841 1.851 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.844 1.845 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.754 1.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.817 1.819 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.760 1.775 multiply_cannon_sync_h2d 18288 15.5 1.427 1.644 1.427 1.644 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.597 1.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.598 1.603 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.563 1.564 cp_fm_redistribute_end 50 14.0 1.152 1.531 1.153 1.532 cp_fm_diag_elpa_base 50 14.0 0.361 1.462 0.375 1.503 ot_diis_step 108 11.5 0.011 0.011 1.486 1.486 mp_sum_l 7207 12.9 1.138 1.479 1.138 1.479 fft3d_ps 1201 14.6 0.508 0.526 1.445 1.458 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.348 1.361 grid_collocate_task_list 119 9.7 1.205 1.300 1.205 1.300 acc_transpose_blocks 18288 15.5 0.075 0.077 1.231 1.255 wfi_extrapolate 11 7.9 0.001 0.001 1.232 1.232 parallel_gemm_fm 81 9.0 0.000 0.000 1.216 1.219 parallel_gemm_fm_cosma 81 10.0 1.216 1.218 1.216 1.218 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.205 1.209 make_images_data 4572 15.5 0.045 0.049 0.883 1.068 make_basis_sm 11 9.8 0.001 0.001 1.007 1.009 qs_energies_init_hamiltonians 11 5.9 0.048 0.069 1.005 1.009 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.766 0.979 mp_alltoall_d11v 2130 13.8 0.763 0.936 0.763 0.936 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.913 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.889 cp_fm_cholesky_invert 11 10.9 0.834 0.838 0.834 0.838 mp_alltoall_z22v 1201 16.6 0.772 0.832 0.772 0.832 acc_transpose_blocks_kernels 18288 16.5 0.207 0.216 0.793 0.808 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.662 0.747 dbcsr_complete_redistribute 329 12.2 0.116 0.151 0.594 0.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.084000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.818182, yerr=1.898238 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 544.821248E+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 0.067 0.233 36.369 36.370 qs_mol_dyn_low 1 2.0 0.003 0.003 35.441 35.449 qs_forces 11 3.9 0.002 0.002 35.380 35.382 qs_energies 11 4.9 0.001 0.003 33.659 33.663 scf_env_do_scf 11 5.9 0.000 0.001 27.568 27.570 scf_env_do_scf_inner_loop 108 6.5 0.008 0.050 24.139 24.139 dbcsr_multiply_generic 2286 12.5 0.099 0.101 18.295 18.402 velocity_verlet 10 3.0 0.001 0.001 17.567 17.569 qs_scf_new_mos 108 7.5 0.001 0.001 16.097 16.155 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.096 16.155 ot_scf_mini 108 9.5 0.002 0.003 15.152 15.203 multiply_cannon 2286 13.5 0.226 0.266 14.601 15.101 multiply_cannon_loop 2286 14.5 0.939 0.965 13.630 14.070 ot_mini 108 10.5 0.001 0.001 9.328 9.394 multiply_cannon_multrec 27432 15.5 2.345 3.011 8.747 9.123 qs_ot_get_derivative 108 11.5 0.001 0.001 7.520 7.573 dbcsr_mm_accdrv_process 47916 15.9 5.277 7.220 6.308 7.519 rebuild_ks_matrix 119 8.3 0.000 0.000 6.388 6.442 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.028 6.388 6.441 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.674 5.721 init_scf_run 11 5.9 0.000 0.003 4.679 4.680 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.679 4.680 sum_up_and_integrate 119 10.3 0.035 0.038 3.676 3.685 integrate_v_rspace 119 11.3 0.002 0.003 3.642 3.650 init_scf_loop 11 6.9 0.000 0.001 3.410 3.411 calculate_first_density_matrix 1 7.0 0.020 0.160 3.325 3.327 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.279 3.315 calculate_rho_elec 119 8.7 0.040 0.046 3.278 3.314 qs_ot_get_p 119 10.4 0.001 0.001 3.206 3.277 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.678 3.152 mp_waitall_1 145218 16.4 2.041 2.677 2.041 2.677 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.106 2.583 apply_single 119 13.6 0.000 0.000 2.105 2.583 prepare_preconditioner 11 7.9 0.000 0.000 2.566 2.573 make_preconditioner 11 8.9 0.000 0.001 2.566 2.573 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.170 2.502 make_m2s 4572 13.5 0.054 0.056 2.184 2.334 make_images 4572 14.5 0.274 0.334 2.076 2.225 calculate_dm_sparse 119 9.5 0.000 0.001 2.068 2.126 rs_pw_transfer 974 11.9 0.009 0.009 1.954 2.072 jit_kernel_multiply 10 16.0 0.972 1.987 0.972 1.987 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.935 1.961 qs_ot_p2m_diag 50 11.0 0.016 0.023 1.929 1.939 grid_integrate_task_list 119 12.3 1.811 1.916 1.811 1.916 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.908 mp_sum_l 7207 12.9 1.290 1.882 1.290 1.882 density_rs2pw 119 9.7 0.004 0.004 1.743 1.844 ot_diis_step 108 11.5 0.012 0.012 1.765 1.765 pw_transfer 1439 11.6 0.063 0.066 1.731 1.763 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.746 1.760 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.641 1.677 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.612 1.612 potential_pw2rs 119 12.3 0.008 0.009 1.477 1.481 acc_transpose_blocks 27432 15.5 0.110 0.112 1.453 1.476 fft_wrap_pw1pw2_140 487 13.2 0.085 0.093 1.353 1.388 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.384 1.385 cp_fm_redistribute_end 50 14.0 0.912 1.363 0.913 1.363 fft3d_ps 1201 14.6 0.535 0.585 1.321 1.349 cp_fm_diag_elpa_base 50 14.0 0.428 1.304 0.448 1.345 grid_collocate_task_list 119 9.7 1.221 1.308 1.221 1.308 wfi_extrapolate 11 7.9 0.001 0.001 1.307 1.307 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.253 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.758 1.241 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.119 1.141 cp_fm_upper_to_full 72 13.5 0.817 1.133 0.817 1.133 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 1.120 1.121 parallel_gemm_fm 81 9.0 0.000 0.000 1.079 1.083 parallel_gemm_fm_cosma 81 10.0 1.079 1.083 1.079 1.083 multiply_cannon_sync_h2d 27432 15.5 1.001 1.066 1.001 1.066 dbcsr_complete_redistribute 329 12.2 0.116 0.145 0.764 1.045 make_basis_sm 11 9.8 0.001 0.001 1.030 1.031 make_images_data 4572 15.5 0.045 0.048 0.844 1.003 mp_alltoall_d11v 2130 13.8 0.746 0.994 0.746 0.994 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.727 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.873 acc_transpose_blocks_kernels 27432 16.5 0.265 0.273 0.829 0.843 mp_sum_d 4123 12.0 0.684 0.842 0.684 0.842 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.558 0.832 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.803 0.810 cp_fm_cholesky_invert 11 10.9 0.750 0.753 0.750 0.753 mp_alltoall_i22 627 13.8 0.435 0.744 0.435 0.744 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.370000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=517.727273, yerr=3.645443 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 614.920192E+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 0.056 0.066 27.573 27.574 qs_mol_dyn_low 1 2.0 0.003 0.003 27.231 27.239 qs_forces 11 3.9 0.001 0.002 27.170 27.171 qs_energies 11 4.9 0.001 0.001 25.496 25.498 scf_env_do_scf 11 5.9 0.000 0.001 20.686 20.686 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.244 18.244 velocity_verlet 10 3.0 0.001 0.001 14.055 14.058 dbcsr_multiply_generic 2286 12.5 0.091 0.093 11.928 12.000 qs_scf_new_mos 108 7.5 0.001 0.001 10.770 10.796 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.769 10.795 ot_scf_mini 108 9.5 0.002 0.002 10.092 10.116 multiply_cannon 2286 13.5 0.230 0.238 9.424 9.880 multiply_cannon_loop 2286 14.5 0.329 0.340 8.497 8.678 multiply_cannon_multrec 9144 15.5 1.570 1.829 5.699 5.898 rebuild_ks_matrix 119 8.3 0.000 0.000 5.728 5.747 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.727 5.747 ot_mini 108 10.5 0.001 0.001 5.574 5.605 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.110 5.129 qs_ot_get_derivative 108 11.5 0.001 0.001 4.291 4.315 dbcsr_mm_accdrv_process 12550 15.8 3.016 3.711 4.029 4.102 sum_up_and_integrate 119 10.3 0.038 0.041 3.500 3.505 integrate_v_rspace 119 11.3 0.002 0.003 3.462 3.467 init_scf_run 11 5.9 0.000 0.002 3.277 3.277 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.276 3.276 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.171 3.178 calculate_rho_elec 119 8.7 0.059 0.061 3.171 3.177 qs_ot_get_p 119 10.4 0.001 0.001 2.739 2.775 init_scf_loop 11 6.9 0.000 0.000 2.422 2.423 mp_waitall_1 121218 16.5 1.613 2.161 1.613 2.161 calculate_first_density_matrix 1 7.0 0.000 0.001 2.159 2.160 make_m2s 4572 13.5 0.034 0.035 1.744 1.938 grid_integrate_task_list 119 12.3 1.850 1.923 1.850 1.923 qs_ot_p2m_diag 50 11.0 0.023 0.023 1.845 1.848 make_images 4572 14.5 0.267 0.301 1.656 1.848 jit_kernel_multiply 10 15.7 0.975 1.742 0.975 1.742 calculate_dm_sparse 119 9.5 0.000 0.000 1.697 1.715 prepare_preconditioner 11 7.9 0.000 0.000 1.710 1.715 make_preconditioner 11 8.9 0.000 0.000 1.710 1.715 rs_pw_transfer 974 11.9 0.008 0.008 1.642 1.713 density_rs2pw 119 9.7 0.003 0.003 1.581 1.668 pw_transfer 1439 11.6 0.063 0.066 1.651 1.659 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.649 1.651 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.599 1.628 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.560 1.569 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.506 1.525 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.519 1.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.350 1.357 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.352 1.353 grid_collocate_task_list 119 9.7 1.272 1.352 1.272 1.352 cp_fm_redistribute_end 50 14.0 0.671 1.333 0.672 1.334 cp_fm_diag_elpa_base 50 14.0 0.615 1.261 0.660 1.320 potential_pw2rs 119 12.3 0.010 0.010 1.302 1.305 qs_energies_init_hamiltonians 11 5.9 0.037 0.049 1.292 1.296 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.272 1.282 ot_diis_step 108 11.5 0.012 0.013 1.271 1.271 fft3d_ps 1201 14.6 0.535 0.549 1.228 1.236 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.155 1.178 apply_single 119 13.6 0.000 0.000 1.155 1.177 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.133 1.146 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.119 1.124 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.806 1.080 wfi_extrapolate 11 7.9 0.001 0.001 1.073 1.073 make_images_data 4572 15.5 0.039 0.042 0.829 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 acc_transpose_blocks 9144 15.5 0.038 0.039 0.887 0.895 cp_fm_cholesky_invert 11 10.9 0.865 0.868 0.865 0.868 mp_alltoall_d11v 2130 13.8 0.782 0.850 0.782 0.850 multiply_cannon_sync_h2d 9144 15.5 0.714 0.802 0.714 0.802 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.674 0.