=== 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: a3b6180a81f48dbca65be969b5f13520da8e081a ################# 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/01 job id: 42408679 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/02 job id: 42408680 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/03 job id: 42408681 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/04 job id: 42408682 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/05 job id: 42408683 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/06 job id: 42408684 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/07 job id: 42408685 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/08 job id: 42408686 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/09 job id: 42408687 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/10 job id: 42408693 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/11 job id: 42408696 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/12 job id: 42408697 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/13 job id: 42408698 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/14 job id: 42408699 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/15 job id: 42408700 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/16 job id: 42408701 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/17 job id: 42408702 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/18 job id: 42408703 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/19 job id: 42408704 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/20 job id: 42408705 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/21 job id: 42408706 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/22 job id: 42408707 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/23 job id: 42408708 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/24 job id: 42408709 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/25 job id: 42408711 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/26 job id: 42408712 --- 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/a3b6180a81f48dbca65be969b5f13520da8e081a_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 0.082 0.124 132.688 132.689 farming_run 1 2.0 132.042 132.045 132.598 132.649 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.464164E+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 0.010 0.027 115.359 115.359 qs_energies 1 2.0 0.000 0.000 115.161 115.164 mp2_main 1 3.0 0.000 0.000 113.262 113.265 mp2_gpw_main 1 4.0 0.019 0.023 112.414 112.417 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.327 93.881 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.384 55.938 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.699 47.208 get_2c_integrals 1 6.0 0.000 0.000 37.334 37.771 integrate_v_rspace 273 8.0 0.440 0.453 25.084 30.369 pw_transfer 6555 10.6 0.372 0.389 27.359 28.261 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.984 26.774 grid_integrate_task_list 273 9.0 20.901 26.668 20.901 26.668 fft_wrap_pw1pw2_100 2178 12.4 1.218 1.419 23.549 24.341 compute_2c_integrals 1 7.0 0.002 0.002 19.686 19.687 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.951 19.418 mp2_eri_2c_integrate_gpw 1 9.0 2.371 2.438 18.948 19.417 rpa_ri_compute_en 1 5.0 0.000 0.000 18.979 19.133 cp_fm_cholesky_decompose 12 8.2 17.621 18.088 17.621 18.088 cholesky_decomp 1 7.0 0.000 0.000 16.493 16.955 fft3d_s 5443 13.4 16.098 16.660 16.120 16.680 ao_to_mo_and_store_B_mult_1 272 7.0 10.846 15.559 10.846 15.559 calculate_wavefunction 272 8.0 5.420 5.572 12.525 13.195 rpa_num_int 1 6.0 0.000 0.000 10.549 10.550 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.522 10.549 calc_mat_Q 8 8.0 0.000 0.000 9.351 9.453 contract_S_to_Q 8 9.0 0.000 0.000 8.771 8.872 calc_potential_gpw 544 9.5 0.005 0.006 8.278 8.692 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.260 8.571 potential_pw2rs 545 10.0 0.108 0.111 7.709 8.496 parallel_gemm_fm 14 9.1 0.000 0.000 8.345 8.443 parallel_gemm_fm_cosma 14 10.1 8.345 8.443 8.345 8.443 create_integ_mat 1 6.0 0.022 0.028 7.781 7.781 collocate_single_gaussian 272 10.0 0.041 0.043 7.497 7.770 array2fm 1 7.0 0.000 0.000 6.707 7.258 pw_scatter_s 2720 13.7 4.405 4.603 4.405 4.603 pw_gather_s 2722 13.2 3.893 4.248 3.893 4.248 array2fm_buffer_send 1 8.0 2.966 3.105 2.966 3.105 pw_poisson_solve 545 10.5 1.128 1.187 2.211 2.415 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.416722, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2737.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.038 395.960 395.961 farming_run 1 2.0 395.117 395.120 395.926 395.928 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.221689E+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 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 0.052 209.927 209.928 qs_energies 1 2.0 0.000 0.000 209.679 209.690 scf_env_do_scf 1 3.0 0.000 0.000 106.677 106.677 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.815 105.822 rebuild_ks_matrix 4 6.0 0.000 0.000 105.806 105.813 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.064 105.806 105.813 hfx_ks_matrix 4 8.0 0.001 0.001 105.420 105.425 integrate_four_center 4 9.0 0.144 0.457 105.419 105.424 mp2_main 1 3.0 0.000 0.000 102.725 102.736 mp2_gpw_main 1 4.0 0.033 0.046 101.895 101.908 integrate_four_center_main 4 10.0 0.119 0.633 96.635 99.528 integrate_four_center_bin 264 11.0 96.516 99.524 96.516 99.524 init_scf_loop 1 4.0 0.000 0.000 92.384 92.384 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.935 76.099 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.584 55.748 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.163 42.169 47.471 integrate_v_rspace 95 8.0 0.399 0.572 28.551 33.633 pw_transfer 2240 10.6 0.145 0.162 29.858 30.259 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.890 29.261 grid_integrate_task_list 95 9.0 23.830 29.079 23.830 29.079 mp2_ri_gpw_compute_en 1 5.0 0.061 0.074 26.817 28.669 ao_to_mo_and_store_B_mult_1 91 7.0 10.726 27.729 10.726 27.729 fft_wrap_pw1pw2_100 730 12.4 1.248 1.409 26.619 27.082 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.898 24.999 25.010 get_2c_integrals 1 6.0 0.000 0.000 20.257 20.287 compute_2c_integrals 1 7.0 0.002 0.003 19.223 19.231 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.786 19.101 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.837 18.785 19.099 fft3d_s 1823 13.4 18.397 18.824 18.410 18.838 scf_env_do_scf_inner_loop 4 4.0 0.000 0.009 14.292 14.292 calculate_wavefunction 91 8.0 2.021 2.052 9.742 10.018 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.743 9.328 potential_pw2rs 186 10.0 0.034 0.035 8.636 9.220 local_gemm 172 8.0 8.187 8.764 8.187 8.764 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.518 8.047 8.751 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.192 8.511 calc_potential_gpw 182 9.5 0.002 0.002 7.873 8.107 collocate_single_gaussian 91 10.0 0.017 0.025 7.851 8.107 mp_sendrecv_dm3 2068 8.0 6.086 6.810 6.086 6.810 mp_sync 38 10.4 3.397 6.689 3.397 6.689 mp2_ri_gpw_compute_en_ener 172 7.0 6.334 6.431 6.334 6.431 pw_gather_s 912 13.2 4.947 5.376 4.947 5.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.903678, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1506.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 451.248128E+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 998544. 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 0.011 0.028 51.981 51.983 qs_mol_dyn_low 1 2.0 0.003 0.005 51.740 51.747 qs_forces 11 3.9 0.002 0.008 51.674 51.675 qs_energies 11 4.9 0.001 0.001 50.226 50.236 scf_env_do_scf 11 5.9 0.000 0.001 44.041 44.042 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 41.935 41.935 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.795 34.296 qs_scf_new_mos 108 7.5 0.000 0.000 32.061 32.335 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.061 32.335 ot_scf_mini 108 9.5 0.002 0.002 30.370 30.572 multiply_cannon 2286 13.5 0.184 0.188 26.163 27.679 multiply_cannon_loop 2286 14.5 1.486 1.557 25.469 27.028 velocity_verlet 10 3.0 0.001 0.002 25.803 25.804 ot_mini 108 10.5 0.001 0.001 19.378 19.618 qs_ot_get_derivative 108 11.5 0.001 0.002 16.379 16.576 mp_waitall_1 267858 16.1 8.653 15.089 8.653 15.089 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.036 13.369 multiply_cannon_multrec 54864 15.5 4.213 6.454 7.797 11.372 rebuild_ks_matrix 119 8.3 0.000 0.000 7.794 7.913 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.015 7.794 7.912 multiply_cannon_sync_h2d 54864 15.5 5.884 7.724 5.884 7.724 mp_sum_l 7207 12.9 5.660 7.355 5.660 7.355 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.874 6.984 qs_ot_get_p 119 10.4 0.001 0.001 6.299 6.646 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.598 6.026 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.434 5.538 init_scf_run 11 5.9 0.000 0.001 4.951 4.951 scf_env_initial_rho_setup 11 6.9 0.000 0.002 4.951 4.951 dbcsr_mm_accdrv_process 76910 16.1 1.189 1.830 3.507 4.885 sum_up_and_integrate 119 10.3 0.012 0.017 4.450 4.457 integrate_v_rspace 119 11.3 0.002 0.002 4.438 4.446 qs_rho_update_rho_low 119 7.7 0.001 0.002 4.043 4.114 calculate_rho_elec 119 8.7 0.011 0.016 4.042 4.114 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.445 3.495 calculate_dm_sparse 119 9.5 0.000 0.000 3.037 3.162 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.906 3.093 apply_single 119 13.6 0.000 0.000 2.905 3.093 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.781 3.051 jit_kernel_multiply 13 15.8 2.254 2.984 2.254 2.984 rs_pw_transfer 974 11.9 0.012 0.013 2.722 2.822 calculate_first_density_matrix 1 7.0 0.000 0.001 2.766 2.769 ot_diis_step 108 11.5 0.006 0.010 2.695 2.696 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.596 2.597 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.492 2.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.450 2.452 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.403 2.404 cp_fm_redistribute_end 50 14.0 2.183 2.383 2.187 2.384 cp_fm_diag_elpa_base 50 14.0 0.196 2.350 0.197 2.360 acc_transpose_blocks 54864 15.5 0.226 0.242 1.794 2.232 density_rs2pw 119 9.7 0.004 0.004 2.125 2.226 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.168 2.217 grid_integrate_task_list 119 12.3 2.015 2.152 2.015 2.152 wfi_extrapolate 11 7.9 0.001 0.003 2.132 2.132 init_scf_loop 11 6.9 0.000 0.000 2.088 2.088 mp_sum_d 4129 12.0 1.369 1.978 1.369 1.978 potential_pw2rs 119 12.3 0.004 0.004 1.813 1.825 make_m2s 4572 13.5 0.053 0.055 1.661 1.722 pw_transfer 1439 11.6 0.051 0.056 1.608 1.684 make_images 4572 14.5 0.133 0.139 1.579 1.639 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.533 1.610 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.508 1.526 mp_alltoall_d11v 2130 13.8 1.346 1.477 1.346 1.477 mp_waitany 12084 13.8 1.260 1.409 1.260 1.409 fft3d_ps 1201 14.6 0.358 0.458 1.312 1.386 grid_collocate_task_list 119 9.7 1.288 1.352 1.288 1.352 fft_wrap_pw1pw2_140 487 13.2 0.080 0.094 1.181 1.256 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.750 1.156 acc_transpose_blocks_kernels 54864 16.5 0.253 0.381 0.812 1.100 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.983000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.624128E+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 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 0.026 0.048 38.724 38.725 qs_mol_dyn_low 1 2.0 0.003 0.003 38.467 38.474 qs_forces 11 3.9 0.002 0.003 38.411 38.412 qs_energies 11 4.9 0.002 0.004 36.706 36.709 scf_env_do_scf 11 5.9 0.001 0.001 31.232 31.233 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.719 28.720 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.673 22.068 qs_scf_new_mos 108 7.5 0.001 0.001 19.736 19.975 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.735 19.974 ot_scf_mini 108 9.5 0.002 0.004 18.839 19.006 multiply_cannon 2286 13.5 0.208 0.218 16.649 18.189 velocity_verlet 10 3.0 0.001 0.001 18.159 18.160 multiply_cannon_loop 2286 14.5 0.902 0.976 15.512 17.011 ot_mini 108 10.5 0.001 0.001 11.779 12.009 mp_waitall_1 217478 16.2 6.218 11.543 6.218 11.543 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.242 9.913 qs_ot_get_derivative 108 11.5 0.001 0.001 9.317 9.489 multiply_cannon_multrec 27432 15.5 1.977 4.531 6.061 9.179 rebuild_ks_matrix 119 8.3 0.000 0.000 7.203 7.343 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.203 7.343 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.348 6.474 dbcsr_mm_accdrv_process 47894 16.0 2.902 5.325 4.015 6.016 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.578 4.426 qs_ot_get_p 119 10.4 0.001 0.002 4.176 4.405 sum_up_and_integrate 119 10.3 0.024 0.027 4.242 4.252 init_scf_run 11 5.9 0.000 0.001 4.230 4.230 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.229 4.230 integrate_v_rspace 119 11.3 0.002 0.002 4.218 4.229 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.051 4.142 apply_single 119 13.6 0.000 0.000 3.051 4.142 mp_sum_l 7207 12.9 2.067 4.016 2.067 4.016 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.604 3.644 calculate_rho_elec 119 8.7 0.021 0.024 3.604 3.643 multiply_cannon_sync_h2d 27432 15.5 2.178 2.777 2.178 2.777 make_m2s 4572 13.5 0.052 0.054 2.515 2.762 rs_pw_transfer 974 11.9 0.010 0.011 2.638 2.737 calculate_first_density_matrix 1 7.0 0.001 0.002 2.729 2.731 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.703 2.721 make_images 4572 14.5 0.201 0.238 2.426 2.671 jit_kernel_multiply 12 16.1 1.061 2.632 1.061 2.632 init_scf_loop 11 6.9 0.000 0.002 2.490 2.494 ot_diis_step 108 11.5 0.011 0.011 2.413 2.414 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.277 2.279 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.272 2.273 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.165 2.252 calculate_dm_sparse 119 9.5 0.000 0.000 2.035 2.115 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.076 2.115 density_rs2pw 119 9.7 0.004 0.004 1.964 2.043 potential_pw2rs 119 12.3 0.006 0.007 1.977 1.990 grid_integrate_task_list 119 12.3 1.839 1.945 1.839 1.945 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.927 1.928 cp_fm_redistribute_end 50 14.0 1.585 1.903 1.589 1.904 acc_transpose_blocks 27432 15.5 0.108 0.114 1.241 1.893 cp_fm_diag_elpa_base 50 14.0 0.306 1.849 0.313 1.881 pw_transfer 1439 11.6 0.063 0.066 1.769 1.802 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.679 1.715 make_images_data 4572 15.5 0.045 0.052 1.154 1.593 prepare_preconditioner 11 7.9 0.000 0.000 1.562 1.588 make_preconditioner 11 8.9 0.000 0.001 1.562 1.588 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.455 1.511 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.007 1.473 wfi_extrapolate 11 7.9 0.001 0.001 1.449 1.450 fft3d_ps 1201 14.6 0.497 0.548 1.387 1.421 fft_wrap_pw1pw2_140 487 13.2 0.075 0.082 1.308 1.342 mp_allgather_i34 2286 14.5 0.582 1.341 0.582 1.341 mp_alltoall_d11v 2130 13.8 1.228 1.334 1.228 1.334 grid_collocate_task_list 119 9.7 1.240 1.307 1.240 1.307 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.709 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.255 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.180 1.226 mp_sum_d 4129 12.0 0.621 1.066 0.621 1.066 jit_kernel_transpose 5 15.5 0.527 1.024 0.527 1.024 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.968 0.968 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.912 0.925 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.606 0.788 0.832 make_images_sizes 4572 15.5 0.005 0.005 0.548 0.789 mp_alltoall_i44 4572 16.5 0.543 0.784 0.543 0.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.725000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.499311 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 521.543680E+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 0.019 0.032 32.380 32.381 qs_mol_dyn_low 1 2.0 0.003 0.005 32.037 32.044 qs_forces 11 3.9 0.002 0.005 31.978 31.979 qs_energies 11 4.9 0.003 0.011 30.412 30.416 scf_env_do_scf 11 5.9 0.001 0.001 25.481 25.481 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 22.830 22.832 dbcsr_multiply_generic 2286 12.5 0.095 0.099 16.631 16.771 velocity_verlet 10 3.0 0.001 0.002 15.270 15.272 qs_scf_new_mos 108 7.5 0.001 0.001 14.852 14.871 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.852 14.871 multiply_cannon 2286 13.5 0.195 0.206 13.235 14.164 ot_scf_mini 108 9.5 0.002 0.003 14.116 14.132 multiply_cannon_loop 2286 14.5 0.635 0.665 12.434 13.416 ot_mini 108 10.5 0.001 0.003 8.807 8.826 qs_ot_get_derivative 108 11.5 0.001 0.004 7.285 7.299 multiply_cannon_multrec 18288 15.5 1.906 2.861 6.883 7.218 rebuild_ks_matrix 119 8.3 0.000 0.000 6.355 6.372 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.016 6.355 6.371 dbcsr_mm_accdrv_process 38222 16.0 4.260 5.573 4.894 5.794 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.617 5.633 mp_waitall_1 169478 16.3 3.053 4.171 3.053 4.171 sum_up_and_integrate 119 10.3 0.029 0.032 3.959 3.976 integrate_v_rspace 119 11.3 0.002 0.003 3.930 3.946 init_scf_run 11 5.9 0.000 0.001 3.708 3.708 scf_env_initial_rho_setup 11 6.9 0.000 0.002 3.708 3.708 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.744 3.364 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.263 3.284 calculate_rho_elec 119 8.7 0.030 0.032 3.262 3.283 qs_ot_get_p 119 10.4 0.001 0.001 3.130 3.157 init_scf_loop 11 6.9 0.002 0.009 2.634 2.636 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.465 2.498 calculate_first_density_matrix 1 7.0 0.001 0.003 2.477 2.478 rs_pw_transfer 974 11.9 0.009 0.010 2.316 2.452 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.049 2.367 apply_single 119 13.6 0.000 0.000 2.049 2.367 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.011 2.017 jit_kernel_multiply 10 16.4 0.582 1.989 0.582 1.989 density_rs2pw 119 9.7 0.004 0.004 1.846 1.987 make_m2s 4572 13.5 0.045 0.046 1.830 1.968 calculate_dm_sparse 119 9.5 0.000 0.000 1.924 1.941 grid_integrate_task_list 119 12.3 1.814 1.900 1.814 1.900 make_images 4572 14.5 0.191 0.204 1.745 1.879 prepare_preconditioner 11 7.9 0.000 0.000 1.796 1.799 make_preconditioner 11 8.9 0.000 0.001 1.796 1.799 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.761 1.763 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.747 1.747 potential_pw2rs 119 12.3 0.007 0.008 1.737 1.746 pw_transfer 1439 11.6 0.063 0.066 1.726 1.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.650 1.734 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.635 1.656 multiply_cannon_sync_h2d 18288 15.5 1.381 1.616 1.381 1.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.600 1.612 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.590 1.597 mp_sum_l 7207 12.9 1.177 1.593 1.177 1.593 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.515 1.516 ot_diis_step 108 11.5 0.011 0.014 1.496 1.496 cp_fm_redistribute_end 50 14.0 1.123 1.495 1.125 1.496 cp_fm_diag_elpa_base 50 14.0 0.355 1.440 0.369 1.479 fft3d_ps 1201 14.6 0.506 0.524 1.323 1.345 grid_collocate_task_list 119 9.7 1.204 1.309 1.204 1.309 fft_wrap_pw1pw2_140 487 13.2 0.086 0.089 1.284 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.282 1.287 acc_transpose_blocks 18288 15.5 0.075 0.076 1.251 1.273 wfi_extrapolate 11 7.9 0.001 0.005 1.173 1.174 make_images_data 4572 15.5 0.045 0.048 0.831 0.997 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.976 0.978 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.895 0.921 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.717 0.903 mp_alltoall_d11v 2130 13.8 0.700 0.877 0.700 0.877 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.811 0.824 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.813 0.816 cp_fm_cholesky_invert 11 10.9 0.756 0.760 0.756 0.760 mp_alltoall_z22v 1201 16.6 0.658 0.738 0.658 0.738 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.714 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.544 0.681 mp_waitany 9880 13.7 0.555 0.679 0.555 0.679 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.381000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.000000, yerr=2.449490 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 553.365504E+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 0.011 0.028 34.164 34.164 qs_mol_dyn_low 1 2.0 0.003 0.004 33.993 34.000 qs_forces 11 3.9 0.002 0.002 33.936 33.937 qs_energies 11 4.9 0.001 0.001 32.243 32.248 scf_env_do_scf 11 5.9 0.000 0.001 27.314 27.316 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.930 23.931 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.008 18.126 velocity_verlet 10 3.0 0.001 0.001 17.476 17.478 qs_scf_new_mos 108 7.5 0.001 0.001 15.858 15.913 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.858 15.912 ot_scf_mini 108 9.5 0.002 0.003 14.932 14.984 multiply_cannon 2286 13.5 0.228 0.271 14.421 14.897 multiply_cannon_loop 2286 14.5 0.934 0.961 13.488 13.859 ot_mini 108 10.5 0.001 0.001 9.194 9.259 multiply_cannon_multrec 27432 15.5 2.340 3.042 8.677 9.060 dbcsr_mm_accdrv_process 47916 15.9 5.558 7.309 6.245 7.452 qs_ot_get_derivative 108 11.5 0.001 0.001 7.394 7.445 rebuild_ks_matrix 119 8.3 0.000 0.000 6.415 6.470 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.414 6.470 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.695 5.745 sum_up_and_integrate 119 10.3 0.034 0.037 3.719 3.728 integrate_v_rspace 119 11.3 0.002 0.003 3.685 3.695 init_scf_run 11 5.9 0.000 0.001 3.616 3.616 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.616 3.616 init_scf_loop 11 6.9 0.000 0.000 3.364 3.365 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.261 3.295 calculate_rho_elec 119 8.7 0.040 0.046 3.260 3.294 qs_ot_get_p 119 10.4 0.001 0.001 3.150 3.221 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.729 3.206 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.111 2.590 apply_single 119 13.6 0.000 0.000 2.111 2.590 mp_waitall_1 145218 16.4 1.994 2.574 1.994 2.574 prepare_preconditioner 11 7.9 0.000 0.000 2.538 2.546 make_preconditioner 11 8.9 0.000 0.000 2.538 2.546 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.143 2.473 calculate_first_density_matrix 1 7.0 0.000 0.000 2.280 2.282 make_m2s 4572 13.5 0.054 0.056 2.160 2.265 make_images 4572 14.5 0.273 0.336 2.052 2.156 calculate_dm_sparse 119 9.5 0.000 0.000 2.078 2.129 rs_pw_transfer 974 11.9 0.009 0.009 1.950 2.076 grid_integrate_task_list 119 12.3 1.826 1.912 1.826 1.912 qs_ot_p2m_diag 50 11.0 0.015 0.022 1.895 1.903 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.877 1.903 jit_kernel_multiply 10 16.3 0.625 1.882 0.625 1.882 density_rs2pw 119 9.7 0.004 0.004 1.743 1.