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.739 0.741 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.721 0.730 mp_allgather_i34 2286 14.5 0.260 0.691 0.260 0.691 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.314 0.654 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.647 0.652 mp_alltoall_z22v 1201 16.6 0.567 0.593 0.567 0.593 qs_create_task_list 11 7.9 0.000 0.001 0.547 0.572 generate_qs_task_list 11 8.9 0.194 0.216 0.547 0.571 dbcsr_complete_redistribute 329 12.2 0.170 0.188 0.526 0.554 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.574000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=581.818182, yerr=7.132744 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 753.684480E+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 0.117 0.121 43.272 43.272 qs_mol_dyn_low 1 2.0 0.003 0.003 42.426 42.433 qs_forces 11 3.9 0.002 0.002 42.333 42.336 qs_energies 11 4.9 0.001 0.001 40.340 40.345 scf_env_do_scf 11 5.9 0.001 0.001 33.773 33.773 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.733 25.734 velocity_verlet 10 3.0 0.001 0.001 23.319 23.328 dbcsr_multiply_generic 2286 12.5 0.097 0.099 18.177 18.314 qs_scf_new_mos 108 7.5 0.001 0.001 16.453 16.554 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.452 16.553 ot_scf_mini 108 9.5 0.002 0.002 15.371 15.469 multiply_cannon 2286 13.5 0.303 0.309 14.117 14.957 multiply_cannon_loop 2286 14.5 0.342 0.348 12.857 13.745 ot_mini 108 10.5 0.001 0.001 9.047 9.159 multiply_cannon_multrec 9144 15.5 3.504 4.922 8.942 9.097 init_scf_loop 11 6.9 0.000 0.000 8.008 8.010 rebuild_ks_matrix 119 8.3 0.000 0.000 7.080 7.216 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.080 7.215 qs_ot_get_derivative 108 11.5 0.001 0.001 7.036 7.134 prepare_preconditioner 11 7.9 0.000 0.000 7.004 7.018 make_preconditioner 11 8.9 0.000 0.000 7.004 7.018 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.535 6.897 dbcsr_mm_accdrv_process 12550 15.8 4.011 5.666 5.317 6.723 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.408 6.531 cp_fm_upper_to_full 72 14.2 3.200 4.624 3.200 4.624 init_scf_run 11 5.9 0.000 0.001 4.447 4.447 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.447 4.447 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.911 3.935 calculate_rho_elec 119 8.7 0.118 0.121 3.911 3.934 mp_waitall_1 97218 16.6 2.847 3.922 2.847 3.922 sum_up_and_integrate 119 10.3 0.064 0.066 3.880 3.887 integrate_v_rspace 119 11.3 0.003 0.003 3.816 3.825 qs_ot_get_p 119 10.4 0.001 0.001 3.624 3.756 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.582 2.995 calculate_first_density_matrix 1 7.0 0.001 0.001 2.980 2.984 dbcsr_complete_redistribute 329 12.2 0.289 0.295 2.082 2.933 make_m2s 4572 13.5 0.038 0.038 2.484 2.643 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.718 2.549 make_images 4572 14.5 0.356 0.387 2.364 2.523 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.164 2.452 apply_single 119 13.6 0.000 0.000 2.164 2.452 mp_alltoall_i22 627 13.8 1.537 2.427 1.537 2.427 calculate_dm_sparse 119 9.5 0.000 0.000 2.388 2.406 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.356 2.307 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.464 2.288 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.230 2.232 pw_transfer 1439 11.6 0.066 0.066 2.226 2.231 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.205 2.206 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.131 2.136 grid_integrate_task_list 119 12.3 2.032 2.053 2.032 2.053 jit_kernel_multiply 10 15.6 1.279 2.053 1.279 2.053 ot_diis_step 108 11.5 0.014 0.015 1.988 1.989 density_rs2pw 119 9.7 0.003 0.004 1.968 1.989 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.941 1.942 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.847 1.906 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 1.821 1.823 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.745 1.796 fft3d_ps 1201 14.6 0.568 0.577 1.765 1.768 fft_wrap_pw1pw2_140 487 13.2 0.088 0.091 1.743 1.748 mp_sum_l 7207 12.9 1.048 1.736 1.048 1.736 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.597 1.597 cp_fm_diag_elpa_base 50 14.0 1.434 1.500 1.593 1.593 rs_pw_transfer 974 11.9 0.009 0.009 1.555 1.593 cp_fm_cholesky_invert 11 10.9 1.509 1.512 1.509 1.512 grid_collocate_task_list 119 9.7 1.469 1.483 1.469 1.483 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.457 1.468 hybrid_alltoall_any 4725 16.4 0.086 0.147 1.186 1.448 make_images_data 4572 15.5 0.043 0.047 1.166 1.420 wfi_extrapolate 11 7.9 0.001 0.001 1.402 1.402 potential_pw2rs 119 12.3 0.014 0.014 1.358 1.362 mp_alltoall_d11v 2130 13.8 1.227 1.248 1.227 1.248 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.157 1.173 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.145 1.162 mp_alltoall_z22v 1201 16.6 1.061 1.077 1.061 1.077 multiply_cannon_sync_h2d 9144 15.5 1.045 1.049 1.045 1.049 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.984 1.048 qs_create_task_list 11 7.9 0.042 0.042 0.993 1.006 acc_transpose_blocks 9144 15.5 0.037 0.038 1.000 1.006 generate_qs_task_list 11 8.9 0.377 0.396 0.951 0.964 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.846 0.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.884 0.897 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.272000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=707.000000, yerr=13.584751 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 501.178368E+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 326702. 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 0.047 0.069 84.234 84.235 qs_mol_dyn_low 1 2.0 0.006 0.027 83.831 83.842 qs_forces 11 3.9 0.002 0.003 83.728 83.731 qs_energies 11 4.9 0.004 0.026 80.899 80.916 scf_env_do_scf 11 5.9 0.002 0.013 71.531 71.534 scf_env_do_scf_inner_loop 99 6.5 0.003 0.012 65.731 65.733 dbcsr_multiply_generic 2055 12.4 0.105 0.110 52.054 52.353 qs_scf_new_mos 99 7.5 0.000 0.001 48.141 48.249 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.140 48.249 ot_scf_mini 99 9.5 0.002 0.003 45.744 45.843 multiply_cannon 2055 13.4 0.174 0.182 42.319 43.362 multiply_cannon_loop 2055 14.4 1.540 1.582 41.292 42.348 velocity_verlet 10 3.0 0.001 0.001 41.842 41.842 ot_mini 99 10.5 0.001 0.001 27.769 27.871 qs_ot_get_derivative 99 11.5 0.001 0.001 20.911 20.996 multiply_cannon_multrec 49320 15.4 12.237 13.114 17.193 18.120 rebuild_ks_matrix 110 8.3 0.000 0.001 14.829 15.015 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.014 14.828 15.014 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.073 13.236 mp_waitall_1 241148 16.1 11.939 12.944 11.939 12.944 multiply_cannon_sync_h2d 49320 15.4 10.046 10.764 10.046 10.764 qs_ot_get_p 110 10.4 0.001 0.001 9.490 9.598 multiply_cannon_metrocomm3 49320 15.4 0.081 0.086 6.697 7.872 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.168 7.659 apply_single 110 13.6 0.000 0.000 7.168 7.659 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.942 7.498 sum_up_and_integrate 110 10.3 0.037 0.044 7.300 7.317 integrate_v_rspace 110 11.3 0.003 0.003 7.263 7.288 init_scf_run 11 5.9 0.000 0.002 7.212 7.213 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.212 7.212 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.125 7.190 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.596 6.726 calculate_rho_elec 110 8.6 0.020 0.024 6.595 6.726 ot_diis_step 99 11.5 0.005 0.005 6.567 6.567 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.170 6.223 mp_sum_l 6514 12.8 5.065 6.023 5.065 6.023 init_scf_loop 11 6.9 0.001 0.008 5.768 5.768 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.209 5.210 dbcsr_mm_accdrv_process 87628 16.1 1.993 2.131 4.836 5.178 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.661 4.662 cp_fm_redistribute_end 48 14.0 4.053 4.632 4.057 4.633 cp_fm_diag_elpa_base 48 14.0 0.569 4.558 0.573 4.584 rs_pw_transfer 902 11.9 0.012 0.014 4.230 4.487 make_m2s 4110 13.4 0.061 0.065 4.219 4.309 make_images 4110 14.4 0.178 0.191 4.123 4.216 wfi_extrapolate 11 7.9 0.001 0.001 4.087 4.087 calculate_dm_sparse 110 9.5 0.000 0.001 3.760 3.847 density_rs2pw 110 9.6 0.004 0.004 3.627 3.813 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.586 3.590 multiply_cannon_metrocomm1 49320 15.4 0.058 0.062 2.591 3.564 prepare_preconditioner 11 7.9 0.000 0.000 3.513 3.533 make_preconditioner 11 8.9 0.002 0.014 3.513 3.533 grid_integrate_task_list 110 12.3 3.238 3.411 3.238 3.411 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.277 3.317 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.229 3.282 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.219 3.266 pw_transfer 1331 11.6 0.054 0.061 3.088 3.174 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.001 3.088 calculate_first_density_matrix 1 7.0 0.001 0.003 3.034 3.038 potential_pw2rs 110 12.3 0.006 0.007 2.870 2.908 mp_alltoall_d11v 2046 13.8 2.270 2.867 2.270 2.867 fft3d_ps 1111 14.6 0.745 0.832 2.541 2.625 jit_kernel_multiply 13 15.9 2.561 2.618 2.561 2.618 fft_wrap_pw1pw2_140 451 13.1 0.168 0.189 2.480 2.565 acc_transpose_blocks 49320 15.4 0.216 0.231 2.138 2.243 grid_collocate_task_list 110 9.6 2.087 2.208 2.087 2.208 mp_sum_d 3879 11.9 1.568 2.112 1.568 2.112 make_images_data 4110 15.4 0.043 0.046 1.910 2.062 mp_waitany 14300 13.8 1.882 2.047 1.882 2.047 hybrid_alltoall_any 4261 16.3 0.083 0.482 1.655 1.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.898 1.925 cp_fm_cholesky_invert 11 10.9 1.914 1.918 1.914 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.235000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.818182, yerr=2.124240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 585.748480E+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 2776267. 