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.780 1.781 pw_transfer 1439 11.6 0.063 0.066 1.747 1.777 ot_diis_step 108 11.5 0.012 0.012 1.759 1.760 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.658 1.692 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.612 1.628 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.588 1.588 mp_sum_l 7207 12.9 0.997 1.545 0.997 1.545 potential_pw2rs 119 12.3 0.008 0.010 1.518 1.521 acc_transpose_blocks 27432 15.5 0.112 0.114 1.462 1.486 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.349 1.383 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.365 1.366 fft3d_ps 1201 14.6 0.534 0.584 1.339 1.364 cp_fm_redistribute_end 50 14.0 0.899 1.344 0.900 1.344 cp_fm_diag_elpa_base 50 14.0 0.422 1.287 0.442 1.328 grid_collocate_task_list 119 9.7 1.221 1.306 1.221 1.306 wfi_extrapolate 11 7.9 0.001 0.001 1.289 1.290 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.271 1.282 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.737 1.220 cp_fm_upper_to_full 72 13.5 0.807 1.120 0.807 1.120 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.087 1.106 multiply_cannon_sync_h2d 27432 15.5 1.002 1.073 1.002 1.073 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.063 1.064 dbcsr_complete_redistribute 329 12.2 0.120 0.147 0.749 1.025 make_images_data 4572 15.5 0.045 0.049 0.836 0.960 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.706 0.895 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.868 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.836 0.854 mp_alltoall_d11v 2130 13.8 0.707 0.832 0.707 0.832 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.558 0.830 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.808 0.812 cp_fm_cholesky_invert 11 10.9 0.753 0.757 0.753 0.757 mp_alltoall_i22 627 13.8 0.425 0.725 0.425 0.725 mp_alltoall_z22v 1201 16.6 0.687 0.719 0.687 0.719 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.164000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=526.363636, yerr=2.672171 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 609.452032E+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 0.012 0.030 28.041 28.041 qs_mol_dyn_low 1 2.0 0.003 0.003 27.849 27.856 qs_forces 11 3.9 0.007 0.008 27.792 27.793 qs_energies 11 4.9 0.001 0.002 26.095 26.098 scf_env_do_scf 11 5.9 0.000 0.001 21.255 21.256 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.715 18.715 velocity_verlet 10 3.0 0.001 0.001 14.202 14.205 dbcsr_multiply_generic 2286 12.5 0.093 0.096 12.428 12.544 qs_scf_new_mos 108 7.5 0.001 0.001 11.191 11.217 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.190 11.216 ot_scf_mini 108 9.5 0.002 0.002 10.511 10.536 multiply_cannon 2286 13.5 0.229 0.237 9.808 10.356 multiply_cannon_loop 2286 14.5 0.332 0.343 8.874 9.077 multiply_cannon_multrec 9144 15.5 1.582 1.910 5.913 6.208 ot_mini 108 10.5 0.001 0.001 5.938 5.971 rebuild_ks_matrix 119 8.3 0.000 0.000 5.809 5.833 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.809 5.832 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.177 5.200 qs_ot_get_derivative 108 11.5 0.001 0.001 4.645 4.671 dbcsr_mm_accdrv_process 12550 15.8 3.334 4.257 4.231 4.319 sum_up_and_integrate 119 10.3 0.037 0.040 3.575 3.580 integrate_v_rspace 119 11.3 0.003 0.003 3.538 3.543 init_scf_run 11 5.9 0.000 0.001 3.400 3.400 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.400 3.400 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.160 3.167 calculate_rho_elec 119 8.7 0.059 0.061 3.160 3.167 qs_ot_get_p 119 10.4 0.001 0.001 2.782 2.820 init_scf_loop 11 6.9 0.000 0.000 2.515 2.516 mp_waitall_1 121218 16.5 1.815 2.298 1.815 2.298 calculate_first_density_matrix 1 7.0 0.000 0.000 2.280 2.282 make_m2s 4572 13.5 0.034 0.035 1.783 1.931 grid_integrate_task_list 119 12.3 1.846 1.914 1.846 1.914 make_images 4572 14.5 0.268 0.302 1.694 1.840 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.813 1.815 prepare_preconditioner 11 7.9 0.000 0.000 1.791 1.795 make_preconditioner 11 8.9 0.000 0.000 1.791 1.795 calculate_dm_sparse 119 9.5 0.000 0.000 1.766 1.784 rs_pw_transfer 974 11.9 0.008 0.008 1.683 1.767 pw_transfer 1439 11.6 0.063 0.065 1.714 1.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.680 1.707 density_rs2pw 119 9.7 0.004 0.004 1.624 1.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.669 1.670 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.624 1.635 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.612 1.613 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.574 1.591 jit_kernel_multiply 8 15.6 0.859 1.565 0.859 1.565 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.449 1.459 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.369 1.370 potential_pw2rs 119 12.3 0.010 0.011 1.362 1.368 grid_collocate_task_list 119 9.7 1.272 1.346 1.272 1.346 cp_fm_redistribute_end 50 14.0 0.678 1.345 0.679 1.345 cp_fm_diag_elpa_base 50 14.0 0.621 1.274 0.665 1.330 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.317 1.328 fft3d_ps 1201 14.6 0.538 0.551 1.291 1.304 ot_diis_step 108 11.5 0.012 0.013 1.277 1.277 qs_energies_init_hamiltonians 11 5.9 0.006 0.012 1.219 1.219 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.201 1.212 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.187 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.146 1.174 apply_single 119 13.6 0.000 0.000 1.146 1.174 wfi_extrapolate 11 7.9 0.001 0.001 1.075 1.075 make_images_data 4572 15.5 0.039 0.043 0.830 1.015 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.749 0.992 cp_fm_cholesky_invert 11 10.9 0.919 0.922 0.919 0.922 acc_transpose_blocks 9144 15.5 0.039 0.041 0.910 0.921 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.916 mp_alltoall_d11v 2130 13.8 0.718 0.794 0.718 0.794 multiply_cannon_sync_h2d 9144 15.5 0.705 0.780 0.705 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.757 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.734 0.743 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.398 0.716 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.666 0.711 mp_allgather_i34 2286 14.5 0.268 0.683 0.268 0.683 acc_transpose_blocks_kernels 9144 16.5 0.117 0.121 0.668 0.674 mp_alltoall_z22v 1201 16.6 0.627 0.661 0.627 0.661 dbcsr_complete_redistribute 329 12.2 0.196 0.231 0.591 0.619 qs_create_task_list 11 7.9 0.000 0.000 0.546 0.568 generate_qs_task_list 11 8.9 0.191 0.213 0.546 0.568 mp_sum_l 7207 12.9 0.405 0.568 0.405 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.041000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=574.000000, yerr=8.790491 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 758.218752E+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 0.043 0.047 40.780 40.781 qs_mol_dyn_low 1 2.0 0.005 0.018 40.303 40.309 qs_forces 11 3.9 0.009 0.010 40.243 40.245 qs_energies 11 4.9 0.001 0.002 38.318 38.321 scf_env_do_scf 11 5.9 0.000 0.001 32.206 32.206 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.322 24.323 velocity_verlet 10 3.0 0.001 0.001 22.608 22.613 dbcsr_multiply_generic 2286 12.5 0.121 0.123 17.438 17.674 qs_scf_new_mos 108 7.5 0.001 0.001 15.549 15.647 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.548 15.646 multiply_cannon 2286 13.5 0.304 0.313 13.761 14.730 ot_scf_mini 108 9.5 0.002 0.002 14.467 14.566 multiply_cannon_loop 2286 14.5 0.344 0.354 12.541 13.548 multiply_cannon_multrec 9144 15.5 3.367 4.729 8.757 8.858 ot_mini 108 10.5 0.001 0.001 8.739 8.852 init_scf_loop 11 6.9 0.000 0.000 7.858 7.859 prepare_preconditioner 11 7.9 0.000 0.000 6.931 6.948 make_preconditioner 11 8.9 0.000 0.000 6.931 6.948 qs_ot_get_derivative 108 11.5 0.001 0.001 6.759 6.855 rebuild_ks_matrix 119 8.3 0.000 0.000 6.692 6.838 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.692 6.837 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.449 6.828 dbcsr_mm_accdrv_process 12550 15.8 4.121 5.722 5.268 6.541 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.057 6.189 cp_fm_upper_to_full 72 14.2 3.292 4.722 3.292 4.722 init_scf_run 11 5.9 0.000 0.001 4.092 4.092 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.092 4.092 sum_up_and_integrate 119 10.3 0.064 0.065 3.597 3.603 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.586 3.594 calculate_rho_elec 119 8.7 0.118 0.121 3.585 3.594 integrate_v_rspace 119 11.3 0.003 0.003 3.533 3.540 mp_waitall_1 97218 16.6 2.362 3.374 2.362 3.374 qs_ot_get_p 119 10.4 0.001 0.001 3.044 3.174 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.518 2.927 dbcsr_complete_redistribute 329 12.2 0.297 0.300 2.003 2.845 calculate_first_density_matrix 1 7.0 0.000 0.000 2.706 2.707 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.712 2.548 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.138 2.423 apply_single 119 13.6 0.000 0.000 2.138 2.423 mp_alltoall_i22 627 13.8 1.468 2.330 1.468 2.330 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.477 2.306 make_m2s 4572 13.5 0.038 0.038 2.140 2.287 calculate_dm_sparse 119 9.5 0.000 0.000 2.227 2.246 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.285 2.228 make_images 4572 14.5 0.356 0.385 2.020 2.167 grid_integrate_task_list 119 12.3 1.992 2.010 1.992 2.010 pw_transfer 1439 11.6 0.066 0.066 1.964 1.967 ot_diis_step 108 11.5 0.014 0.014 1.956 1.956 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.952 1.952 jit_kernel_multiply 10 15.4 1.119 1.933 1.119 1.933 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.869 1.872 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.819 1.821 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.758 1.760 density_rs2pw 119 9.7 0.003 0.004 1.704 1.722 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.680 1.710 mp_sum_l 7207 12.9 0.974 1.695 0.974 1.695 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.639 1.688 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.569 1.574 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.555 1.555 fft3d_ps 1201 14.6 0.568 0.577 1.503 1.506 grid_collocate_task_list 119 9.7 1.437 1.444 1.437 1.444 cp_fm_cholesky_invert 11 10.9 1.417 1.420 1.417 1.420 rs_pw_transfer 974 11.9 0.009 0.009 1.317 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.329 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.291 1.300 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.280 1.280 cp_fm_diag_elpa_base 50 14.0 1.132 1.186 1.278 1.278 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.037 1.248 potential_pw2rs 119 12.3 0.014 0.015 1.165 1.167 make_images_data 4572 15.5 0.043 0.046 0.972 1.155 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.125 1.146 mp_alltoall_d11v 2130 13.8 1.104 1.127 1.104 1.127 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.086 1.104 multiply_cannon_sync_h2d 9144 15.5 1.038 1.042 1.038 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.036 acc_transpose_blocks 9144 15.5 0.039 0.039 0.958 0.977 qs_create_task_list 11 7.9 0.000 0.000 0.938 0.952 generate_qs_task_list 11 8.9 0.372 0.392 0.938 0.952 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.846 0.859 mp_alltoall_z22v 1201 16.6 0.800 0.819 0.800 0.819 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=40.781000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=706.545455, yerr=16.832672 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 501.403648E+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 1884922. 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 0.033 0.116 82.466 82.466 qs_mol_dyn_low 1 2.0 0.003 0.004 81.997 82.006 qs_forces 11 3.9 0.002 0.004 81.911 81.912 qs_energies 11 4.9 0.003 0.010 79.118 79.132 scf_env_do_scf 11 5.9 0.001 0.001 70.072 70.075 scf_env_do_scf_inner_loop 99 6.5 0.004 0.021 64.562 64.563 dbcsr_multiply_generic 2055 12.4 0.109 0.138 52.287 52.686 qs_scf_new_mos 99 7.5 0.000 0.001 47.877 48.016 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.877 48.015 ot_scf_mini 99 9.5 0.002 0.003 45.446 45.557 multiply_cannon 2055 13.4 0.179 0.242 42.737 43.964 multiply_cannon_loop 2055 14.4 1.541 1.586 41.728 42.963 velocity_verlet 10 3.0 0.001 0.001 41.447 41.447 ot_mini 99 10.5 0.001 0.001 27.645 27.753 qs_ot_get_derivative 99 11.5 0.001 0.001 20.807 20.925 multiply_cannon_multrec 49320 15.4 12.163 12.955 17.168 17.789 rebuild_ks_matrix 110 8.3 0.000 0.001 14.296 14.451 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.296 14.450 mp_waitall_1 241148 16.1 11.822 12.831 11.822 12.831 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.566 12.711 multiply_cannon_sync_h2d 49320 15.4 10.031 10.621 10.031 10.621 qs_ot_get_p 110 10.4 0.001 0.001 9.266 9.421 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.055 8.631 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.178 7.639 apply_single 110 13.6 0.000 0.000 7.178 7.639 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.964 7.515 init_scf_run 11 5.9 0.000 0.002 6.957 6.958 scf_env_initial_rho_setup 11 6.9 0.001 0.004 6.957 6.958 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.843 6.920 sum_up_and_integrate 110 10.3 0.037 0.043 6.893 6.907 integrate_v_rspace 110 11.3 0.003 0.003 6.856 6.879 ot_diis_step 99 11.5 0.005 0.006 6.561 6.561 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.072 6.180 calculate_rho_elec 110 8.6 0.020 0.024 6.071 6.180 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.082 6.109 mp_sum_l 6514 12.8 5.208 5.855 5.208 5.855 init_scf_loop 11 6.9 0.001 0.003 5.481 5.482 dbcsr_mm_accdrv_process 87628 16.1 2.061 2.171 4.884 5.219 cp_dbcsr_syevd 48 12.0 0.003 0.004 5.058 5.059 cp_fm_diag_elpa 48 13.0 0.000 0.001 4.582 4.584 cp_fm_redistribute_end 48 14.0 3.986 4.556 3.990 4.558 cp_fm_diag_elpa_base 48 14.0 0.561 4.493 0.565 4.518 make_m2s 4110 13.4 0.060 0.067 3.924 4.075 make_images 4110 14.4 0.179 0.192 3.829 3.983 wfi_extrapolate 11 7.9 0.001 0.002 3.979 3.979 calculate_dm_sparse 110 9.5 0.000 0.001 3.819 3.901 multiply_cannon_metrocomm1 49320 15.4 0.059 0.062 2.717 3.812 rs_pw_transfer 902 11.9 0.012 0.013 3.637 3.811 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.748 3.752 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.386 3.451 grid_integrate_task_list 110 12.3 3.244 3.424 3.244 3.424 prepare_preconditioner 11 7.9 0.000 0.000 3.327 3.353 make_preconditioner 11 8.9 0.000 0.001 3.327 3.353 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.256 3.306 density_rs2pw 110 9.6 0.004 0.005 3.145 3.284 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.119 3.180 calculate_first_density_matrix 1 7.0 0.005 0.021 2.875 2.888 pw_transfer 1331 11.6 0.053 0.063 2.766 2.821 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.679 2.735 jit_kernel_multiply 13 15.9 2.538 2.575 2.538 2.575 mp_alltoall_d11v 2046 13.8 2.083 2.533 2.083 2.533 potential_pw2rs 110 12.3 0.006 0.007 2.493 2.514 fft_wrap_pw1pw2_140 451 13.1 0.169 0.188 2.267 2.324 fft3d_ps 1111 14.6 0.752 0.837 2.216 2.258 acc_transpose_blocks 49320 15.4 0.212 0.223 2.142 2.215 grid_collocate_task_list 110 9.6 2.085 2.189 2.085 2.189 mp_waitany 14300 13.8 1.822 2.150 1.822 2.150 mp_sum_d 3883 11.9 1.480 2.026 1.480 2.026 make_images_data 4110 15.4 0.043 0.047 1.820 2.000 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.869 1.891 hybrid_alltoall_any 4261 16.3 0.082 0.477 1.548 1.853 cp_fm_cholesky_invert 11 10.9 1.830 1.834 1.830 1.834 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.659 1.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.466000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.454545, yerr=2.462949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 585.265152E+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 1916874. 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 0.034 0.088 68.378 68.381 qs_mol_dyn_low 1 2.0 0.003 0.006 67.786 67.795 qs_forces 11 3.9 0.003 0.005 67.724 67.724 qs_energies 11 4.9 0.005 0.020 64.445 64.449 scf_env_do_scf 11 5.9 0.001 0.001 55.660 55.663 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 48.113 48.113 dbcsr_multiply_generic 2055 12.4 0.118 0.135 37.652 37.880 velocity_verlet 10 3.0 0.001 0.001 35.635 35.636 qs_scf_new_mos 99 7.5 0.001 0.001 32.366 32.534 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.366 32.534 multiply_cannon 2055 13.4 0.221 0.246 31.053 32.249 multiply_cannon_loop 2055 14.4 0.930 0.956 29.759 30.967 ot_scf_mini 99 9.5 0.003 0.003 30.703 30.852 ot_mini 99 10.5 0.001 0.001 18.117 18.277 multiply_cannon_multrec 24660 15.4 7.613 9.100 13.826 15.094 rebuild_ks_matrix 110 8.3 0.000 0.001 13.613 13.779 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 13.613 13.779 qs_ot_get_derivative 99 11.5 0.001 0.001 12.297 12.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.022 12.166 mp_waitall_1 186928 16.3 7.883 10.654 7.883 10.654 multiply_cannon_sync_h2d 24660 15.4 7.060 8.254 7.060 8.254 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.190 7.728 init_scf_loop 11 6.9 0.001 0.005 7.512 7.513 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.535 7.229 apply_single 110 13.6 0.000 0.001 6.535 7.228 sum_up_and_integrate 110 10.3 0.052 0.058 6.548 6.562 integrate_v_rspace 110 11.3 0.002 0.003 6.496 6.510 dbcsr_mm_accdrv_process 52282 16.1 4.674 5.728 6.053 6.393 init_scf_run 11 5.9 0.000 0.001 6.269 6.272 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.269 6.272 qs_ot_get_p 110 10.4 0.001 0.001 5.958 6.136 ot_diis_step 99 11.5 0.010 0.010 5.772 5.772 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.647 5.655 calculate_rho_elec 110 8.6 0.039 0.047 5.646 5.655 prepare_preconditioner 11 7.9 0.000 0.000 5.503 5.517 make_preconditioner 11 8.9 0.000 0.003 5.503 5.517 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.613 5.313 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.100 5.257 make_m2s 4110 13.4 0.057 0.060 4.152 4.626 make_images 4110 14.4 0.403 0.443 4.043 4.516 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.089 4.110 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.659 3.660 wfi_extrapolate 11 7.9 0.001 0.001 3.476 3.476 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.383 3.385 pw_transfer 1331 11.6 0.065 0.070 3.212 3.344 grid_integrate_task_list 110 12.3 3.158 3.308 3.158 3.308 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.176 3.252 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.106 3.239 rs_pw_transfer 902 11.9 0.012 0.014 3.005 3.188 density_rs2pw 110 9.6 0.004 0.004 2.960 3.128 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.124 3.125 cp_fm_redistribute_end 48 14.0 2.326 3.100 2.328 3.100 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.016 3.085 cp_fm_diag_elpa_base 48 14.0 0.737 2.990 0.768 3.076 calculate_dm_sparse 110 9.5 0.001 0.001 2.981 3.010 make_images_data 4110 15.4 0.048 0.059 2.315 2.791 fft_wrap_pw1pw2_140 451 13.1 0.201 0.218 2.608 2.740 hybrid_alltoall_any 4261 16.3 0.102 0.442 1.991 2.737 calculate_first_density_matrix 1 7.0 0.001 0.005 2.706 2.710 fft3d_ps 1111 14.6 1.073 1.266 2.463 2.585 cp_fm_cholesky_invert 11 10.9 2.576 2.583 2.576 2.583 mp_sum_l 6514 12.8 1.854 2.498 1.854 2.498 potential_pw2rs 110 12.3 0.008 0.009 2.425 2.445 jit_kernel_multiply 12 16.4 1.024 2.290 1.024 2.290 grid_collocate_task_list 110 9.6 2.118 2.278 2.118 2.278 qs_energies_init_hamiltonians 11 5.9 0.018 0.027 1.997 1.998 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.946 1.969 mp_alltoall_d11v 2046 13.8 1.704 1.883 1.704 1.883 multiply_cannon_metrocomm4 22605 15.4 0.074 0.076 0.780 1.781 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.755 1.775 mp_allgather_i34 2055 14.4 0.619 1.636 0.619 1.636 mp_irecv_dv 57340 16.2 0.656 1.591 0.656 1.591 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.561 acc_transpose_blocks 24660 15.4 0.107 0.111 1.526 1.554 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.367 1.477 dbcsr_complete_redistribute 325 12.2 0.252 0.310 1.135 1.394 mp_sum_d 3883 11.9 0.979 1.390 0.979 1.390 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.381000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.545455, yerr=6.329780 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.258816E+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 0.073 0.131 59.791 59.794 qs_mol_dyn_low 1 2.0 0.003 0.003 59.156 59.165 qs_forces 11 3.9 0.003 0.009 58.855 58.856 qs_energies 11 4.9 0.033 0.045 55.690 55.694 scf_env_do_scf 11 5.9 0.001 0.001 47.240 47.241 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 38.879 38.879 velocity_verlet 10 3.0 0.001 0.001 32.102 32.111 dbcsr_multiply_generic 2055 12.4 0.109 0.120 28.370 28.696 qs_scf_new_mos 99 7.5 0.001 0.001 24.592 24.687 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.591 24.687 ot_scf_mini 99 9.5 0.003 0.003 23.383 23.493 multiply_cannon 2055 13.4 0.213 0.231 21.902 23.204 multiply_cannon_loop 2055 14.4 0.620 0.635 20.776 21.882 ot_mini 99 10.5 0.001 0.001 13.323 13.435 rebuild_ks_matrix 110 8.3 0.000 0.000 12.044 12.206 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.043 12.205 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.625 10.773 multiply_cannon_multrec 16440 15.4 3.874 4.679 9.609 10.401 mp_waitall_1 146766 16.3 7.058 9.776 7.058 9.776 qs_ot_get_derivative 99 11.5 0.001 0.001 8.909 9.021 init_scf_loop 11 6.9 0.001 0.005 8.328 8.329 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.253 7.043 prepare_preconditioner 11 7.9 0.000 0.000 6.596 6.617 make_preconditioner 11 8.9 0.000 0.000 6.596 6.617 sum_up_and_integrate 110 10.3 0.060 0.061 6.301 6.316 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.954 6.289 integrate_v_rspace 110 11.3 0.002 0.003 6.241 6.256 init_scf_run 11 5.9 0.000 0.001 5.844 5.845 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.844 5.844 dbcsr_mm_accdrv_process 34862 16.1 4.704 5.251 5.589 5.726 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.421 5.430 calculate_rho_elec 110 8.6 0.058 0.058 5.420 5.430 qs_ot_get_p 110 10.4 0.001 0.001 5.258 5.398 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.933 5.316 apply_single 110 13.6 0.000 0.000 4.932 5.316 make_m2s 4110 13.4 0.050 0.052 4.303 4.642 make_images 4110 14.4 0.410 0.512 4.188 4.524 ot_diis_step 99 11.5 0.010 0.011 4.386 4.387 multiply_cannon_sync_h2d 16440 15.4 3.708 4.179 3.708 4.179 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.039 3.698 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.676 3.679 grid_integrate_task_list 110 12.3 3.206 3.421 3.206 3.421 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.347 3.348 pw_transfer 1331 11.6 0.064 0.070 3.117 3.123 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.018 3.020 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.011 3.018 density_rs2pw 110 9.6 0.004 0.004 2.709 2.950 wfi_extrapolate 11 7.9 0.001 0.001 2.925 2.925 make_images_data 4110 15.4 0.044 0.048 2.472 2.908 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.168 2.871 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.848 2.849 calculate_first_density_matrix 1 7.0 0.002 0.005 2.839 2.841 rs_pw_transfer 902 11.9 0.010 0.011 2.654 2.838 cp_fm_redistribute_end 48 14.0 1.769 2.826 1.771 2.826 cp_fm_diag_elpa_base 48 14.0 0.994 2.698 1.052 2.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.654 2.729 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.595 2.602 cp_fm_cholesky_invert 11 10.9 2.575 2.581 2.575 2.581 calculate_dm_sparse 110 9.5 0.001 0.001 2.482 2.510 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.878 2.464 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.276 2.346 mp_irecv_dv 48980 15.7 0.807 2.339 0.807 2.339 fft3d_ps 1111 14.6 1.063 1.072 2.325 2.331 grid_collocate_task_list 110 9.6 2.180 2.317 2.180 2.317 mp_sum_l 6514 12.8 1.553 2.158 1.553 2.158 potential_pw2rs 110 12.3 0.010 0.010 2.147 2.155 qs_energies_init_hamiltonians 11 5.9 0.011 0.036 2.132 2.134 mp_alltoall_d11v 2046 13.8 1.686 2.035 1.686 2.035 dbcsr_complete_redistribute 325 12.2 0.341 0.454 1.426 1.937 cp_fm_upper_to_full 70 13.6 1.367 1.814 1.367 1.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.618 1.637 mp_allgather_i34 2055 14.4 0.451 1.602 0.451 1.602 cp_fm_cholesky_decompose 22 10.9 1.543 1.562 1.543 1.562 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.363 1.483 jit_kernel_multiply 8 16.9 0.491 1.479 0.491 1.479 copy_fm_to_dbcsr 174 11.2 0.001 0.002 0.959 1.478 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.464 mp_waitany 17072 13.8 1.215 1.422 1.215 1.422 qs_env_update_s_mstruct 11 6.9 0.002 0.007 1.193 1.318 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.228 1.237 acc_transpose_blocks 16440 15.4 0.073 0.075 1.210 1.224 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.794000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.818182, yerr=9.291425 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 727.937024E+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 0.023 0.059 64.259 64.260 qs_mol_dyn_low 1 2.0 0.003 0.004 63.822 63.831 qs_forces 11 3.9 0.003 0.006 63.759 63.760 qs_energies 11 4.9 0.003 0.012 60.389 60.393 scf_env_do_scf 11 5.9 0.001 0.002 51.923 51.926 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.252 40.252 velocity_verlet 10 3.0 0.001 0.001 36.439 36.442 dbcsr_multiply_generic 2055 12.4 0.116 0.119 29.426 29.655 qs_scf_new_mos 99 7.5 0.001 0.001 26.075 26.172 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.074 26.171 ot_scf_mini 99 9.5 0.003 0.004 24.459 24.557 multiply_cannon 2055 13.4 0.246 0.265 22.465 23.659 multiply_cannon_loop 2055 14.4 0.882 0.898 21.033 21.592 ot_mini 99 10.5 0.001 0.001 14.080 14.191 multiply_cannon_multrec 24660 15.4 4.206 6.946 12.684 13.826 rebuild_ks_matrix 110 8.3 0.000 0.000 11.796 11.902 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 11.796 11.901 init_scf_loop 11 6.9 0.001 0.003 11.631 11.632 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.430 10.522 qs_ot_get_derivative 99 11.5 0.001 0.001 9.930 10.028 prepare_preconditioner 