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 0.026 0.052 71.867 71.871 qs_mol_dyn_low 1 2.0 0.003 0.003 71.067 71.077 qs_forces 11 3.9 0.002 0.002 70.700 70.702 qs_energies 11 4.9 0.005 0.032 67.200 67.204 scf_env_do_scf 11 5.9 0.001 0.002 57.512 57.515 scf_env_do_scf_inner_loop 99 6.5 0.006 0.034 49.394 49.413 dbcsr_multiply_generic 2055 12.4 0.112 0.117 38.090 38.264 velocity_verlet 10 3.0 0.001 0.001 37.036 37.043 qs_scf_new_mos 99 7.5 0.001 0.001 32.971 33.131 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.970 33.131 multiply_cannon 2055 13.4 0.223 0.246 31.206 32.424 ot_scf_mini 99 9.5 0.003 0.003 31.301 31.477 multiply_cannon_loop 2055 14.4 0.925 0.947 29.820 30.620 ot_mini 99 10.5 0.001 0.001 18.320 18.495 multiply_cannon_multrec 24660 15.4 7.698 9.573 13.850 15.499 rebuild_ks_matrix 110 8.3 0.000 0.001 14.095 14.195 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.095 14.194 qs_ot_get_derivative 99 11.5 0.001 0.001 12.544 12.718 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.311 12.394 mp_waitall_1 186928 16.3 8.259 11.024 8.259 11.024 multiply_cannon_sync_h2d 24660 15.4 7.024 8.152 7.024 8.152 init_scf_loop 11 6.9 0.001 0.004 8.079 8.104 multiply_cannon_metrocomm3 24660 15.4 0.069 0.072 5.123 7.886 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.479 7.216 apply_single 110 13.6 0.000 0.001 6.479 7.216 init_scf_run 11 5.9 0.000 0.002 7.100 7.101 scf_env_initial_rho_setup 11 6.9 0.001 0.005 7.100 7.101 sum_up_and_integrate 110 10.3 0.052 0.058 6.930 6.958 integrate_v_rspace 110 11.3 0.002 0.002 6.878 6.908 qs_ot_get_p 110 10.4 0.007 0.037 6.453 6.629 dbcsr_mm_accdrv_process 52282 16.1 4.623 5.230 5.991 6.259 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.065 6.073 calculate_rho_elec 110 8.6 0.039 0.047 6.064 6.072 prepare_preconditioner 11 7.9 0.000 0.000 5.860 5.875 make_preconditioner 11 8.9 0.000 0.002 5.860 5.875 ot_diis_step 99 11.5 0.010 0.010 5.729 5.730 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.426 5.589 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.717 5.425 make_m2s 4110 13.4 0.057 0.060 4.410 4.915 make_images 4110 14.4 0.400 0.443 4.305 4.808 qs_ot_p2m_diag 48 11.0 0.032 0.069 4.278 4.298 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.811 3.813 pw_transfer 1331 11.6 0.065 0.070 3.614 3.784 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.508 3.678 rs_pw_transfer 902 11.9 0.012 0.014 3.373 3.623 wfi_extrapolate 11 7.9 0.001 0.001 3.612 3.612 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.532 3.534 density_rs2pw 110 9.6 0.004 0.004 3.322 3.531 calculate_first_density_matrix 1 7.0 0.004 0.024 3.396 3.399 grid_integrate_task_list 110 12.3 3.136 3.348 3.136 3.348 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.240 3.313 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.159 3.219 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.204 3.207 cp_fm_redistribute_end 48 14.0 2.387 3.174 2.389 3.176 cp_fm_diag_elpa_base 48 14.0 0.750 3.042 0.782 3.130 fft_wrap_pw1pw2_140 451 13.1 0.199 0.217 2.896 3.065 fft3d_ps 1111 14.6 1.072 1.268 2.866 3.028 calculate_dm_sparse 110 9.5 0.005 0.017 2.988 3.021 make_images_data 4110 15.4 0.047 0.050 2.447 2.956 hybrid_alltoall_any 4261 16.3 0.103 0.445 2.149 2.880 potential_pw2rs 110 12.3 0.008 0.009 2.742 2.788 cp_fm_cholesky_invert 11 10.9 2.698 2.705 2.698 2.705 mp_sum_l 6514 12.8 1.892 2.611 1.892 2.611 grid_collocate_task_list 110 9.6 2.121 2.312 2.121 2.312 mp_alltoall_d11v 2046 13.8 1.949 2.199 1.949 2.199 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.996 1.998 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.963 1.983 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.950 1.967 multiply_cannon_metrocomm4 22605 15.4 0.075 0.078 0.793 1.806 acc_transpose_blocks 24660 15.4 0.110 0.112 1.625 1.758 jit_kernel_multiply 9 16.3 1.022 1.746 1.022 1.746 mp_allgather_i34 2055 14.4 0.680 1.724 0.680 1.724 mp_irecv_dv 57340 16.2 0.666 1.685 0.666 1.685 mp_alltoall_z22v 1111 16.6 1.527 1.641 1.527 1.641 mp_sum_d 3879 11.9 1.069 1.569 1.069 1.569 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.553 1.567 make_basis_sm 11 9.8 0.001 0.001 1.528 1.532 parallel_gemm_fm 81 9.0 0.000 0.000 1.523 1.529 parallel_gemm_fm_cosma 81 10.0 1.523 1.529 1.523 1.529 dbcsr_complete_redistribute 325 12.2 0.235 0.294 1.239 1.503 mp_waitany 10164 13.8 1.274 1.494 1.274 1.494 cp_fm_cholesky_decompose 22 10.9 1.480 1.487 1.480 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.369 1.478 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=71.871000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.272727, yerr=6.180481 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 659.034112E+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 0.034 0.075 61.991 61.993 qs_mol_dyn_low 1 2.0 0.003 0.003 61.437 61.448 qs_forces 11 3.9 0.007 0.023 61.369 61.371 qs_energies 11 4.9 0.007 0.041 58.183 58.192 scf_env_do_scf 11 5.9 0.001 0.001 49.072 49.072 scf_env_do_scf_inner_loop 99 6.5 0.008 0.026 40.303 40.304 velocity_verlet 10 3.0 0.001 0.001 32.800 32.802 dbcsr_multiply_generic 2055 12.4 0.107 0.109 28.779 29.009 qs_scf_new_mos 99 7.5 0.001 0.001 25.211 25.297 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.211 25.297 ot_scf_mini 99 9.5 0.003 0.003 23.983 24.088 multiply_cannon 2055 13.4 0.211 0.223 22.223 23.484 multiply_cannon_loop 2055 14.4 0.614 0.631 20.954 21.992 ot_mini 99 10.5 0.001 0.001 13.660 13.767 rebuild_ks_matrix 110 8.3 0.000 0.000 12.487 12.612 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.487 12.612 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.038 11.156 multiply_cannon_multrec 16440 15.4 3.935 5.016 9.651 10.705 mp_waitall_1 146766 16.3 7.543 10.260 7.543 10.260 qs_ot_get_derivative 99 11.5 0.001 0.001 9.238 9.344 init_scf_loop 11 6.9 0.001 0.002 8.727 8.727 prepare_preconditioner 11 7.9 0.000 0.000 6.956 6.977 make_preconditioner 11 8.9 0.000 0.002 6.956 6.977 multiply_cannon_metrocomm3 16440 15.4 0.044 0.046 4.165 6.875 sum_up_and_integrate 110 10.3 0.060 0.061 6.693 6.707 integrate_v_rspace 110 11.3 0.002 0.003 6.632 6.647 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.274 6.636 init_scf_run 11 5.9 0.001 0.004 6.528 6.528 scf_env_initial_rho_setup 11 6.9 0.003 0.018 6.527 6.528 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.849 5.858 calculate_rho_elec 110 8.6 0.058 0.058 5.848 5.857 dbcsr_mm_accdrv_process 34862 16.1 4.374 4.938 5.570 5.739 qs_ot_get_p 110 10.4 0.001 0.001 5.505 5.628 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.953 5.350 apply_single 110 13.6 0.000 0.000 4.953 5.350 make_m2s 4110 13.4 0.050 0.051 4.360 4.751 make_images 4110 14.4 0.392 0.508 4.249 4.641 ot_diis_step 99 11.5 0.010 0.011 4.394 4.394 multiply_cannon_sync_h2d 16440 15.4 3.735 3.994 3.735 3.994 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.143 3.810 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.780 3.784 calculate_first_density_matrix 1 7.0 0.011 0.043 3.427 3.432 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.418 3.419 pw_transfer 1331 11.6 0.064 0.071 3.404 3.416 grid_integrate_task_list 110 12.3 3.184 3.406 3.184 3.406 density_rs2pw 110 9.6 0.004 0.004 3.086 3.330 rs_pw_transfer 902 11.9 0.011 0.012 3.073 3.315 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.298 3.311 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.071 3.073 make_images_data 4110 15.4 0.044 0.048 2.494 3.030 wfi_extrapolate 11 7.9 0.001 0.001 3.007 3.007 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.210 2.951 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.834 2.839 cp_fm_redistribute_end 48 14.0 1.757 2.801 1.759 2.803 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.719 2.776 cp_fm_cholesky_invert 11 10.9 2.761 2.767 2.761 2.767 fft_wrap_pw1pw2_140 451 13.1 0.208 0.212 2.753 2.767 cp_fm_diag_elpa_base 48 14.0 0.975 2.651 1.033 2.761 fft3d_ps 1111 14.6 1.064 1.076 2.614 2.626 calculate_dm_sparse 110 9.5 0.007 0.031 2.538 2.570 potential_pw2rs 110 12.3 0.011 0.011 2.485 2.502 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.419 2.477 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.862 2.348 grid_collocate_task_list 110 9.6 2.180 2.345 2.180 2.345 mp_alltoall_d11v 2046 13.8 1.921 2.261 1.921 2.261 mp_irecv_dv 48980 15.7 0.792 2.223 0.792 2.223 mp_sum_l 6514 12.8 1.580 2.117 1.580 2.117 qs_energies_init_hamiltonians 11 5.9 0.011 0.041 2.096 2.109 dbcsr_complete_redistribute 325 12.2 0.322 0.349 1.536 1.987 cp_fm_upper_to_full 70 13.6 1.385 1.822 1.385 1.822 cp_fm_cholesky_decompose 22 10.9 1.674 1.698 1.674 1.698 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.666 mp_allgather_i34 2055 14.4 0.561 1.628 0.561 1.628 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.493 1.514 jit_kernel_multiply 8 16.2 0.815 1.492 0.815 1.492 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.038 1.481 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.356 1.473 acc_transpose_blocks 16440 15.4 0.071 0.073 1.305 1.422 mp_waitany 17072 13.8 1.191 1.413 1.191 1.413 make_basis_sm 11 9.8 0.008 0.038 1.375 1.378 mp_alltoall_z22v 1111 16.6 1.336 1.351 1.336 1.351 parallel_gemm_fm 81 9.0 0.000 0.000 1.350 1.351 parallel_gemm_fm_cosma 81 10.0 1.350 1.351 1.350 1.351 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.267 1.275 rs_gather_matrices 110 12.3 0.138 0.150 0.904 1.259 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.149 1.240 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.993000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.909091, yerr=8.360671 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 735.543296E+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 0.070 0.186 67.139 67.140 qs_mol_dyn_low 1 2.0 0.003 0.003 66.437 66.454 qs_forces 11 3.9 0.008 0.020 66.307 66.308 qs_energies 11 4.9 0.004 0.021 62.909 62.913 scf_env_do_scf 11 5.9 0.001 0.002 53.912 53.917 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.584 41.585 velocity_verlet 10 3.0 0.001 0.001 37.255 37.259 dbcsr_multiply_generic 2055 12.4 0.113 0.116 30.057 30.306 qs_scf_new_mos 99 7.5 0.001 0.001 26.797 26.900 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.797 26.899 ot_scf_mini 99 9.5 0.007 0.021 25.172 25.278 multiply_cannon 2055 13.4 0.244 0.267 22.669 24.002 multiply_cannon_loop 2055 14.4 0.891 0.910 21.203 21.855 ot_mini 99 10.5 0.001 0.001 14.472 14.609 multiply_cannon_multrec 24660 15.4 4.245 6.898 12.717 13.907 rebuild_ks_matrix 110 8.3 0.000 0.000 12.456 12.566 qs_ks_build_kohn_sham_matrix 110 9.3 0.025 0.079 12.455 12.565 init_scf_loop 11 6.9 0.005 0.023 12.284 12.287 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.070 11.164 qs_ot_get_derivative 99 11.5 0.001 0.001 10.299 10.415 prepare_preconditioner 11 7.9 0.000 0.000 10.229 10.249 make_preconditioner 11 8.9 0.015 0.052 10.229 10.249 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.449 9.905 dbcsr_mm_accdrv_process 52304 16.0 6.790 8.534 8.326 9.281 mp_waitall_1 126806 16.4 5.019 6.933 5.019 6.933 sum_up_and_integrate 110 10.3 0.067 0.070 6.503 6.518 integrate_v_rspace 110 11.3 0.002 0.003 6.436 6.449 init_scf_run 11 5.9 0.000 0.001 6.101 6.