11 7.9 0.000 0.000 9.907 9.925 make_preconditioner 11 8.9 0.000 0.001 9.907 9.925 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.122 9.608 dbcsr_mm_accdrv_process 52304 16.0 6.977 8.616 8.331 9.270 sum_up_and_integrate 110 10.3 0.067 0.070 6.238 6.251 mp_waitall_1 126806 16.4 4.391 6.217 4.391 6.217 integrate_v_rspace 110 11.3 0.003 0.003 6.171 6.186 make_m2s 4110 13.4 0.059 0.062 5.352 5.746 init_scf_run 11 5.9 0.000 0.001 5.702 5.702 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.702 5.702 make_images 4110 14.4 0.579 0.700 5.211 5.602 qs_ot_get_p 110 10.4 0.001 0.001 5.399 5.544 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.496 5.505 calculate_rho_elec 110 8.6 0.077 0.081 5.496 5.504 cp_fm_upper_to_full 70 13.8 3.326 4.734 3.326 4.734 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.023 4.114 apply_single 110 13.6 0.000 0.000 4.023 4.114 ot_diis_step 99 11.5 0.011 0.011 4.113 4.113 dbcsr_complete_redistribute 325 12.2 0.444 0.530 2.815 3.970 qs_ot_p2m_diag 48 11.0 0.055 0.063 3.648 3.662 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.282 3.425 grid_integrate_task_list 110 12.3 3.285 3.422 3.285 3.422 multiply_cannon_sync_h2d 24660 15.4 3.198 3.356 3.198 3.356 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.291 3.342 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.414 3.282 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.170 3.171 pw_transfer 1331 11.6 0.064 0.073 3.133 3.155 make_images_data 4110 15.4 0.048 0.060 2.695 3.095 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.300 3.086 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.006 3.059 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.028 3.053 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.977 2.978 calculate_dm_sparse 110 9.5 0.001 0.001 2.930 2.958 wfi_extrapolate 11 7.9 0.001 0.001 2.936 2.936 mp_alltoall_i22 605 13.7 1.730 2.930 1.730 2.930 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.777 2.906 density_rs2pw 110 9.6 0.004 0.004 2.668 2.807 cp_fm_cholesky_invert 11 10.9 2.695 2.704 2.695 2.704 calculate_first_density_matrix 1 7.0 0.004 0.017 2.681 2.683 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.646 2.649 cp_fm_redistribute_end 48 14.0 1.318 2.623 1.320 2.624 cp_fm_diag_elpa_base 48 14.0 1.221 2.491 1.301 2.604 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.575 2.600 rs_pw_transfer 902 11.9 0.010 0.011 2.357 2.554 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.463 2.506 grid_collocate_task_list 110 9.6 2.224 2.456 2.224 2.456 fft3d_ps 1111 14.6 1.061 1.096 2.331 2.355 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.266 2.267 potential_pw2rs 110 12.3 0.013 0.013 2.035 2.044 mp_alltoall_d11v 2046 13.8 1.736 1.973 1.736 1.973 jit_kernel_multiply 10 15.7 1.020 1.844 1.020 1.844 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.684 1.719 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.598 1.698 cp_fm_cholesky_decompose 22 10.9 1.642 1.687 1.642 1.687 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.611 1.623 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.563 mp_allgather_i34 2055 14.4 0.533 1.528 0.533 1.528 acc_transpose_blocks 24660 15.4 0.105 0.106 1.494 1.515 mp_sum_l 6514 12.8 0.953 1.512 0.953 1.512 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.839 1.495 mp_irecv_dv 62702 16.1 0.740 1.413 0.740 1.413 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.206 1.302 mp_waitany 13376 13.8 1.041 1.294 1.041 1.294 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.260000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.818182, yerr=7.505370 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 827.789312E+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 0.013 0.028 53.816 53.817 qs_mol_dyn_low 1 2.0 0.003 0.004 53.545 53.554 qs_forces 11 3.9 0.002 0.003 53.484 53.484 qs_energies 11 4.9 0.001 0.001 49.902 49.907 scf_env_do_scf 11 5.9 0.000 0.001 41.750 41.750 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 34.139 34.139 velocity_verlet 10 3.0 0.001 0.001 30.265 30.268 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.677 22.839 qs_scf_new_mos 99 7.5 0.001 0.001 20.178 20.236 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.177 20.236 ot_scf_mini 99 9.5 0.002 0.003 18.947 18.998 multiply_cannon 2055 13.4 0.247 0.257 17.341 18.707 multiply_cannon_loop 2055 14.4 0.323 0.335 15.994 16.298 rebuild_ks_matrix 110 8.3 0.000 0.000 11.387 11.430 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.386 11.430 ot_mini 99 10.5 0.001 0.001 10.314 10.362 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.149 10.189 multiply_cannon_multrec 8220 15.4 3.180 4.471 7.439 8.550 mp_waitall_1 106626 16.5 6.152 7.850 6.152 7.850 init_scf_loop 11 6.9 0.000 0.000 7.565 7.566 qs_ot_get_derivative 99 11.5 0.001 0.001 6.476 6.526 sum_up_and_integrate 110 10.3 0.079 0.082 6.140 6.152 integrate_v_rspace 110 11.3 0.003 0.003 6.060 6.073 prepare_preconditioner 11 7.9 0.000 0.000 5.952 5.960 make_preconditioner 11 8.9 0.000 0.000 5.952 5.960 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.530 5.599 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.511 5.523 calculate_rho_elec 110 8.6 0.115 0.115 5.510 5.522 init_scf_run 11 5.9 0.000 0.001 5.019 5.019 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.018 5.019 dbcsr_mm_accdrv_process 17442 15.9 2.734 3.765 4.128 5.017 qs_ot_get_p 110 10.4 0.001 0.001 4.719 4.787 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.164 4.512 make_m2s 4110 13.4 0.039 0.040 4.100 4.347 make_images 4110 14.4 0.639 0.696 3.970 4.219 ot_diis_step 99 11.5 0.012 0.012 3.819 3.819 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.735 3.780 apply_single 110 13.6 0.000 0.000 3.735 3.779 grid_integrate_task_list 110 12.3 3.372 3.493 3.372 3.493 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.420 3.424 pw_transfer 1331 11.6 0.065 0.070 3.174 3.179 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.125 3.125 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.067 3.075 multiply_cannon_sync_h2d 8220 15.4 2.901 3.060 2.901 3.060 cp_fm_cholesky_invert 11 10.9 2.876 2.880 2.876 2.880 density_rs2pw 110 9.6 0.004 0.004 2.579 2.775 make_images_data 4110 15.4 0.038 0.044 2.341 2.748 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.692 2.693 wfi_extrapolate 11 7.9 0.001 0.001 2.658 2.658 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 2.637 2.647 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.268 2.644 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.619 2.619 cp_fm_redistribute_end 48 14.0 0.660 2.593 0.664 2.594 cp_fm_diag_elpa_base 48 14.0 1.745 2.405 1.922 2.567 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.537 2.539 grid_collocate_task_list 110 9.6 2.323 2.529 2.323 2.529 calculate_dm_sparse 110 9.5 0.001 0.001 2.450 2.480 rs_pw_transfer 902 11.9 0.010 0.010 2.155 2.392 fft3d_ps 1111 14.6 1.108 1.120 2.325 2.336 calculate_first_density_matrix 1 7.0 0.000 0.000 2.274 2.276 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.196 2.213 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.984 potential_pw2rs 110 12.3 0.015 0.016 1.896 1.899 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.861 1.890 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.756 1.783 mp_alltoall_d11v 2046 13.8 1.600 1.729 1.600 1.729 cp_fm_cholesky_decompose 22 10.9 1.630 1.645 1.630 1.645 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.516 1.627 mp_allgather_i34 2055 14.4 0.535 1.603 0.535 1.603 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.567 1.573 jit_kernel_multiply 9 15.9 1.081 1.553 1.081 1.553 dbcsr_complete_redistribute 325 12.2 0.565 0.584 1.418 1.499 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.418 1.427 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.481 1.330 qs_create_task_list 11 7.9 0.000 0.000 1.220 1.317 generate_qs_task_list 11 8.9 0.379 0.446 1.220 1.317 mp_waitany 9240 13.8 1.040 1.299 1.040 1.299 mp_irecv_dv 24056 15.7 0.456 1.287 0.456 1.287 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.113 1.146 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=53.817000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.454545, yerr=9.048994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 1.385144E+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 0.017 0.037 85.466 85.466 qs_mol_dyn_low 1 2.0 0.003 0.004 85.147 85.156 qs_forces 11 3.9 0.002 0.002 85.082 85.083 qs_energies 11 4.9 0.003 0.004 81.045 81.048 scf_env_do_scf 11 5.9 0.000 0.001 71.002 71.003 velocity_verlet 10 3.0 0.001 0.001 54.929 54.934 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.215 42.216 dbcsr_multiply_generic 2055 12.4 0.125 0.131 28.596 28.735 init_scf_loop 11 6.9 0.000 0.000 28.715 28.716 prepare_preconditioner 11 7.9 0.000 0.000 26.771 26.783 make_preconditioner 11 8.9 0.000 0.000 26.771 26.783 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.817 26.236 qs_scf_new_mos 99 7.5 0.001 0.001 25.941 26.015 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.941 26.015 ot_scf_mini 99 9.5 0.002 0.002 24.196 24.262 multiply_cannon 2055 13.4 0.338 0.372 21.635 22.325 multiply_cannon_loop 2055 14.4 0.343 0.347 19.844 20.186 cp_fm_upper_to_full 70 14.2 13.014 18.687 13.014 18.687 ot_mini 99 10.5 0.001 0.001 13.488 13.552 rebuild_ks_matrix 110 8.3 0.001 0.001 13.091 13.170 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.090 13.170 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.905 11.977 dbcsr_complete_redistribute 325 12.2 1.037 1.075 7.549 10.781 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.518 9.765 multiply_cannon_multrec 8220 15.4 4.346 4.518 9.566 9.675 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.938 9.163 qs_ot_get_derivative 99 11.5 0.001 0.001 8.840 8.911 mp_alltoall_i22 605 13.7 5.568 8.824 5.568 8.824 mp_waitall_1 87304 16.6 7.704 8.683 7.704 8.683 sum_up_and_integrate 110 10.3 0.152 0.153 6.592 6.605 integrate_v_rspace 110 11.3 0.004 0.004 6.440 6.454 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.402 6.436 calculate_rho_elec 110 8.6 0.227 0.227 6.401 6.435 init_scf_run 11 5.9 0.000 0.001 5.800 5.800 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.799 5.800 make_m2s 4110 13.4 0.043 0.044 5.211 5.729 qs_ot_get_p 110 10.4 0.001 0.001 5.540 5.615 make_images 4110 14.4 0.879 0.929 5.024 5.542 dbcsr_mm_accdrv_process 11614 15.7 3.251 3.721 5.076 5.346 cp_fm_cholesky_invert 11 10.9 5.272 5.276 5.272 5.276 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.729 5.235 apply_single 110 13.6 0.000 0.000 4.729 5.235 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.746 4.981 ot_diis_step 99 11.5 0.015 0.016 4.624 4.624 qs_ot_p2m_diag 48 11.0 0.151 0.155 3.987 3.994 multiply_cannon_sync_h2d 8220 15.4 3.948 3.954 3.948 3.954 grid_integrate_task_list 110 12.3 3.655 3.714 3.655 3.714 pw_transfer 1331 11.6 0.073 0.073 3.692 3.703 qs_energies_init_hamiltonians 11 5.9 0.007 0.011 3.647 3.648 hybrid_alltoall_any 4261 16.3 0.256 0.550 2.868 3.627 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.576 3.587 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.548 3.549 make_images_data 4110 15.4 0.042 0.045 2.851 3.491 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.814 3.273 wfi_extrapolate 11 7.9 0.001 0.001 3.195 3.195 calculate_dm_sparse 110 9.5 0.001 0.001 3.146 3.170 fft_wrap_pw1pw2_140 451 13.1 0.215 0.217 3.106 3.119 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.052 3.057 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.953 2.954 cp_fm_diag_elpa_base 48 14.0 2.411 2.615 2.951 2.951 density_rs2pw 110 9.6 0.004 0.004 2.901 2.915 fft3d_ps 1111 14.6 1.269 1.277 2.787 2.799 grid_collocate_task_list 110 9.6 2.625 2.655 2.625 2.655 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.533 2.560 calculate_first_density_matrix 1 7.0 0.000 0.000 2.503 2.504 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.350 2.386 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.165 2.220 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.099 2.204 rs_pw_transfer 902 11.9 0.011 0.011 2.077 2.116 mp_alltoall_d11v 2046 13.8 2.038 2.076 2.038 2.076 potential_pw2rs 110 12.3 0.022 0.022 2.027 2.033 cp_fm_cholesky_decompose 22 10.9 1.990 2.008 1.990 2.008 qs_create_task_list 11 7.9 0.000 0.000 1.894 1.938 generate_qs_task_list 11 8.9 0.739 0.791 1.894 1.938 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.778 1.