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.100 6.101 make_m2s 4110 13.4 0.059 0.062 5.719 6.094 make_images 4110 14.4 0.579 0.702 5.578 5.950 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.822 5.834 calculate_rho_elec 110 8.6 0.077 0.081 5.822 5.834 qs_ot_get_p 110 10.4 0.001 0.001 5.613 5.752 cp_fm_upper_to_full 70 13.8 3.354 4.752 3.354 4.752 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.027 4.142 apply_single 110 13.6 0.000 0.000 4.027 4.142 ot_diis_step 99 11.5 0.011 0.011 4.134 4.135 dbcsr_complete_redistribute 325 12.2 0.416 0.465 2.729 3.874 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.747 3.762 grid_integrate_task_list 110 12.3 3.269 3.428 3.269 3.428 pw_transfer 1331 11.6 0.064 0.072 3.395 3.422 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.344 3.400 make_images_data 4110 15.4 0.047 0.051 2.946 3.392 hybrid_alltoall_any 4261 16.3 0.120 0.455 2.449 3.370 multiply_cannon_sync_h2d 24660 15.4 3.177 3.329 3.177 3.329 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.290 3.319 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.154 3.287 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.256 3.257 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.542 3.237 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.114 3.182 density_rs2pw 110 9.6 0.004 0.004 2.940 3.094 wfi_extrapolate 11 7.9 0.001 0.001 3.050 3.050 calculate_dm_sparse 110 9.5 0.001 0.001 2.974 3.014 calculate_first_density_matrix 1 7.0 0.001 0.003 2.943 2.946 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.919 2.921 mp_alltoall_i22 605 13.7 1.674 2.878 1.674 2.878 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.756 2.865 cp_fm_cholesky_invert 11 10.9 2.853 2.861 2.853 2.861 rs_pw_transfer 902 11.9 0.010 0.011 2.701 2.822 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.754 2.784 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.647 2.651 fft3d_ps 1111 14.6 1.066 1.102 2.593 2.612 cp_fm_redistribute_end 48 14.0 1.312 2.607 1.314 2.609 cp_fm_diag_elpa_base 48 14.0 1.209 2.466 1.289 2.580 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.533 2.568 grid_collocate_task_list 110 9.6 2.223 2.394 2.223 2.394 qs_energies_init_hamiltonians 11 5.9 0.017 0.032 2.391 2.392 jit_kernel_multiply 12 15.7 1.207 2.276 1.207 2.276 potential_pw2rs 110 12.3 0.012 0.013 2.242 2.257 mp_alltoall_d11v 2046 13.8 1.929 2.154 1.929 2.154 cp_fm_cholesky_decompose 22 10.9 1.716 1.766 1.716 1.766 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.715 1.750 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.602 1.699 mp_allgather_i34 2055 14.4 0.567 1.679 0.567 1.679 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.637 1.647 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.588 1.604 acc_transpose_blocks 24660 15.4 0.103 0.105 1.509 1.565 mp_sum_l 6514 12.8 0.991 1.561 0.991 1.561 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.841 1.554 mp_irecv_dv 62702 16.1 0.742 1.472 0.742 1.472 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.297 1.424 mp_sum_d 3877 11.9 0.913 1.397 0.913 1.397 mp_alltoall_z22v 1111 16.6 1.310 1.373 1.310 1.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.140000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.272727, yerr=11.362545 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 816.336896E+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 0.077 0.246 57.871 57.872 qs_mol_dyn_low 1 2.0 0.003 0.003 57.102 57.112 qs_forces 11 3.9 0.072 0.101 57.036 57.037 qs_energies 11 4.9 0.001 0.001 53.342 53.417 scf_env_do_scf 11 5.9 0.000 0.001 44.063 44.063 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 36.072 36.073 velocity_verlet 10 3.0 0.001 0.001 31.197 31.200 dbcsr_multiply_generic 2055 12.4 0.104 0.105 23.245 23.383 qs_scf_new_mos 99 7.5 0.001 0.001 21.580 21.619 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.579 21.619 ot_scf_mini 99 9.5 0.002 0.002 20.331 20.352 multiply_cannon 2055 13.4 0.257 0.281 17.478 18.918 multiply_cannon_loop 2055 14.4 0.321 0.333 16.102 16.429 rebuild_ks_matrix 110 8.3 0.000 0.000 11.673 11.701 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.672 11.701 ot_mini 99 10.5 0.001 0.001 10.509 10.524 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.412 10.436 multiply_cannon_multrec 8220 15.4 3.211 4.531 7.540 8.563 mp_waitall_1 106626 16.5 6.483 8.071 6.483 8.071 init_scf_loop 11 6.9 0.000 0.000 7.938 7.938 qs_ot_get_derivative 99 11.5 0.001 0.001 6.705 6.725 sum_up_and_integrate 110 10.3 0.080 0.081 6.341 6.354 prepare_preconditioner 11 7.9 0.000 0.000 6.299 6.303 make_preconditioner 11 8.9 0.000 0.000 6.299 6.303 integrate_v_rspace 110 11.3 0.003 0.003 6.261 6.273 init_scf_run 11 5.9 0.000 0.002 6.071 6.071 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.071 6.071 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.875 5.961 qs_ot_get_p 110 10.4 0.001 0.001 5.825 5.845 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.811 5.822 calculate_rho_elec 110 8.6 0.115 0.115 5.810 5.822 dbcsr_mm_accdrv_process 17442 15.9 2.795 3.919 4.199 5.131 make_m2s 4110 13.4 0.039 0.040 4.449 4.723 make_images 4110 14.4 0.639 0.693 4.321 4.594 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.953 4.425 qs_ot_p2m_diag 48 11.0 0.081 0.085 4.303 4.307 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.982 3.983 ot_diis_step 99 11.5 0.012 0.012 3.780 3.781 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.688 3.712 apply_single 110 13.6 0.000 0.000 3.688 3.712 grid_integrate_task_list 110 12.3 3.348 3.488 3.348 3.488 pw_transfer 1331 11.6 0.064 0.069 3.439 3.449 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.385 3.387 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.333 3.347 cp_fm_redistribute_end 48 14.0 0.854 3.337 0.860 3.339 cp_fm_diag_elpa_base 48 14.0 2.282 3.126 2.461 3.288 calculate_first_density_matrix 1 7.0 0.001 0.001 3.221 3.222 multiply_cannon_sync_h2d 8220 15.4 2.922 3.087 2.922 3.087 cp_fm_cholesky_invert 11 10.9 3.030 3.034 3.030 3.034 make_images_data 4110 15.4 0.039 0.044 2.527 2.975 density_rs2pw 110 9.6 0.004 0.004 2.822 2.921 hybrid_alltoall_any 4261 16.3 0.200 0.867 2.414 2.891 qs_energies_init_hamiltonians 11 5.9 0.013 0.031 2.742 2.813 fft_wrap_pw1pw2_140 451 13.1 0.211 0.216 2.796 2.808 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.768 2.770 wfi_extrapolate 11 7.9 0.001 0.001 2.759 2.760 fft3d_ps 1111 14.6 1.120 1.153 2.592 2.606 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.514 rs_pw_transfer 902 11.9 0.010 0.010 2.344 2.478 grid_collocate_task_list 110 9.6 2.329 2.476 2.329 2.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.376 2.387 potential_pw2rs 110 12.3 0.015 0.015 2.053 2.057 mp_alltoall_d11v 2046 13.8 1.788 2.029 1.788 2.029 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.984 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.939 1.952 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.877 1.892 cp_fm_cholesky_decompose 22 10.9 1.810 1.823 1.810 1.823 mp_allgather_i34 2055 14.4 0.553 1.682 0.553 1.682 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.563 1.665 dbcsr_complete_redistribute 325 12.2 0.558 0.581 1.534 1.634 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.597 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.446 make_basis_sm 11 9.8 0.027 0.034 1.424 1.426 jit_kernel_multiply 8 15.7 1.101 1.414 1.101 1.414 parallel_gemm_fm 81 9.0 0.000 0.000 1.372 1.374 parallel_gemm_fm_cosma 81 10.0 1.372 1.374 1.372 1.374 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.913 1.369 qs_create_task_list 11 7.9 0.001 0.001 1.240 1.336 generate_qs_task_list 11 8.9 0.386 0.460 1.239 1.335 mp_alltoall_z22v 1111 16.6 1.237 1.260 1.237 1.260 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.206 1.243 mp_waitany 9240 13.8 1.090 1.216 1.090 1.216 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.872000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=770.818182, yerr=9.093636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 1.332474E+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 0.031 0.046 86.196 86.197 qs_mol_dyn_low 1 2.0 0.003 0.003 85.848 85.857 qs_forces 11 3.9 0.006 0.010 85.403 85.405 qs_energies 11 4.9 0.001 0.001 81.392 81.398 scf_env_do_scf 11 5.9 0.000 0.001 71.330 71.330 velocity_verlet 10 3.0 0.001 0.001 55.285 55.338 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.919 42.921 dbcsr_multiply_generic 2055 12.4 0.118 0.120 28.504 28.567 init_scf_loop 11 6.9 0.000 0.000 28.338 28.340 prepare_preconditioner 11 7.9 0.000 0.000 26.380 26.388 make_preconditioner 11 8.9 0.000 0.000 26.380 26.388 qs_scf_new_mos 99 7.5 0.001 0.001 26.271 26.306 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.270 26.305 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.593 25.842 ot_scf_mini 99 9.5 0.002 0.002 24.520 24.539 multiply_cannon 2055 13.4 0.351 0.379 21.373 22.234 multiply_cannon_loop 2055 14.4 0.340 0.343 19.602 19.939 cp_fm_upper_to_full 70 14.2 12.642 18.146 12.642 18.146 ot_mini 99 10.5 0.001 0.001 13.528 13.550 rebuild_ks_matrix 110 8.3 0.000 0.001 13.301 13.329 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.300 13.328 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.103 12.129 dbcsr_complete_redistribute 325 12.2 1.030 1.054 7.447 10.691 multiply_cannon_multrec 8220 15.4 4.414 4.621 9.660 9.795 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.375 9.612 qs_ot_get_derivative 99 11.5 0.001 0.001 9.038 9.058 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.772 8.979 mp_alltoall_i22 605 13.7 5.379 8.629 5.379 8.629 mp_waitall_1 87304 16.6 7.788 8.589 7.788 8.589 sum_up_and_integrate 110 10.3 0.151 0.153 6.773 6.789 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.603 6.641 calculate_rho_elec 110 8.6 0.227 0.227 6.603 6.640 integrate_v_rspace 110 11.3 0.003 0.004 6.622 6.639 make_m2s 4110 13.4 0.043 0.044 5.434 5.928 qs_ot_get_p 110 10.4 0.001 0.001 5.826 5.850 init_scf_run 11 5.9 0.000 0.001 5.774 5.774 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.774 5.774 make_images 4110 14.4 0.888 0.942 5.247 5.742 dbcsr_mm_accdrv_process 11614 15.7 3.282 3.625 5.105 5.346 cp_fm_cholesky_invert 11 10.9 5.334 5.339 5.334 5.339 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.602 5.088 apply_single 110 13.6 0.000 0.000 4.601 5.088 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.589 4.964 ot_diis_step 99 11.5 0.015 0.016 4.473 4.474 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.246 4.254 multiply_cannon_sync_h2d 8220 15.4 3.950 3.955 3.950 3.955 pw_transfer 1331 11.6 0.073 0.073 3.921 3.928 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.805 3.813 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.799 3.800 grid_integrate_task_list 110 12.3 3.655 3.713 3.655 3.713 hybrid_alltoall_any 4261 16.3 0.257 0.557 2.939 3.685 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.672 3.677 make_images_data 4110 15.4 0.042 0.045 2.962 3.632 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.896 3.346 fft_wrap_pw1pw2_140 451 13.1 0.216 0.222 3.276 3.290 wfi_extrapolate 11 7.9 0.001 0.001 3.252 3.252 calculate_dm_sparse 110 9.5 0.001 0.001 3.220 3.242 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.189 3.189 cp_fm_diag_elpa_base 48 14.0 2.635 2.845 3.187 3.187 density_rs2pw 110 9.6 0.004 0.004 3.097 3.117 fft3d_ps 1111 14.6 1.273 1.281 3.014 3.024 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.874 2.876 grid_collocate_task_list 110 9.6 2.626 2.657 2.626 2.657 calculate_first_density_matrix 1 7.0 0.001 0.001 2.414 2.415 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.396 2.407 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.337 2.343 rs_pw_transfer 902 11.9 0.010 0.011 2.243 2.304 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.214 2.292 mp_alltoall_d11v 2046 13.8 2.135 2.184 2.135 2.184 potential_pw2rs 110 12.3 0.021 0.021 2.165 2.168 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.077 2.137 cp_fm_cholesky_decompose 22 10.9 2.090 2.103 2.090 2.103 qs_create_task_list 11 7.9 0.017 0.017 1.930 1.974 generate_qs_task_list 11 8.9 0.750 0.804 1.914 1.957 jit_kernel_multiply 10 15.3 1.624 1.841 1.624 1.841 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.812 1.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.805 1.807 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.770 1.786 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.197000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1198.636364, yerr=53.105851 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 625.377280E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2740610. 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 0.056 0.138 205.493 205.495 qs_mol_dyn_low 1 2.0 0.003 0.004 204.924 204.937 qs_forces 11 3.9 0.003 0.004 204.812 204.813 qs_energies 11 4.9 0.005 0.028 199.175 199.188 scf_env_do_scf 11 5.9 0.001 0.002 182.304 182.308 scf_env_do_scf_inner_loop 117 6.6 0.004 0.016 161.232 161.234 dbcsr_multiply_generic 2507 12.6 0.174 0.179 125.018 125.551 velocity_verlet 10 3.0 0.001 0.001 123.658 123.659 qs_scf_new_mos 117 7.6 0.001 0.001 121.979 122.206 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.979 122.205 ot_scf_mini 117 9.6 0.003 0.003 115.360 115.536 multiply_cannon 2507 13.6 0.238 0.245 101.205 103.047 multiply_cannon_loop 2507 14.6 2.107 2.175 98.951 100.469 ot_mini 117 10.6 0.001 0.001 66.219 66.411 multiply_cannon_multrec 60168 15.6 33.291 35.078 41.662 43.367 qs_ot_get_derivative 117 11.6 0.001 0.001 41.306 41.499 rebuild_ks_matrix 128 8.3 0.001 0.001 33.679 33.855 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.034 33.678 33.855 mp_waitall_1 291448 16.2 29.301 32.480 29.301 32.480 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.259 30.405 multiply_cannon_sync_h2d 60168 15.6 27.390 28.924 27.390 28.924 qs_ot_get_p 128 10.4 0.001 0.002 27.219 27.504 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.384 24.952 apply_single 128 13.6 0.001 0.001 24.383 24.951 ot_diis_step 117 11.6 0.007 0.008 24.506 24.507 init_scf_loop 11 6.9 0.001 0.005 20.990 20.991 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.363 20.439 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.035 19.157 multiply_cannon_metrocomm3 60168 15.6 0.119 0.123 15.849 18.470 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.645 17.647 prepare_preconditioner 11 7.9 0.000 0.000 16.309 16.345 make_preconditioner 11 8.9 0.001 0.002 16.309 16.345 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.538 15.711 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.742 14.748 cp_fm_redistribute_end 83 14.4 11.614 14.664 11.628 14.669 cp_fm_diag_elpa_base 83 14.4 2.997 14.425 3.029 14.547 make_m2s 5014 13.6 0.103 0.111 14.110 14.439 sum_up_and_integrate 128 10.3 0.089 0.107 14.314 14.351 make_images 5014 14.6 0.405 0.425 13.935 14.278 integrate_v_rspace 128 11.3 0.003 0.004 14.224 14.263 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.926 13.021 calculate_rho_elec 128 8.7 0.045 0.063 12.925 13.020 init_scf_run 11 5.9 0.000 0.002 12.587 12.587 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.586 12.587 cp_fm_cholesky_invert 11 10.9 9.359 9.368 9.359 9.368 mp_sum_l 7870 13.0 8.462 9.277 8.462 9.277 wfi_extrapolate 11 7.9 0.001 0.001 9.092 9.092 dbcsr_mm_accdrv_process 124484 16.2 3.204 3.363 7.933 8.512 calculate_dm_sparse 128 9.5 0.001 0.001 8.414 8.490 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.239 8.371 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.881 7.960 make_images_data 5014 15.6 0.076 0.121 6.995 7.938 grid_integrate_task_list 128 12.3 7.077 7.504 7.077 7.504 hybrid_alltoall_any 5200 16.5 0.293 2.268 6.111 7.380 pw_transfer 1547 11.6 0.075 0.106 7.075 7.366 density_rs2pw 128 9.7 0.006 0.007 6.661 7.320 multiply_cannon_metrocomm1 60168 15.6 0.085 0.089 6.016 7.312 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.871 7.132 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.815 6.823 rs_pw_transfer 1046 11.9 0.017 0.018 5.988 6.599 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.953 6.049 fft_wrap_pw1pw2_140 523 13.2 0.444 0.512 5.847 6.041 fft3d_ps 1291 14.7 2.091 2.579 5.664 5.862 mp_alltoall_d11v 2415 14.1 4.391 5.740 4.391 5.740 grid_collocate_task_list 128 9.7 4.730 5.131 4.730 5.131 potential_pw2rs 128 12.3 0.009 0.011 4.764 4.819 cp_fm_cholesky_decompose 22 10.9 4.755 4.770 4.755 4.770 mp_sum_d 4455 12.2 3.653 4.151 3.653 4.151 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.495000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.818182, yerr=6.235065 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/16/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 2.183246E+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 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 826.986496E+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 1463834560144 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 6533303. 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 0.029 0.063 193.011 193.013 qs_mol_dyn_low 1 2.0 0.005 0.014 192.422 192.435 qs_forces 11 3.9 0.005 0.017 191.431 191.433 qs_energies 11 4.9 0.004 0.021 184.687 184.699 scf_env_do_scf 11 5.9 0.001 0.002 167.930 167.940 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.688 134.690 velocity_verlet 10 3.0 0.001 0.001 121.135 121.155 dbcsr_multiply_generic 2507 12.6 0.184 0.189 97.765 98.944 qs_scf_new_mos 117 7.6 0.001 0.001 95.698 96.268 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.698 96.268 ot_scf_mini 117 9.6 0.004 0.004 90.894 91.535 multiply_cannon 2507 13.6 0.484 0.534 77.577 81.671 multiply_cannon_loop 2507 14.6 1.246 1.284 74.179 76.634 ot_mini 117 10.6 0.001 0.001 49.890 50.455 mp_waitall_1 226760 16.4 25.297 38.972 25.297 38.972 multiply_cannon_multrec 30084 15.6 21.960 26.345 31.617 36.514 rebuild_ks_matrix 128 8.3 0.001 0.001 32.918 33.446 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.918 33.446 init_scf_loop 11 6.9 0.001 0.003 33.147 33.148 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.670 30.141 multiply_cannon_metrocomm3 30084 15.6 0.093 0.097 15.818 29.450 prepare_preconditioner 11 7.9 0.000 0.000 28.740 28.826 make_preconditioner 11 8.9 0.000 0.002 28.740 28.826 qs_ot_get_derivative 117 11.6 0.001 0.002 27.902 28.536 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.445 27.991 qs_ot_get_p 128 10.4 0.001 0.001 22.719 23.454 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.017 23.104 apply_single 128 13.6 0.001 0.001 22.016 23.104 multiply_cannon_sync_h2d 30084 15.6 19.253 21.935 19.253 21.935 ot_diis_step 117 11.6 0.014 0.015 21.813 21.816 qs_ot_p2m_diag 83 11.4 0.188 0.216 17.834 17.868 cp_fm_cholesky_invert 11 10.9 16.845 16.857 16.845 16.857 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.688 16.690 make_m2s 5014 13.6 0.093 0.098 14.454 16.047 make_images 5014 14.6 1.162 1.349 14.251 15.844 sum_up_and_integrate 128 10.3 0.115 0.134 14.454 14.486 integrate_v_rspace 128 11.3 0.003 0.004 14.339 14.379 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.560 13.568 cp_fm_redistribute_end 83 14.4 7.881 13.487 7.892 13.488 cp_fm_diag_elpa_base 83 14.4 5.355 13.028 5.577 13.396 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.215 13.248 calculate_rho_elec 128 8.7 0.087 0.105 13.214 13.247 init_scf_run 11 5.9 0.000 0.002 11.765 11.767 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.765 11.767 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.995 11.483 make_images_data 5014 15.6 0.067 0.073 8.798 10.844 multiply_cannon_metrocomm4 27577 15.6 0.099 0.114 3.780 10.760 mp_irecv_dv 69486 16.3 3.582 10.371 3.582 10.371 hybrid_alltoall_any 5200 16.5 0.345 1.521 7.460 10.161 dbcsr_mm_accdrv_process 62242 16.2 4.501 5.467 9.115 9.775 wfi_extrapolate 11 7.9 0.001 0.001 8.409 8.409 pw_transfer 1547 11.6 0.085 0.097 7.901 7.947 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.677 7.719 grid_integrate_task_list 128 12.3 7.190 7.583 7.190 7.583 density_rs2pw 128 9.7 0.006 0.006 6.957 7.355 cp_fm_cholesky_decompose 22 10.9 7.019 7.104 7.019 7.104 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.219 6.966 fft_wrap_pw1pw2_140 523 13.2 0.468 0.515 6.666 6.714 calculate_dm_sparse 128 9.5 0.001 0.002 6.521 6.659 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.331 6.342 rs_pw_transfer 1046 11.9 0.014 0.016 5.840 6.332 fft3d_ps 1291 14.7 2.777 2.921 6.048 6.096 mp_sum_l 7870 13.0 3.964 6.017 3.964 6.017 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.424 5.561 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.317 5.380 grid_collocate_task_list 128 9.7 4.910 5.342 4.910 5.342 mp_allgather_i34 2507 14.6 1.967 5.120 1.967 5.120 potential_pw2rs 128 12.3 0.015 0.017 4.899 4.923 mp_alltoall_d11v 2415 14.1 4.191 4.870 4.191 4.870 mp_sum_d 4455 12.2 2.694 4.124 2.694 4.124 dbcsr_complete_redistribute 395 12.7 0.769 0.856 3.169 4.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.013000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 934.002688E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 57139. MP_Allreduce 11261 985. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.238 174.165 174.167 qs_mol_dyn_low 1 2.0 0.003 0.003 173.482 173.496 qs_forces 11 3.9 0.003 0.003 173.381 173.385 qs_energies 11 4.9 0.002 0.006 166.852 166.863 scf_env_do_scf 11 5.9 0.001 0.002 151.330 151.331 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.314 116.317 velocity_verlet 10 3.0 0.001 0.001 110.838 110.840 dbcsr_multiply_generic 2485 12.5 0.177 0.179 80.500 81.563 qs_scf_new_mos 116 7.6 0.001 0.001 80.263 80.585 qs_scf_loop_do_ot 116 8.6 0.001 0.002 80.262 80.584 ot_scf_mini 116 9.6 0.004 0.004 76.146 76.560 multiply_cannon 2485 13.5 0.497 0.525 61.042 65.543 multiply_cannon_loop 2485 14.5 0.845 0.874 58.054 60.647 ot_mini 116 10.6 0.001 0.001 42.105 42.506 init_scf_loop 11 6.9 0.001 0.004 34.916 34.919 mp_waitall_1 176908 16.5 24.780 34.082 24.780 34.082 prepare_preconditioner 11 7.9 0.000 0.000 30.928 30.969 make_preconditioner 11 8.9 0.000 0.002 30.928 30.969 rebuild_ks_matrix 127 8.3 0.001 0.001 29.713 30.194 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.712 30.193 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.560 29.975 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.746 27.189 multiply_cannon_multrec 19880 15.5 13.640 16.688 22.211 25.414 multiply_cannon_metrocomm3 19880 15.5 0.060 0.064 14.829 24.477 qs_ot_get_derivative 116 11.6 0.001 0.002 22.663 23.069 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.565 20.551 apply_single 127 13.6 0.001 0.001 19.564 20.550 qs_ot_get_p 127 10.4 0.001 0.002 19.752 20.213 ot_diis_step 116 11.6 0.017 0.018 19.340 19.340 multiply_cannon_sync_h2d 19880 15.5 14.129 15.671 14.129 15.671 qs_ot_p2m_diag 82 11.4 0.263 0.270 15.457 15.473 make_m2s 4970 13.5 0.082 0.089 14.417 15.275 make_images 4970 14.5 1.166 1.270 14.191 15.046 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.455 14.456 cp_fm_cholesky_invert 11 10.9 14.417 14.426 14.417 14.426 sum_up_and_integrate 127 10.3 0.131 0.142 14.089 14.117 integrate_v_rspace 127 11.3 0.003 0.004 13.958 13.990 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.906 12.929 calculate_rho_elec 127 8.7 0.130 0.146 12.906 12.929 cp_fm_diag_elpa 82 13.4 0.001 0.001 11.383 11.384 cp_fm_redistribute_end 82 14.4 4.275 11.328 4.290 11.331 cp_fm_diag_elpa_base 82 14.4 6.610 10.751 7.018 11.247 init_scf_run 11 5.9 0.000 0.001 10.484 10.485 scf_env_initial_rho_setup 11 6.9 0.003 0.019 10.483 10.485 make_images_data 4970 15.5 0.061 0.068 8.884 10.042 hybrid_alltoall_any 5155 16.4 0.428 1.951 7.711 9.356 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.855 9.151 multiply_cannon_metrocomm4 17395 15.5 0.059 0.069 3.334 8.943 mp_irecv_dv 49801 16.2 3.215 8.696 3.215 8.696 dbcsr_mm_accdrv_process 41158 16.2 4.414 5.222 8.024 8.183 pw_transfer 1535 11.6 0.083 0.103 7.765 7.872 grid_integrate_task_list 127 12.3 7.222 7.705 7.222 7.705 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.544 7.658 cp_fm_cholesky_decompose 22 10.9 7.419 7.462 7.419 7.462 cp_fm_upper_to_full 104 14.5 5.691 7.408 5.691 7.408 wfi_extrapolate 11 7.9 0.001 0.001 7.268 7.268 fft_wrap_pw1pw2_140 519 13.2 0.475 0.523 6.625 6.747 density_rs2pw 127 9.7 0.005 0.006 6.414 6.707 dbcsr_complete_redistribute 393 12.7 1.172 1.198 4.597 6.367 fft3d_ps 1281 14.7 2.676 2.882 5.826 5.892 calculate_dm_sparse 127 9.5 0.001 0.003 5.762 5.862 grid_collocate_task_list 127 9.7 5.029 5.522 5.029 5.522 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.399 5.406 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.574 5.272 rs_pw_transfer 1038 11.9 0.013 0.014 4.983 5.241 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.429 5.195 mp_alltoall_d11v 2401 14.1 4.237 4.792 4.237 4.792 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.497 4.621 potential_pw2rs 127 12.3 0.020 0.022 4.527 4.551 mp_sum_l 7804 13.0 3.282 4.448 3.282 4.448 mp_allgather_i34 2485 14.5 1.523 4.081 1.523 4.081 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.348 4.038 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.976 4.001 mp_alltoall_i22 712 14.1 2.021 3.918 2.021 3.918 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.761 3.762 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.475 3.512 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=174.167000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=886.818182, yerr=7.732084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/18/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 4.353788E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.157861E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+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 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.121 188.263 188.263 qs_mol_dyn_low 1 2.0 0.003 0.003 187.268 187.726 qs_forces 11 3.9 0.003 0.003 186.630 186.638 qs_energies 11 4.9 0.002 0.007 179.633 179.648 scf_env_do_scf 11 5.9 0.001 0.002 162.688 162.703 velocity_verlet 10 3.0 0.001 0.001 124.330 124.354 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 115.809 115.810 qs_scf_new_mos 117 7.6 0.001 0.001 80.128 80.401 qs_scf_loop_do_ot 117 8.6 0.001 0.002 80.127 80.400 dbcsr_multiply_generic 2507 12.6 0.187 0.193 79.187 80.078 ot_scf_mini 117 9.6 0.003 0.004 75.633 75.901 multiply_cannon 2507 13.6 0.557 0.583 54.805 57.954 multiply_cannon_loop 2507 14.6 1.184 1.210 51.000 53.084 init_scf_loop 11 6.9 0.001 0.003 46.759 46.760 prepare_preconditioner 11 7.9 0.000 0.000 42.714 42.747 make_preconditioner 11 8.9 0.000 0.001 42.714 42.747 ot_mini 117 10.6 0.001 0.001 42.096 42.346 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.273 41.381 multiply_cannon_multrec 30084 15.6 14.247 20.087 26.275 31.543 rebuild_ks_matrix 128 8.3 0.001 0.001 28.807 29.056 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 28.807 29.056 mp_waitall_1 153770 16.5 17.117 27.444 17.117 27.444 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.046 26.267 qs_ot_get_derivative 117 11.6 0.001 0.002 22.487 22.756 make_m2s 5014 13.6 0.096 0.101 19.965 21.235 make_images 5014 14.6 1.951 2.313 19.661 20.930 qs_ot_get_p 128 10.4 0.001 0.001 19.832 20.102 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.965 19.606 apply_single 128 13.6 0.001 0.001 18.965 19.606 ot_diis_step 117 11.6 0.018 0.020 19.488 19.490 cp_fm_upper_to_full 105 14.7 11.304 16.771 11.304 16.771 cp_fm_cholesky_invert 11 10.9 16.269 16.279 16.269 16.279 qs_ot_p2m_diag 83 11.4 0.344 0.391 15.726 15.779 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.242 15.300 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.449 14.450 sum_up_and_integrate 128 10.3 0.141 0.153 13.984 14.011 integrate_v_rspace 128 11.3 0.003 0.004 13.843 13.874 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.011 13.054 calculate_rho_elec 128 8.7 0.176 0.190 13.010 13.053 dbcsr_complete_redistribute 395 12.7 1.511 1.656 9.006 12.902 make_images_data 5014 15.6 0.065 0.069 10.685 12.629 multiply_cannon_sync_h2d 30084 15.6 11.772 12.625 11.772 12.625 dbcsr_mm_accdrv_process 62264 16.2 7.417 8.322 11.605 12.098 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.658 11.520 hybrid_alltoall_any 5200 16.5 0.526 2.224 9.567 11.397 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.380 11.382 cp_fm_redistribute_end 83 14.4 1.944 11.309 1.959 11.315 cp_fm_diag_elpa_base 83 14.4 8.729 10.704 9.323 11.197 init_scf_run 11 5.9 0.000 0.001 11.017 11.019 scf_env_initial_rho_setup 11 6.9 0.004 0.005 11.017 11.019 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.418 10.189 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.284 9.479 mp_alltoall_i22 716 14.1 5.630 9.463 5.630 9.463 grid_integrate_task_list 128 12.3 7.515 7.958 7.515 7.958 pw_transfer 1547 11.6 0.085 0.100 7.709 7.789 cp_fm_cholesky_decompose 22 10.9 7.645 7.725 7.645 7.725 wfi_extrapolate 11 7.9 0.001 0.001 7.577 7.577 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.485 7.569 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.764 6.926 fft_wrap_pw1pw2_140 523 13.2 0.481 0.494 6.612 6.705 mp_irecv_dv 76098 16.2 2.619 6.653 2.619 6.653 density_rs2pw 128 9.7 0.005 0.006 6.142 6.644 calculate_dm_sparse 128 9.5 0.001 0.001 6.287 6.426 fft3d_ps 1291 14.7 2.793 2.892 5.719 5.786 mp_alltoall_d11v 2415 14.1 4.712 5.647 4.712 5.647 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.537 5.611 grid_collocate_task_list 128 9.7 5.206 5.543 5.206 5.543 rs_pw_transfer 1046 11.9 0.013 0.015 4.604 5.050 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.486 4.568 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.352 4.441 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.423 4.425 potential_pw2rs 128 12.3 0.022 0.023 4.250 4.267 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.164 4.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=188.263000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1089.545455, yerr=19.434569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.495007E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57602. MP_Allreduce 11109 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.038 171.554 171.554 qs_mol_dyn_low 1 2.0 0.003 0.003 171.109 171.123 qs_forces 11 3.9 0.010 0.011 171.010 171.012 qs_energies 11 4.9 0.001 0.001 163.762 163.770 scf_env_do_scf 11 5.9 0.001 0.001 146.094 146.103 velocity_verlet 10 3.0 0.001 0.001 111.971 111.975 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 110.640 110.641 dbcsr_multiply_generic 2529 12.6 0.182 0.187 75.387 75.803 qs_scf_new_mos 118 7.6 0.001 0.001 74.841 74.930 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.840 74.930 ot_scf_mini 118 9.6 0.004 0.004 70.386 70.455 multiply_cannon 2529 13.6 0.603 0.649 55.577 60.027 multiply_cannon_loop 2529 14.6 0.449 0.460 50.984 51.764 ot_mini 118 10.6 0.001 0.001 40.142 40.203 init_scf_loop 11 6.9 0.000 0.000 35.302 35.303 mp_waitall_1 130746 16.6 26.296 33.235 26.296 33.235 prepare_preconditioner 11 7.9 0.000 0.000 31.413 31.442 make_preconditioner 11 8.9 0.000 0.000 31.413 31.442 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.304 29.586 rebuild_ks_matrix 129 8.3 0.001 0.001 28.721 28.821 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 28.720 28.820 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.267 26.360 multiply_cannon_multrec 10116 15.6 10.605 14.760 18.518 21.305 qs_ot_get_derivative 118 11.6 0.002 0.002 20.167 20.242 ot_diis_step 118 11.6 0.020 0.021 19.908 19.908 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.439 19.645 apply_single 129 13.6 0.001 0.001 19.439 19.645 multiply_cannon_metrocomm3 10116 15.6 0.023 0.025 12.349 19.308 make_m2s 5058 13.6 0.068 0.073 16.222 18.578 make_images 5058 14.6 2.317 2.682 15.916 18.270 cp_fm_cholesky_invert 11 10.9 17.923 17.930 17.923 17.930 qs_ot_get_p 129 10.4 0.001 0.001 17.245 17.347 sum_up_and_integrate 129 10.3 0.182 0.192 14.050 14.095 integrate_v_rspace 129 11.3 0.004 0.004 13.868 13.922 qs_ot_p2m_diag 84 11.4 0.502 0.508 13.517 13.535 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.172 13.210 calculate_rho_elec 129 8.7 0.261 0.271 13.171 13.210 make_images_data 5058 15.6 0.055 0.062 9.929 12.392 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.371 12.372 hybrid_alltoall_any 5245 16.5 0.851 3.863 9.735 12.224 multiply_cannon_sync_h2d 10116 15.6 11.766 12.075 11.766 12.075 init_scf_run 11 5.9 0.000 0.002 10.903 10.903 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.903 10.903 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.370 9.372 cp_fm_diag_elpa_base 84 14.4 9.108 9.187 9.353 9.353 dbcsr_mm_accdrv_process 20934 16.1 2.838 3.513 7.539 8.307 grid_integrate_task_list 129 12.3 7.830 8.209 7.830 8.209 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.003 8.056 cp_fm_cholesky_decompose 22 10.9 7.861 7.958 7.861 7.958 pw_transfer 1559 11.6 0.083 0.089 7.770 7.783 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.548 7.559 multiply_cannon_metrocomm1 10116 15.6 0.028 0.029 4.632 7.429 mp_allgather_i34 2529 14.6 2.778 7.327 2.778 7.327 wfi_extrapolate 11 7.9 0.001 0.001 7.261 7.261 fft_wrap_pw1pw2_140 527 13.2 0.507 0.529 6.667 6.683 calculate_dm_sparse 129 9.5 0.001 0.001 6.447 6.530 density_rs2pw 129 9.7 0.005 0.006 5.937 6.337 grid_collocate_task_list 129 9.7 5.528 5.929 5.528 5.929 fft3d_ps 1301 14.7 2.747 2.818 5.692 5.720 dbcsr_complete_redistribute 397 12.7 2.155 2.247 5.252 5.619 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.316 5.331 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.236 5.236 mp_alltoall_d11v 2429 14.1 4.673 5.220 4.673 5.220 rs_pw_transfer 1054 12.0 0.013 0.013 4.030 4.