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.790 1.797 jit_kernel_multiply 10 15.5 1.621 1.763 1.621 1.763 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.696 1.713 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=85.466000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1220.181818, yerr=81.000357 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.402496E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 3606774. 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 0.075 0.271 202.595 202.607 qs_mol_dyn_low 1 2.0 0.007 0.034 201.644 201.659 qs_forces 11 3.9 0.010 0.037 201.539 201.543 qs_energies 11 4.9 0.003 0.012 196.022 196.048 scf_env_do_scf 11 5.9 0.001 0.001 179.367 179.371 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 158.937 158.939 dbcsr_multiply_generic 2507 12.6 0.175 0.179 123.879 124.529 velocity_verlet 10 3.0 0.001 0.001 121.480 121.481 qs_scf_new_mos 117 7.6 0.001 0.001 120.631 120.921 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.630 120.920 ot_scf_mini 117 9.6 0.003 0.004 114.078 114.295 multiply_cannon 2507 13.6 0.238 0.244 100.745 103.160 multiply_cannon_loop 2507 14.6 2.101 2.177 98.565 100.655 ot_mini 117 10.6 0.001 0.001 65.591 65.819 multiply_cannon_multrec 60168 15.6 33.269 34.958 41.677 43.311 qs_ot_get_derivative 117 11.6 0.001 0.001 40.558 40.821 rebuild_ks_matrix 128 8.3 0.001 0.001 33.034 33.461 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.034 33.460 mp_waitall_1 291448 16.2 28.245 31.164 28.245 31.164 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.731 30.120 multiply_cannon_sync_h2d 60168 15.6 27.357 29.163 27.357 29.163 qs_ot_get_p 128 10.4 0.001 0.001 26.699 27.093 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.447 25.438 apply_single 128 13.6 0.001 0.001 24.447 25.438 ot_diis_step 117 11.6 0.007 0.008 24.629 24.631 init_scf_loop 11 6.9 0.000 0.001 20.355 20.356 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.085 20.150 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.880 19.041 multiply_cannon_metrocomm3 60168 15.6 0.108 0.114 15.564 17.505 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.462 17.463 prepare_preconditioner 11 7.9 0.000 0.000 15.767 15.838 make_preconditioner 11 8.9 0.000 0.000 15.767 15.838 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.965 15.155 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.652 14.661 cp_fm_redistribute_end 83 14.4 11.551 14.588 11.563 14.590 cp_fm_diag_elpa_base 83 14.4 2.986 14.369 3.017 14.491 make_m2s 5014 13.6 0.103 0.113 13.692 14.263 make_images 5014 14.6 0.403 0.422 13.512 14.095 sum_up_and_integrate 128 10.3 0.089 0.106 13.775 13.792 integrate_v_rspace 128 11.3 0.003 0.004 13.685 13.704 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.417 12.478 calculate_rho_elec 128 8.7 0.045 0.063 12.416 12.477 init_scf_run 11 5.9 0.000 0.001 12.124 12.124 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.124 12.124 mp_sum_l 7870 13.0 8.288 9.519 8.288 9.519 wfi_extrapolate 11 7.9 0.001 0.001 8.919 8.919 cp_fm_cholesky_invert 11 10.9 8.908 8.916 8.908 8.916 dbcsr_mm_accdrv_process 124484 16.2 3.331 3.477 7.970 8.546 calculate_dm_sparse 128 9.5 0.001 0.001 8.373 8.506 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.783 7.947 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.729 7.812 make_images_data 5014 15.6 0.069 0.075 6.734 7.723 grid_integrate_task_list 128 12.3 7.078 7.450 7.078 7.450 multiply_cannon_metrocomm1 60168 15.6 0.085 0.089 5.970 7.358 hybrid_alltoall_any 5200 16.5 0.289 2.244 5.872 7.205 pw_transfer 1547 11.6 0.074 0.103 6.669 6.928 fft_wrap_pw1pw2 1291 12.7 0.009 0.012 6.465 6.701 density_rs2pw 128 9.7 0.006 0.007 6.255 6.655 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.005 6.592 6.601 rs_pw_transfer 1046 11.9 0.017 0.019 5.586 6.120 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.773 5.858 fft_wrap_pw1pw2_140 523 13.2 0.445 0.504 5.606 5.790 fft3d_ps 1291 14.7 2.090 2.565 5.258 5.439 mp_alltoall_d11v 2415 14.1 4.190 5.395 4.190 5.395 grid_collocate_task_list 128 9.7 4.722 5.092 4.722 5.092 cp_fm_cholesky_decompose 22 10.9 4.643 4.659 4.643 4.659 mp_sum_d 4459 12.1 3.711 4.407 3.711 4.407 potential_pw2rs 128 12.3 0.009 0.010 4.279 4.298 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=202.607000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.454545, yerr=7.177789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 830.595072E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56904. MP_Allreduce 11307 944. MP_Sync 170 MP_Alltoall 1983 6328514. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.135 190.049 190.049 qs_mol_dyn_low 1 2.0 0.003 0.013 189.255 189.405 qs_forces 11 3.9 0.010 0.012 189.172 189.173 qs_energies 11 4.9 0.006 0.008 182.484 182.500 scf_env_do_scf 11 5.9 0.001 0.001 166.345 166.355 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.651 133.654 velocity_verlet 10 3.0 0.001 0.001 119.573 119.574 dbcsr_multiply_generic 2527 12.6 0.186 0.193 97.873 99.164 qs_scf_new_mos 118 7.6 0.001 0.001 94.904 95.370 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.903 95.369 ot_scf_mini 118 9.6 0.004 0.004 90.061 90.606 multiply_cannon 2527 13.6 0.480 0.530 77.730 82.480 multiply_cannon_loop 2527 14.6 1.267 1.303 74.469 77.075 ot_mini 118 10.6 0.001 0.001 50.169 50.640 mp_waitall_1 228564 16.4 25.233 38.630 25.233 38.630 multiply_cannon_multrec 30324 15.6 21.949 27.629 31.652 37.569 rebuild_ks_matrix 129 8.3 0.001 0.001 32.834 33.397 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 32.833 33.396 init_scf_loop 11 6.9 0.000 0.000 32.605 32.606 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.632 30.123 qs_ot_get_derivative 118 11.6 0.001 0.002 27.915 28.459 prepare_preconditioner 11 7.9 0.000 0.000 28.290 28.364 make_preconditioner 11 8.9 0.000 0.000 28.290 28.364 multiply_cannon_metrocomm3 30324 15.6 0.095 0.101 16.084 28.146 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.995 27.526 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.274 23.446 apply_single 129 13.6 0.001 0.001 22.274 23.446 qs_ot_get_p 129 10.4 0.001 0.001 21.537 22.214 ot_diis_step 118 11.6 0.014 0.015 22.076 22.078 multiply_cannon_sync_h2d 30324 15.6 19.178 21.388 19.178 21.388 qs_ot_p2m_diag 83 11.4 0.187 0.216 16.742 16.779 cp_fm_cholesky_invert 11 10.9 16.468 16.480 16.468 16.480 make_m2s 5054 13.6 0.091 0.098 14.418 15.970 make_images 5054 14.6 1.170 1.356 14.208 15.764 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.593 15.595 sum_up_and_integrate 129 10.3 0.115 0.132 14.383 14.410 integrate_v_rspace 129 11.3 0.003 0.004 14.268 14.300 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.928 12.961 calculate_rho_elec 129 8.7 0.088 0.106 12.927 12.960 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.513 12.525 cp_fm_redistribute_end 83 14.4 7.278 12.460 7.287 12.462 cp_fm_diag_elpa_base 83 14.4 4.936 12.010 5.152 12.373 init_scf_run 11 5.9 0.000 0.001 11.419 11.420 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.418 11.420 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.916 11.306 multiply_cannon_metrocomm4 27797 15.6 0.097 0.111 3.857 11.063 mp_irecv_dv 70031 16.3 3.660 10.674 3.660 10.674 make_images_data 5054 15.6 0.067 0.075 8.740 10.553 hybrid_alltoall_any 5240 16.5 0.345 1.507 7.547 9.797 dbcsr_mm_accdrv_process 62734 16.2 4.666 5.399 9.160 9.756 wfi_extrapolate 11 7.9 0.001 0.001 8.268 8.268 pw_transfer 1559 11.6 0.085 0.097 7.790 7.860 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.565 7.626 grid_integrate_task_list 129 12.3 7.208 7.614 7.208 7.614 density_rs2pw 129 9.7 0.006 0.006 6.699 7.174 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.275 7.074 cp_fm_cholesky_decompose 22 10.9 6.858 6.937 6.858 6.937 fft_wrap_pw1pw2_140 527 13.2 0.473 0.522 6.613 6.684 calculate_dm_sparse 129 9.5 0.001 0.001 6.526 6.683 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.112 6.120 rs_pw_transfer 1054 12.0 0.014 0.016 5.503 5.982 fft3d_ps 1301 14.7 2.790 2.939 5.929 5.962 mp_sum_l 7930 13.1 4.026 5.952 4.026 5.952 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.404 5.477 grid_collocate_task_list 129 9.7 4.948 5.400 4.948 5.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.266 5.391 mp_allgather_i34 2527 14.6 1.837 5.043 1.837 5.043 mp_alltoall_d11v 2423 14.1 4.162 4.942 4.162 4.942 potential_pw2rs 129 12.3 0.015 0.016 4.774 4.796 mp_sum_d 4493 12.2 2.646 4.038 2.646 4.038 dbcsr_complete_redistribute 395 12.7 0.773 0.866 3.160 4.000 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.049000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.454545, yerr=3.447361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 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 11444707676160 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 92.796579E+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 6705500928 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 920.670208E+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 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4079 57277. MP_Allreduce 11236 986. 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 0.095 0.151 173.312 173.315 qs_mol_dyn_low 1 2.0 0.003 0.004 172.473 172.486 qs_forces 11 3.9 0.034 0.056 172.384 172.389 qs_energies 11 4.9 0.016 0.026 165.907 165.945 scf_env_do_scf 11 5.9 0.001 0.001 150.315 150.316 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.759 115.759 velocity_verlet 10 3.0 0.001 0.001 110.269 110.271 dbcsr_multiply_generic 2485 12.5 0.180 0.185 80.798 81.992 qs_scf_new_mos 116 7.6 0.001 0.001 79.779 80.067 qs_scf_loop_do_ot 116 8.6 0.001 0.002 79.778 80.066 ot_scf_mini 116 9.6 0.004 0.005 75.684 76.067 multiply_cannon 2485 13.5 0.494 0.516 61.242 65.370 multiply_cannon_loop 2485 14.5 0.856 0.885 58.099 60.917 ot_mini 116 10.6 0.001 0.001 42.012 42.390 mp_waitall_1 176908 16.5 24.886 34.598 24.886 34.598 init_scf_loop 11 6.9 0.001 0.004 34.457 34.458 prepare_preconditioner 11 7.9 0.000 0.000 30.456 30.508 make_preconditioner 11 8.9 0.000 0.001 30.456 30.508 rebuild_ks_matrix 127 8.3 0.001 0.001 29.705 30.252 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.704 30.251 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.122 29.512 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.812 27.321 multiply_cannon_multrec 19880 15.5 13.361 16.462 21.987 25.109 multiply_cannon_metrocomm3 19880 15.5 0.057 0.060 14.943 24.319 qs_ot_get_derivative 116 11.6 0.001 0.002 22.570 22.960 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.614 20.542 apply_single 127 13.6 0.001 0.001 19.614 20.541 qs_ot_get_p 127 10.4 0.001 0.001 19.199 19.620 ot_diis_step 116 11.6 0.017 0.018 19.341 19.342 multiply_cannon_sync_h2d 19880 15.5 14.281 16.094 14.281 16.094 make_m2s 4970 13.5 0.080 0.085 14.598 15.619 make_images 4970 14.5 1.166 1.289 14.367 15.385 qs_ot_p2m_diag 82 11.4 0.262 0.269 14.992 15.000 cp_fm_cholesky_invert 11 10.9 14.256 14.265 14.256 14.265 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.007 14.008 sum_up_and_integrate 127 10.3 0.130 0.142 13.894 13.915 integrate_v_rspace 127 11.3 0.003 0.004 13.764 13.789 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.828 12.866 calculate_rho_elec 127 8.7 0.131 0.145 12.827 12.865 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.056 11.056 cp_fm_redistribute_end 82 14.4 4.151 11.002 4.164 11.004 cp_fm_diag_elpa_base 82 14.4 6.417 10.442 6.821 10.926 make_images_data 4970 15.5 0.060 0.068 9.027 10.515 init_scf_run 11 5.9 0.001 0.001 10.506 10.506 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.505 10.506 hybrid_alltoall_any 5155 16.4 0.425 1.942 7.859 9.680 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.865 9.161 multiply_cannon_metrocomm4 17395 15.5 0.062 0.071 3.405 8.992 mp_irecv_dv 49801 16.2 3.282 8.744 3.282 8.744 dbcsr_mm_accdrv_process 41158 16.2 4.424 5.181 8.076 8.280 grid_integrate_task_list 127 12.3 7.297 7.769 7.297 7.769 pw_transfer 1535 11.6 0.084 0.103 7.615 7.719 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.394 7.503 wfi_extrapolate 11 7.9 0.001 0.001 7.371 7.371 cp_fm_upper_to_full 104 14.5 5.680 7.351 5.680 7.351 cp_fm_cholesky_decompose 22 10.9 7.143 7.172 7.143 7.172 density_rs2pw 127 9.7 0.006 0.006 6.359 6.697 fft_wrap_pw1pw2_140 519 13.2 0.475 0.532 6.506 6.623 dbcsr_complete_redistribute 393 12.7 1.169 1.201 4.416 6.127 calculate_dm_sparse 127 9.5 0.001 0.001 5.684 5.793 fft3d_ps 1281 14.7 2.672 2.873 5.675 5.755 grid_collocate_task_list 127 9.7 5.027 5.545 5.027 5.