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.193 4.229 potential_pw2rs 129 12.3 0.026 0.027 4.016 4.027 multiply_cannon_metrocomm4 7587 15.6 0.023 0.026 1.874 3.868 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.825 mp_irecv_dv 29102 15.9 1.838 3.804 1.838 3.804 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.425 3.748 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.709 3.744 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.543 3.617 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.486 3.506 calculate_first_density_matrix 1 7.0 0.001 0.001 3.449 3.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.554000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1398.090909, yerr=41.469495 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_performance_tests/20/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 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.133280E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59908. MP_Allreduce 10870 1518. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.063 290.228 290.228 qs_mol_dyn_low 1 2.0 0.003 0.003 289.560 289.573 qs_forces 11 3.9 0.011 0.021 289.058 289.060 qs_energies 11 4.9 0.001 0.001 280.384 280.408 scf_env_do_scf 11 5.9 0.001 0.001 258.225 258.235 velocity_verlet 10 3.0 0.001 0.001 208.643 208.702 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.555 133.558 init_scf_loop 11 6.9 0.000 0.000 124.405 124.409 prepare_preconditioner 11 7.9 0.000 0.000 119.631 119.662 make_preconditioner 11 8.9 0.000 0.000 119.631 119.662 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.692 116.755 qs_scf_new_mos 116 7.6 0.001 0.001 91.447 91.577 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.446 91.577 ot_scf_mini 116 9.6 0.004 0.004 86.683 86.717 dbcsr_multiply_generic 2485 12.5 0.209 0.217 82.922 83.443 cp_fm_upper_to_full 104 14.8 52.763 75.690 52.763 75.690 multiply_cannon 2485 13.5 0.701 0.734 58.417 59.052 multiply_cannon_loop 2485 14.5 0.469 0.479 54.808 56.085 ot_mini 116 10.6 0.001 0.001 44.256 44.290 dbcsr_complete_redistribute 393 12.7 3.989 4.029 29.667 42.605 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.232 39.161 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.893 36.718 mp_alltoall_i22 712 14.1 21.659 34.424 21.659 34.424 rebuild_ks_matrix 127 8.3 0.001 0.001 33.665 33.704 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.664 33.704 cp_fm_cholesky_invert 11 10.9 33.588 33.594 33.588 33.594 mp_waitall_1 104546 16.7 28.208 32.172 28.208 32.172 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.361 31.402 qs_ot_get_p 127 10.4 0.001 0.001 27.063 27.073 qs_ot_get_derivative 116 11.6 0.002 0.002 24.381 24.416 qs_ot_p2m_diag 82 11.4 0.868 0.873 22.984 23.007 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.282 21.284 make_m2s 4970 13.5 0.076 0.079 19.999 20.925 make_images 4970 14.5 3.747 3.882 19.527 20.456 ot_diis_step 116 11.6 0.022 0.022 19.851 19.851 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 18.519 19.724 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.087 19.275 apply_single 127 13.6 0.001 0.001 19.087 19.274 multiply_cannon_multrec 9940 15.5 10.588 12.434 18.192 18.267 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.054 18.055 cp_fm_diag_elpa_base 82 14.4 13.665 15.292 18.049 18.050 sum_up_and_integrate 127 10.3 0.317 0.320 15.898 15.984 integrate_v_rspace 127 11.3 0.004 0.004 15.580 15.666 multiply_cannon_sync_h2d 9940 15.5 15.538 15.552 15.538 15.552 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.089 15.118 calculate_rho_elec 127 8.7 0.479 0.480 15.089 15.117 hybrid_alltoall_any 5155 16.4 1.291 3.018 10.807 12.716 make_images_data 4970 15.5 0.061 0.064 10.658 12.471 init_scf_run 11 5.9 0.000 0.001 12.120 12.121 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.120 12.121 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.369 9.400 dbcsr_mm_accdrv_process 20590 16.0 3.744 5.704 7.371 9.174 cp_fm_cholesky_decompose 22 10.9 9.081 9.144 9.081 9.144 wfi_extrapolate 11 7.9 0.001 0.001 8.703 8.703 grid_integrate_task_list 127 12.3 8.481 8.655 8.481 8.655 pw_transfer 1535 11.6 0.089 0.090 8.563 8.569 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.331 8.338 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 8.035 8.045 mp_alltoall_d11v 2401 14.1 7.240 7.389 7.240 7.389 fft_wrap_pw1pw2_140 519 13.2 0.535 0.537 7.344 7.361 calculate_dm_sparse 127 9.5 0.001 0.001 6.813 6.928 fft3d_ps 1281 14.7 2.734 2.750 6.391 6.406 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.337 6.393 grid_collocate_task_list 127 9.7 6.289 6.323 6.289 6.323 density_rs2pw 127 9.7 0.005 0.005 6.181 6.219 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.115 6.181 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.228000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2776.272727, yerr=192.603734 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 1.261478E+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 39652582. 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 0.044 0.069 86.182 86.184 qs_energies 1 2.0 0.000 0.001 85.530 85.547 ls_scf 1 3.0 0.002 0.009 84.516 84.533 dbcsr_multiply_generic 111 6.7 0.017 0.021 72.906 73.092 multiply_cannon 111 7.7 0.019 0.025 56.162 57.353 multiply_cannon_loop 111 8.7 0.204 0.218 52.590 53.982 ls_scf_main 1 4.0 0.000 0.002 52.379 52.384 density_matrix_trs4 2 5.0 0.003 0.005 46.894 47.003 ls_scf_init_scf 1 4.0 0.000 0.002 29.041 29.054 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.613 27.685 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.465 25.499 mp_waitall_1 11316 10.9 22.751 25.437 22.751 25.437 multiply_cannon_multrec 2664 9.7 8.203 8.949 15.528 17.257 multiply_cannon_sync_h2d 2664 9.7 13.522 15.719 13.522 15.719 make_m2s 222 7.7 0.010 0.013 13.175 13.736 make_images 222 8.7 0.104 0.125 13.152 13.716 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.822 12.269 make_images_data 222 9.7 0.005 0.006 7.690 8.240 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.549 8.141 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.629 6.942 7.914 hybrid_alltoall_any 227 10.6 0.219 1.848 6.583 7.909 dbcsr_mm_accdrv_process_sort 4760 11.4 6.231 7.102 6.231 7.102 calculate_norms 4752 9.8 5.522 6.154 5.522 6.154 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.054 5.194 mp_sum_l 807 5.4 3.035 4.202 3.035 4.202 make_images_sizes 222 9.7 0.000 0.000 0.699 3.551 mp_alltoall_i44 222 10.7 0.699 3.551 0.699 3.551 arnoldi_extremal 4 6.8 0.000 0.002 3.335 3.383 arnoldi_normal_ev 4 7.8 0.001 0.007 3.334 3.383 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.242 3.306 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.072 3.253 mp_irecv_dv 6231 10.9 2.053 3.223 2.053 3.223 build_subspace 16 8.4 0.009 0.013 3.164 3.166 ls_scf_post 1 4.0 0.004 0.025 3.093 3.106 ls_scf_store_result 1 5.0 0.000 0.000 2.848 2.891 dbcsr_special_finalize 555 9.7 0.006 0.007 2.416 2.807 dbcsr_merge_single_wm 555 10.7 0.489 0.629 2.407 2.798 make_images_pack 222 9.7 2.216 2.643 2.218 2.645 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.354 2.592 dbcsr_sort_data 658 11.4 2.176 2.493 2.176 2.493 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.461 2.069 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.224 2.296 buffer_matrices_ensure_size 222 8.7 1.842 2.207 1.842 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.880 1.894 rebuild_ks_matrix 3 7.3 0.000 0.000 1.870 1.884 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.028 1.870 1.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.184000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 2.098946E+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 0.040 0.071 91.613 91.619 qs_energies 1 2.0 0.000 0.000 90.918 90.924 ls_scf 1 3.0 0.000 0.001 89.585 89.591 dbcsr_multiply_generic 111 6.7 0.016 0.020 75.448 75.797 multiply_cannon 111 7.7 0.037 0.065 53.233 56.993 ls_scf_main 1 4.0 0.000 0.001 54.993 54.999 multiply_cannon_loop 111 8.7 0.118 0.125 49.963 52.851 density_matrix_trs4 2 5.0 0.002 0.003 49.419 49.580 ls_scf_init_scf 1 4.0 0.000 0.002 31.082 31.084 mp_waitall_1 9246 10.9 21.181 30.310 21.181 30.310 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.871 29.995 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.367 27.382 multiply_cannon_multrec 1332 9.7 13.204 16.886 22.462 27.274 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.650 21.031 make_m2s 222 7.7 0.008 0.009 15.574 16.294 make_images 222 8.7 1.587 1.943 15.543 16.264 dbcsr_mm_accdrv_process 4041 10.4 0.278 0.455 8.858 10.495 make_images_data 222 9.7 0.004 0.005 8.965 10.066 dbcsr_mm_accdrv_process_sort 4041 11.4 8.449 10.040 8.449 10.040 hybrid_alltoall_any 227 10.6 0.523 2.429 8.415 9.351 mp_sum_l 807 5.4 5.355 8.043 5.355 8.043 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.250 7.607 mp_irecv_dv 3311 11.0 3.229 7.553 3.229 7.553 calculate_norms 2376 9.8 6.025 6.889 6.025 6.889 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.162 6.589 multiply_cannon_sync_h2d 1332 9.7 4.807 5.887 4.807 5.887 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.046 5.246 arnoldi_extremal 4 6.8 0.000 0.000 4.642 4.672 arnoldi_normal_ev 4 7.8 0.001 0.004 4.642 4.672 build_subspace 16 8.4 0.014 0.021 4.380 4.384 ls_scf_post 1 4.0 0.000 0.001 3.510 3.516 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.119 3.352 ls_scf_store_result 1 5.0 0.000 0.000 3.200 3.334 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.215 2.735 3.217 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.194 2.715 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.501 2.582 mp_allgather_i34 111 8.7 0.957 2.482 0.957 2.482 make_images_pack 222 9.7 2.075 2.465 2.078 2.467 dbcsr_sort_data 436 11.2 1.826 2.071 1.826 2.071 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.848 1.850 dbcsr_data_new 4174 10.1 1.613 1.849 1.613 1.849 rebuild_ks_matrix 3 7.3 0.000 0.000 1.835 1.837 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.835 1.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.619000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1743.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 2.718560E+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 0.045 0.062 94.460 94.462 qs_energies 1 2.0 0.000 0.000 93.674 93.677 ls_scf 1 3.0 0.000 0.001 92.250 92.256 dbcsr_multiply_generic 111 6.7 0.016 0.018 76.743 76.994 ls_scf_main 1 4.0 0.000 0.001 57.566 57.570 multiply_cannon 111 7.7 0.041 0.066 53.203 57.183 multiply_cannon_loop 111 8.7 0.101 0.107 49.648 53.148 density_matrix_trs4 2 5.0 0.002 0.003 51.673 51.837 mp_waitall_1 7374 11.0 24.123 33.494 24.123 33.494 ls_scf_init_scf 1 4.0 0.000 0.001 31.041 31.044 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.894 29.976 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.474 27.487 multiply_cannon_multrec 888 9.7 12.696 15.152 21.264 24.239 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.289 22.718 make_m2s 222 7.7 0.007 0.008 17.005 18.213 make_images 222 8.7 1.982 2.321 16.966 18.176 make_images_data 222 9.7 0.004 0.005 9.835 10.844 hybrid_alltoall_any 227 10.6 0.623 2.868 9.501 10.711 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.418 8.070 9.283 dbcsr_mm_accdrv_process_sort 3754 11.4 7.698 8.864 7.698 8.864 mp_sum_l 807 5.4 4.975 8.189 4.975 8.189 multiply_cannon_sync_h2d 888 9.7 6.035 7.672 6.035 7.672 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.475 6.753 mp_irecv_dv 2335 11.1 2.460 6.709 2.460 6.709 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.622 6.497 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.846 6.412 arnoldi_extremal 4 6.8 0.000 0.002 5.176 5.192 arnoldi_normal_ev 4 7.8 0.002 0.010 5.176 5.192 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.942 5.091 build_subspace 16 8.4 0.014 0.021 4.871 4.879 calculate_norms 1584 9.8 4.335 4.810 4.335 4.810 mp_allgather_i34 111 8.7 1.391 3.806 1.391 3.806 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.436 3.776 ls_scf_post 1 4.0 0.000 0.001 3.643 3.648 dbcsr_matrix_vector_mult_local 304 10.0 3.011 3.586 3.014 3.589 ls_scf_store_result 1 5.0 0.000 0.000 3.315 3.412 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.764 2.887 make_images_pack 222 9.7 1.816 2.133 1.819 2.136 dbcsr_sort_data 325 11.1 1.887 2.129 1.887 2.129 make_images_sizes 222 9.7 0.000 0.000 0.821 1.988 mp_alltoall_i44 222 10.7 0.821 1.988 0.821 1.988 dbcsr_data_release 9322 10.9 1.346 1.934 1.346 1.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.462000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2204.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 3.319292E+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 0.045 0.076 97.678 97.680 qs_energies 1 2.0 0.003 0.012 96.945 96.966 ls_scf 1 3.0 0.005 0.014 95.288 95.301 dbcsr_multiply_generic 111 6.7 0.017 0.019 78.781 78.997 ls_scf_main 1 4.0 0.002 0.014 58.998 59.001 multiply_cannon 111 7.7 0.048 0.147 51.738 55.513 density_matrix_trs4 2 5.0 0.003 0.017 52.994 53.100 multiply_cannon_loop 111 8.7 0.116 0.128 46.730 49.649 ls_scf_init_scf 1 4.0 0.001 0.001 33.022 33.027 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.546 31.611 mp_waitall_1 6438 11.0 22.868 29.491 22.868 29.491 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.051 29.065 multiply_cannon_multrec 1332 9.7 14.270 17.296 22.164 25.193 make_m2s 222 7.7 0.008 0.009 21.304 22.640 make_images 222 8.7 3.147 3.608 21.253 22.592 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.288 16.788 make_images_data 222 9.7 0.004 0.005 11.937 13.506 hybrid_alltoall_any 227 10.6 0.802 3.842 11.305 13.043 dbcsr_mm_accdrv_process 3641 10.4 0.230 0.414 7.540 9.080 dbcsr_mm_accdrv_process_sort 3641 11.4 7.164 8.658 7.164 8.658 mp_sum_l 807 5.4 4.153 7.124 4.153 7.124 multiply_cannon_sync_h2d 1332 9.7 5.566 6.347 5.566 6.347 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.096 6.101 mp_irecv_dv 3229 10.9 2.074 6.033 2.074 6.033 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.132 5.543 arnoldi_extremal 4 6.8 0.000 0.002 5.269 5.282 arnoldi_normal_ev 4 7.8 0.002 0.005 5.268 5.282 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.478 5.119 build_subspace 16 8.4 0.015 0.021 4.907 4.917 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.510 4.668 calculate_norms 2376 9.8 4.194 4.555 4.194 4.555 mp_allgather_i34 111 8.7 2.120 4.324 2.120 4.324 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.584 3.896 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.708 3.192 3.710 dbcsr_sort_data 658 11.4 3.096 3.587 3.096 3.587 ls_scf_post 1 4.0 0.008 0.017 3.262 3.275 dbcsr_special_finalize 555 9.7 0.006 0.007 2.849 3.271 dbcsr_merge_single_wm 555 10.7 0.538 0.672 2.841 3.263 ls_scf_store_result 1 5.0 0.000 0.000 2.980 3.040 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.910 2.969 dbcsr_data_release 10477 10.7 1.589 2.473 1.589 2.473 dbcsr_finalize 304 7.8 0.050 0.062 1.806 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.019 2.021 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.998 qs_ks_build_kohn_sham_matrix 3 8.3 0.018 0.023 1.996 1.998 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.680000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 4.637393E+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 0.064 0.103 98.979 98.980 qs_energies 1 2.0 0.000 0.000 98.026 98.032 ls_scf 1 3.0 0.000 0.000 96.106 96.111 dbcsr_multiply_generic 111 6.7 0.018 0.018 77.809 78.037 ls_scf_main 1 4.0 0.000 0.000 61.852 61.852 multiply_cannon 111 7.7 0.084 0.134 55.331 60.418 density_matrix_trs4 2 5.0 0.002 0.003 54.810 54.900 multiply_cannon_loop 111 8.7 0.070 0.078 50.770 52.491 mp_waitall_1 5481 11.0 26.330 32.257 26.330 32.257 ls_scf_init_scf 1 4.0 0.000 0.000 30.668 30.673 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.430 29.462 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.258 27.275 multiply_cannon_multrec 444 9.7 14.112 16.675 21.136 23.767 make_m2s 222 7.7 0.006 0.006 17.699 20.241 make_images 222 8.7 3.729 4.441 17.635 20.179 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.020 16.231 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.223 15.371 make_images_data 222 9.7 0.003 0.004 9.890 12.383 hybrid_alltoall_any 227 10.6 0.791 3.778 9.688 12.218 dbcsr_mm_accdrv_process 3003 10.4 0.191 0.354 6.730 7.864 multiply_cannon_sync_h2d 444 9.7 6.507 7.797 6.507 7.797 dbcsr_mm_accdrv_process_sort 3003 11.4 6.412 7.521 6.412 7.521 mp_allgather_i34 111 8.7 2.791 6.988 2.791 6.988 arnoldi_extremal 4 6.8 0.000 0.000 5.797 5.813 arnoldi_normal_ev 4 7.8 0.002 0.005 5.797 5.812 build_subspace 16 8.4 0.015 0.020 5.396 5.405 mp_sum_l 807 5.4 2.813 4.807 2.813 4.807 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.614 4.781 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.152 4.332 dbcsr_matrix_vector_mult_local 304 10.0 3.677 4.140 3.679 4.142 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.677 3.962 mp_irecv_dv 1241 11.2 1.659 3.941 1.659 3.941 calculate_norms 792 9.8 3.559 3.709 3.559 3.709 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.957 3.679 ls_scf_post 1 4.0 0.000 0.000 3.586 3.592 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.414 3.520 make_images_sizes 222 9.7 0.000 0.000 1.141 3.494 mp_alltoall_i44 222 10.7 1.140 3.493 1.140 3.493 ls_scf_store_result 1 5.0 0.000 0.000 3.368 3.411 dbcsr_finalize 304 7.8 0.062 0.078 2.210 2.281 dbcsr_merge_all 275 8.9 0.476 0.540 2.058 2.114 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.024 2.025 dbcsr_data_release 10123 10.8 1.336 1.993 1.336 1.993 rebuild_ks_matrix 3 7.3 0.000 0.000 1.990 1.991 qs_ks_build_kohn_sham_matrix 3 8.3 0.019 0.024 1.990 1.991 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.980000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3701.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d7412ee8675cb76473b00ae3e51e7018021298a2_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 8.823116E+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 0.116 0.136 108.252 108.253 qs_energies 1 2.0 0.000 0.000 106.656 106.671 ls_scf 1 3.0 0.000 0.000 103.765 103.779 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.810 77.928 ls_scf_main 1 4.0 0.000 0.000 65.924 65.924 density_matrix_trs4 2 5.0 0.002 0.003 57.037 57.110 multiply_cannon 111 7.7 0.185 0.260 49.992 51.567 multiply_cannon_loop 111 8.7 0.068 0.070 46.583 47.569 ls_scf_init_scf 1 4.0 0.000 0.000 34.135 34.136 ls_scf_init_matrix_S 1 5.0 0.002 0.014 32.733 32.749 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.857 29.865 mp_waitall_1 4569 11.1 22.409 26.436 22.409 26.436 make_m2s 222 7.7 0.007 0.007 24.054 24.973 make_images 222 8.7 4.596 4.969 23.943 24.863 multiply_cannon_multrec 444 9.7 17.906 18.622 22.524 23.167 hybrid_alltoall_any 227 10.6 1.664 3.644 12.961 15.862 make_images_data 222 9.7 0.003 0.004 13.240 15.634 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.577 11.514 multiply_cannon_sync_h2d 444 9.7 8.865 9.030 8.865 9.030 arnoldi_extremal 4 6.8 0.000 0.000 7.342 7.347 arnoldi_normal_ev 4 7.8 0.003 0.009 7.342 7.347 build_subspace 16 8.4 0.026 0.036 6.794 6.807 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.424 5.561 apply_matrix_preconditioner 6 5.3 0.009 0.024 5.073 5.314 dbcsr_matrix_vector_mult_local 304 10.0 5.008 5.295 5.011 5.298 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.823 4.918 dbcsr_mm_accdrv_process 1814 10.4 0.267 0.345 4.447 4.571 dbcsr_mm_accdrv_process_sort 1814 11.4 4.134 4.269 4.134 4.269 ls_scf_post 1 4.0 0.000 0.000 3.706 3.721 make_images_sizes 222 9.7 0.000 0.000 1.450 3.627 mp_alltoall_i44 222 10.7 1.450 3.627 1.450 3.627 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.440 calculate_norms 792 9.8 3.222 3.287 3.222 3.287 mp_allgather_i34 111 8.7 1.038 3.264 1.038 3.264 dbcsr_finalize 304 7.8 0.082 0.090 3.093 3.129 dbcsr_complete_redistribute 5 7.6 1.445 1.478 2.807 2.932 dbcsr_merge_all 275 8.9 0.895 0.922 2.877 2.907 dbcsr_data_release 12724 10.6 2.341 2.894 2.341 2.894 qs_energies_init_hamiltonians 1 3.0 0.003 0.005 2.861 2.861 matrix_ls_to_qs 2 6.0 0.000 0.000 2.451 2.578 dbcsr_sort_data 325 11.1 2.440 2.491 2.440 2.491 dbcsr_new_transposed 4 7.5 0.254 0.275 2.315 2.325 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.270 2.273 dbcsr_frobenius_norm 74 6.6 2.058 2.140 2.194 2.244 mp_sum_l 807 5.4 1.421 2.210 1.421 2.210 dbcsr_add_d 103 6.2 0.000 0.000 2.136 2.208 dbcsr_add_anytype 103 7.2 0.859 0.892 2.136 2.208 rebuild_ks_matrix 3 7.3 0.000 0.000 2.204 2.206 qs_ks_build_kohn_sham_matrix 3 8.3 0.037 0.040 2.204 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.253000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6862.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d7412ee8675cb76473b00ae3e51e7018021298a2 Summary: empty Status: OK