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.402 5.408 rs_pw_transfer 1038 11.9 0.013 0.014 4.915 5.270 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.535 5.211 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.263 4.976 mp_sum_l 7804 13.0 3.221 4.797 3.221 4.797 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.511 4.651 mp_allgather_i34 2485 14.5 1.693 4.604 1.693 4.604 mp_alltoall_d11v 2401 14.1 3.999 4.533 3.999 4.533 potential_pw2rs 127 12.3 0.020 0.023 4.363 4.382 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.315 4.006 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.955 3.982 mp_alltoall_i22 712 14.1 1.958 3.843 1.958 3.843 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 3.773 3.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.315000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=874.727273, yerr=6.876718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/18/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 4.387242E+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 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.138602E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509604. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.062 187.139 187.141 qs_mol_dyn_low 1 2.0 0.003 0.004 186.636 186.650 qs_forces 11 3.9 0.003 0.005 186.472 186.480 qs_energies 11 4.9 0.005 0.015 179.480 179.493 scf_env_do_scf 11 5.9 0.001 0.001 162.895 162.906 velocity_verlet 10 3.0 0.001 0.001 124.563 124.568 scf_env_do_scf_inner_loop 118 6.6 0.009 0.015 116.293 116.294 qs_scf_new_mos 118 7.6 0.001 0.001 80.550 80.921 qs_scf_loop_do_ot 118 8.6 0.001 0.002 80.549 80.920 dbcsr_multiply_generic 2529 12.6 0.201 0.206 79.904 80.808 ot_scf_mini 118 9.6 0.004 0.004 76.020 76.356 multiply_cannon 2529 13.6 0.570 0.608 55.311 58.782 multiply_cannon_loop 2529 14.6 1.192 1.220 51.331 53.464 init_scf_loop 11 6.9 0.001 0.004 46.481 46.482 ot_mini 118 10.6 0.001 0.001 42.508 42.840 prepare_preconditioner 11 7.9 0.000 0.000 42.423 42.456 make_preconditioner 11 8.9 0.000 0.001 42.423 42.456 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.997 41.161 multiply_cannon_multrec 30348 15.6 14.182 19.717 26.260 31.292 rebuild_ks_matrix 129 8.3 0.001 0.001 29.187 29.561 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.187 29.561 mp_waitall_1 155106 16.5 17.434 27.320 17.434 27.320 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.416 26.752 qs_ot_get_derivative 118 11.6 0.001 0.002 22.810 23.151 make_m2s 5058 13.6 0.097 0.101 20.220 21.505 make_images 5058 14.6 1.969 2.303 19.910 21.195 qs_ot_get_p 129 10.4 0.001 0.001 19.625 19.952 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.078 19.706 apply_single 129 13.6 0.001 0.001 19.077 19.705 ot_diis_step 118 11.6 0.018 0.019 19.579 19.580 cp_fm_upper_to_full 106 14.7 11.339 16.791 11.339 16.791 cp_fm_cholesky_invert 11 10.9 16.041 16.052 16.041 16.052 qs_ot_p2m_diag 84 11.4 0.347 0.395 15.492 15.545 multiply_cannon_metrocomm3 30348 15.6 0.047 0.049 6.452 15.170 cp_dbcsr_syevd 84 12.4 0.005 0.005 14.208 14.209 sum_up_and_integrate 129 10.3 0.141 0.153 14.087 14.112 integrate_v_rspace 129 11.3 0.004 0.004 13.945 13.975 multiply_cannon_sync_h2d 30348 15.6 11.858 13.410 11.858 13.410 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.127 13.165 calculate_rho_elec 129 8.7 0.176 0.192 13.126 13.165 dbcsr_complete_redistribute 397 12.7 1.512 1.630 9.017 12.889 make_images_data 5058 15.6 0.064 0.069 10.850 12.811 dbcsr_mm_accdrv_process 62780 16.2 7.495 8.532 11.649 12.173 hybrid_alltoall_any 5245 16.5 0.530 2.226 9.690 11.699 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.632 11.512 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.101 11.104 cp_fm_redistribute_end 84 14.4 1.889 11.037 1.902 11.043 cp_fm_diag_elpa_base 84 14.4 8.514 10.442 9.108 10.939 init_scf_run 11 5.9 0.000 0.001 10.638 10.640 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.638 10.640 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.403 10.163 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.494 9.748 mp_alltoall_i22 720 14.1 5.656 9.496 5.656 9.496 grid_integrate_task_list 129 12.3 7.574 8.016 7.574 8.016 pw_transfer 1559 11.6 0.086 0.099 7.926 7.998 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.699 7.775 cp_fm_cholesky_decompose 22 10.9 7.531 7.620 7.531 7.620 wfi_extrapolate 11 7.9 0.001 0.001 7.558 7.558 multiply_cannon_metrocomm4 25290 15.6 0.076 0.086 2.796 7.127 fft_wrap_pw1pw2_140 527 13.2 0.486 0.496 6.793 6.878 mp_irecv_dv 76751 16.2 2.649 6.853 2.649 6.853 density_rs2pw 129 9.7 0.006 0.006 6.253 6.619 calculate_dm_sparse 129 9.5 0.001 0.001 6.195 6.307 fft3d_ps 1301 14.7 2.813 2.881 5.915 5.976 grid_collocate_task_list 129 9.7 5.248 5.769 5.248 5.769 mp_alltoall_d11v 2429 14.1 4.728 5.518 4.728 5.518 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.464 5.518 rs_pw_transfer 1054 12.0 0.013 0.014 4.541 4.964 qs_energies_init_hamiltonians 11 5.9 0.006 0.012 4.475 4.477 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.373 4.459 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.346 4.442 potential_pw2rs 129 12.3 0.023 0.023 4.258 4.275 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.187 4.257 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.141000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.363636, yerr=19.685125 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.530368E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58308. MP_Allreduce 10985 1176. MP_Sync 86 MP_Alltoall 1700 18828162. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.035 169.707 169.708 qs_mol_dyn_low 1 2.0 0.003 0.003 169.286 169.298 qs_forces 11 3.9 0.003 0.003 168.916 168.923 qs_energies 11 4.9 0.002 0.002 161.603 161.610 scf_env_do_scf 11 5.9 0.001 0.001 144.539 144.552 velocity_verlet 10 3.0 0.001 0.001 111.423 111.443 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 108.545 108.546 dbcsr_multiply_generic 2485 12.5 0.180 0.184 73.708 74.061 qs_scf_new_mos 116 7.6 0.001 0.001 73.378 73.517 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.377 73.516 ot_scf_mini 116 9.6 0.003 0.004 69.012 69.078 multiply_cannon 2485 13.5 0.584 0.624 53.987 58.662 multiply_cannon_loop 2485 14.5 0.447 0.456 49.124 49.867 ot_mini 116 10.6 0.001 0.001 39.063 39.135 init_scf_loop 11 6.9 0.000 0.000 35.842 35.843 mp_waitall_1 128490 16.6 25.823 32.986 25.823 32.986 prepare_preconditioner 11 7.9 0.000 0.000 31.975 32.007 make_preconditioner 11 8.9 0.000 0.000 31.975 32.007 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.843 30.113 rebuild_ks_matrix 127 8.3 0.001 0.001 28.301 28.403 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 28.300 28.403 qs_ks_update_qs_env 127 7.6 0.001 0.001 25.789 25.883 multiply_cannon_multrec 9940 15.5 10.343 14.493 17.622 20.637 ot_diis_step 116 11.6 0.019 0.020 19.694 19.694 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.254 19.487 apply_single 127 13.6 0.001 0.001 19.254 19.486 qs_ot_get_derivative 116 11.6 0.002 0.002 19.303 19.379 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 11.949 18.931 make_m2s 4970 13.5 0.066 0.070 16.184 18.872 make_images 4970 14.5 2.311 2.626 15.881 18.573 cp_fm_cholesky_invert 11 10.9 18.295 18.300 18.295 18.300 qs_ot_get_p 127 10.4 0.001 0.001 16.961 17.054 sum_up_and_integrate 127 10.3 0.178 0.188 13.831 13.880 integrate_v_rspace 127 11.3 0.004 0.004 13.653 13.710 qs_ot_p2m_diag 82 11.4 0.490 0.495 13.284 13.300 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.981 13.036 calculate_rho_elec 127 8.7 0.256 0.267 12.981 13.036 make_images_data 4970 15.5 0.052 0.061 9.942 12.486 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.167 12.168 multiply_cannon_sync_h2d 9940 15.5 11.528 12.154 11.528 12.154 hybrid_alltoall_any 5155 16.4 0.809 3.816 9.740 12.056 init_scf_run 11 5.9 0.000 0.001 10.277 10.278 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.277 10.278 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.222 9.223 cp_fm_diag_elpa_base 82 14.4 8.981 9.053 9.215 9.215 grid_integrate_task_list 127 12.3 7.687 8.142 7.687 8.142 cp_fm_cholesky_decompose 22 10.9 8.008 8.123 8.008 8.123 mp_allgather_i34 2485 14.5 3.095 7.840 3.095 7.840 multiply_cannon_metrocomm1 9940 15.5 0.029 0.030 4.443 7.834 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.698 7.739 pw_transfer 1535 11.6 0.082 0.089 7.635 7.658 dbcsr_mm_accdrv_process 20590 16.1 2.669 3.436 6.912 7.580 fft_wrap_pw1pw2 1281 12.7 0.009 0.010 7.416 7.444 wfi_extrapolate 11 7.9 0.001 0.001 7.374 7.374 fft_wrap_pw1pw2_140 519 13.2 0.497 0.519 6.542 6.568 density_rs2pw 127 9.7 0.005 0.006 5.864 6.337 calculate_dm_sparse 127 9.5 0.001 0.001 6.019 6.090 grid_collocate_task_list 127 9.7 5.434 5.823 5.434 5.823 fft3d_ps 1281 14.7 2.699 2.772 5.592 5.618 dbcsr_complete_redistribute 393 12.7 2.132 2.184 5.182 5.569 mp_alltoall_d11v 2401 14.1 4.703 5.266 4.703 5.266 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.223 5.237 qs_energies_init_hamiltonians 11 5.9 0.007 0.012 5.230 5.231 rs_pw_transfer 1038 11.9 0.012 0.013 4.036 4.545 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.835 4.179 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.145 4.174 mp_irecv_dv 28618 15.9 1.798 4.108 1.798 4.108 potential_pw2rs 127 12.3 0.026 0.028 3.977 3.992 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.548 3.855 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.396 3.725 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.636 3.665 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.523 3.582 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.424 3.444 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.411 3.420 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.708000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1431.727273, yerr=57.877757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.977071E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4049 58480. MP_Allreduce 11119 1510. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.041 290.745 290.745 qs_mol_dyn_low 1 2.0 0.003 0.003 290.234 290.245 qs_forces 11 3.9 0.003 0.003 290.143 290.148 qs_energies 11 4.9 0.002 0.002 281.610 281.620 scf_env_do_scf 11 5.9 0.001 0.001 260.103 260.112 velocity_verlet 10 3.0 0.001 0.001 210.555 210.562 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 133.844 133.847 init_scf_loop 11 6.9 0.000 0.000 126.003 126.007 prepare_preconditioner 11 7.9 0.000 0.000 121.400 121.427 make_preconditioner 11 8.9 0.000 0.000 121.400 121.427 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.689 118.484 qs_scf_new_mos 118 7.6 0.001 0.001 91.943 92.040 qs_scf_loop_do_ot 118 8.6 0.001 0.001 91.942 92.039 ot_scf_mini 118 9.6 0.004 0.004 87.108 87.153 dbcsr_multiply_generic 2535 12.6 0.217 0.223 83.830 84.442 cp_fm_upper_to_full 106 14.8 54.553 78.265 54.553 78.265 multiply_cannon 2535 13.6 0.711 0.770 58.914 59.747 multiply_cannon_loop 2535 14.6 0.482 0.493 55.072 56.475 ot_mini 118 10.6 0.001 0.001 45.055 45.092 dbcsr_complete_redistribute 397 12.7 3.996 4.048 30.351 43.968 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.971 40.600 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.665 38.037 mp_alltoall_i22 720 14.1 22.501 36.188 22.501 36.188 rebuild_ks_matrix 129 8.3 0.001 0.001 33.367 33.422 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.366 33.421 cp_fm_cholesky_invert 11 10.9 33.247 33.253 33.247 33.253 mp_waitall_1 106626 16.7 28.549 32.364 28.549 32.364 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.200 31.256 qs_ot_get_p 129 10.4 0.001 0.001 26.431 26.512 qs_ot_get_derivative 118 11.6 0.001 0.002 24.267 24.315 qs_ot_p2m_diag 84 11.4 0.889 0.894 22.358 22.388 make_m2s 5070 13.6 0.077 0.079 20.269 21.433 make_images 5070 14.6 3.759 3.938 19.786 20.953 ot_diis_step 118 11.6 0.022 0.022 20.752 20.753 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.624 20.626 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.986 20.198 apply_single 129 13.6 0.001 0.001 19.986 20.198 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 18.665 19.923 multiply_cannon_multrec 10140 15.6 10.645 12.418 18.093 18.160 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.407 17.408 cp_fm_diag_elpa_base 84 14.4 12.817 14.542 17.403 17.403 multiply_cannon_sync_h2d 10140 15.6 15.776 15.794 15.776 15.794 sum_up_and_integrate 129 10.3 0.322 0.325 15.591 15.676 integrate_v_rspace 129 11.3 0.004 0.004 15.269 15.356 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.884 14.901 calculate_rho_elec 129 8.7 0.486 0.487 14.883 14.900 hybrid_alltoall_any 5257 16.5 1.312 3.071 11.167 13.128 make_images_data 5070 15.6 0.060 0.066 11.052 12.990 init_scf_run 11 5.9 0.000 0.001 11.577 11.577 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.576 11.577 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.513 9.558 dbcsr_mm_accdrv_process 20958 16.1 3.785 5.643 7.210 9.035 cp_fm_cholesky_decompose 22 10.9 8.805 8.872 8.805 8.872 grid_integrate_task_list 129 12.3 8.600 8.774 8.600 8.774 wfi_extrapolate 11 7.9 0.001 0.001 8.513 8.513 pw_transfer 1559 11.6 0.091 0.091 8.272 8.283 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.037 8.048 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.963 7.964 fft_wrap_pw1pw2_140 527 13.2 0.544 0.547 7.134 7.150 mp_alltoall_d11v 2429 14.1 6.740 6.886 6.740 6.886 calculate_dm_sparse 129 9.5 0.001 0.001 6.673 6.742 grid_collocate_task_list 129 9.7 6.386 6.419 6.386 6.419 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.245 6.302 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.043 6.121 fft3d_ps 1301 14.7 2.773 2.778 6.071 6.084 density_rs2pw 129 9.7 0.005 0.005 5.923 5.952 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=290.745000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2639.909091, yerr=179.887063 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 1.261269E+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 32651994. 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 0.025 0.040 84.967 84.969 qs_energies 1 2.0 0.000 0.001 84.444 84.462 ls_scf 1 3.0 0.000 0.001 83.396 83.412 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.426 72.625 multiply_cannon 111 7.7 0.017 0.020 55.726 57.014 multiply_cannon_loop 111 8.7 0.210 0.225 52.306 53.654 ls_scf_main 1 4.0 0.000 0.002 52.189 52.199 density_matrix_trs4 2 5.0 0.002 0.017 46.718 46.822 ls_scf_init_scf 1 4.0 0.013 0.053 28.179 28.195 ls_scf_init_matrix_S 1 5.0 0.000 0.003 27.081 27.164 mp_waitall_1 11316 10.9 22.245 25.771 22.245 25.771 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.873 24.892 multiply_cannon_multrec 2664 9.7 8.157 8.859 15.529 17.258 multiply_cannon_sync_h2d 2664 9.7 13.706 15.529 13.706 15.529 make_m2s 222 7.7 0.008 0.011 13.123 13.567 make_images 222 8.7 0.099 0.109 13.101 13.547 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.428 11.819 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.454 9.487 make_images_data 222 9.7 0.004 0.005 7.673 8.216 hybrid_alltoall_any 227 10.6 0.215 1.836 6.566 8.139 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.613 6.992 7.954 dbcsr_mm_accdrv_process_sort 4760 11.4 6.284 7.180 6.284 7.180 calculate_norms 4752 9.8 5.530 6.121 5.530 6.121 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.022 5.187 mp_sum_l 807 5.4 3.049 4.554 3.049 4.554 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.059 4.248 mp_irecv_dv 6231 10.9 2.041 4.214 2.041 4.214 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.249 3.595 make_images_sizes 222 9.7 0.000 0.000 0.729 3.463 mp_alltoall_i44 222 10.7 0.728 3.462 0.728 3.462 arnoldi_extremal 4 6.8 0.000 0.000 3.205 3.228 arnoldi_normal_ev 4 7.8 0.001 0.004 3.205 3.228 build_subspace 16 8.4 0.009 0.012 3.109 3.110 ls_scf_post 1 4.0 0.004 0.022 3.028 3.046 ls_scf_store_result 1 5.0 0.000 0.000 2.833 2.884 dbcsr_special_finalize 555 9.7 0.005 0.006 2.367 2.777 dbcsr_merge_single_wm 555 10.7 0.455 0.584 2.359 2.769 make_images_pack 222 9.7 2.208 2.636 2.210 2.638 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.309 2.567 dbcsr_sort_data 658 11.4 2.161 2.522 2.161 2.522 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.475 2.067 2.477 ls_scf_dm_to_ks 2 5.0 0.001 0.026 2.229 2.330 buffer_matrices_ensure_size 222 8.7 1.754 2.078 1.754 2.078 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.762 1.764 rebuild_ks_matrix 3 7.3 0.000 0.000 1.753 1.755 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.753 1.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.969000, 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/a3b6180a81f48dbca65be969b5f13520da8e081a_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 2.110960E+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 0.032 0.052 91.811 91.812 qs_energies 1 2.0 0.000 0.000 90.901 90.906 ls_scf 1 3.0 0.000 0.000 89.573 89.579 dbcsr_multiply_generic 111 6.7 0.015 0.015 75.625 75.934 multiply_cannon 111 7.7 0.028 0.041 53.341 56.620 ls_scf_main 1 4.0 0.000 0.000 55.180 55.181 multiply_cannon_loop 111 8.7 0.116 0.125 50.048 53.418 density_matrix_trs4 2 5.0 0.002 0.003 49.512 49.663 ls_scf_init_scf 1 4.0 0.000 0.002 30.838 30.840 mp_waitall_1 9246 10.9 21.593 30.466 21.593 30.466 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.712 29.809 multiply_cannon_multrec 1332 9.7 13.000 17.097 22.281 27.561 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.270 27.293 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.985 21.092 make_m2s 222 7.7 0.006 0.007 15.500 16.143 make_images 222 8.7 1.578 1.919 15.470 16.114 dbcsr_mm_accdrv_process 4041 10.4 0.257 0.436 8.891 10.421 dbcsr_mm_accdrv_process_sort 4041 11.4 8.486 9.992 8.486 9.992 make_images_data 222 9.7 0.004 0.004 8.976 9.902 hybrid_alltoall_any 227 10.6 0.519 2.449 8.457 9.396 mp_sum_l 807 5.4 5.491 8.654 5.491 8.654 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.253 7.812 mp_irecv_dv 3311 11.0 3.233 7.763 3.233 7.763 calculate_norms 2376 9.8 5.999 6.840 5.999 6.840 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.273 6.774 multiply_cannon_sync_h2d 1332 9.7 4.705 6.078 4.705 6.078 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.038 5.225 arnoldi_extremal 4 6.8 0.000 0.000 4.717 4.748 arnoldi_normal_ev 4 7.8 0.001 0.005 4.717 4.748 build_subspace 16 8.4 0.014 0.021 4.457 4.459 ls_scf_post 1 4.0 0.000 0.000 3.555 3.559 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.212 3.467 ls_scf_store_result 1 5.0 0.000 0.000 3.257 3.378 dbcsr_matrix_vector_mult_local 304 10.0 2.781 3.274 2.783 3.276 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.255 2.680 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.532 2.638 make_images_pack 222 9.7 2.028 2.414 2.030 2.416 mp_allgather_i34 111 8.7 0.998 2.389 0.998 2.389 dbcsr_sort_data 436 11.2 1.848 2.051 1.848 2.051 dbcsr_data_new 4174 10.1 1.609 1.845 1.609 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.812000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1749.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 2.706338E+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 0.038 0.057 94.791 94.793 qs_energies 1 2.0 0.000 0.000 93.926 93.929 ls_scf 1 3.0 0.001 0.003 92.300 92.303 dbcsr_multiply_generic 111 6.7 0.036 0.172 77.019 77.315 ls_scf_main 1 4.0 0.000 0.000 57.737 57.741 multiply_cannon 111 7.7 0.046 0.123 53.216 57.681 multiply_cannon_loop 111 8.7 0.100 0.105 49.592 53.237 density_matrix_trs4 2 5.0 0.002 0.003 51.774 51.937 mp_waitall_1 7374 11.0 24.157 34.566 24.157 34.566 ls_scf_init_scf 1 4.0 0.001 0.004 30.916 30.919 ls_scf_init_matrix_S 1 5.0 0.001 0.006 29.795 29.862 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.441 27.459 multiply_cannon_multrec 888 9.7 12.566 15.021 21.154 24.193 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.431 23.731 make_m2s 222 7.7 0.006 0.008 17.135 18.347 make_images 222 8.7 1.977 2.292 17.097 18.309 make_images_data 222 9.7 0.004 0.009 9.777 10.785 hybrid_alltoall_any 227 10.6 0.622 2.867 9.485 10.756 dbcsr_mm_accdrv_process 3754 10.4 0.234 0.415 8.121 9.345 dbcsr_mm_accdrv_process_sort 3754 11.4 7.751 8.930 7.751 8.930 mp_sum_l 807 5.4 5.081 8.758 5.081 8.758 multiply_cannon_sync_h2d 888 9.7 6.086 7.257 6.086 7.257 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.895 6.883 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.570 6.734 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.465 6.438 mp_irecv_dv 2335 11.1 2.449 6.388 2.449 6.388 arnoldi_extremal 4 6.8 0.000 0.000 5.149 5.171 arnoldi_normal_ev 4 7.8 0.001 0.005 5.149 5.171 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.964 5.171 build_subspace 16 8.4 0.014 0.020 4.825 4.831 calculate_norms 1584 9.8 4.271 4.618 4.271 4.618 mp_allgather_i34 111 8.7 1.455 3.832 1.455 3.832 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.524 3.827 ls_scf_post 1 4.0 0.000 0.001 3.646 3.649 dbcsr_matrix_vector_mult_local 304 10.0 3.065 3.639 3.067 3.641 ls_scf_store_result 1 5.0 0.000 0.000 3.353 3.469 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.801 2.920 dbcsr_sort_data 325 11.1 1.909 2.135 1.909 2.135 make_images_sizes 222 9.7 0.000 0.000 1.017 2.131 mp_alltoall_i44 222 10.7 1.017 2.130 1.017 2.130 make_images_pack 222 9.7 1.812 2.122 1.815 2.125 dbcsr_data_release 9322 10.9 1.300 1.921 1.300 1.921 dbcsr_finalize 304 7.8 0.026 0.032 1.618 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.793000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2165.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 3.339710E+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 0.032 0.057 97.394 97.395 qs_energies 1 2.0 0.000 0.000 96.787 96.793 ls_scf 1 3.0 0.000 0.000 95.066 95.070 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.853 79.103 ls_scf_main 1 4.0 0.000 0.000 59.164 59.168 multiply_cannon 111 7.7 0.054 0.123 51.768 56.456 density_matrix_trs4 2 5.0 0.002 0.003 53.081 53.233 multiply_cannon_loop 111 8.7 0.114 0.125 46.683 49.347 ls_scf_init_scf 1 4.0 0.000 0.000 32.639 32.640 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.476 31.566 mp_waitall_1 6438 11.0 22.927 29.252 22.927 29.252 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.001 29.014 multiply_cannon_multrec 1332 9.7 14.217 17.255 22.127 25.142 make_m2s 222 7.7 0.006 0.008 21.251 22.686 make_images 222 8.7 3.140 3.601 21.201 22.638 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.310 17.250 make_images_data 222 9.7 0.004 0.004 11.942 13.592 hybrid_alltoall_any 227 10.6 0.798 3.838 11.311 13.285 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.412 7.555 9.134 dbcsr_mm_accdrv_process_sort 3641 11.4 7.182 8.710 7.182 8.710 mp_sum_l 807 5.4 4.256 7.737 4.256 7.737 multiply_cannon_sync_h2d 1332 9.7 5.551 6.202 5.551 6.202 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.243 6.124 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.077 6.058 mp_irecv_dv 3229 10.9 2.056 6.013 2.056 6.013 arnoldi_extremal 4 6.8 0.000 0.000 5.411 5.423 arnoldi_normal_ev 4 7.8 0.001 0.004 5.411 5.423 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.498 5.221 build_subspace 16 8.4 0.014 0.021 5.061 5.069 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.593 4.800 mp_allgather_i34 111 8.7 2.209 4.692 2.209 4.692 calculate_norms 2376 9.8 4.186 4.560 4.186 4.560 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.670 4.017 dbcsr_matrix_vector_mult_local 304 10.0 3.229 3.752 3.231 3.754 dbcsr_sort_data 658 11.4 3.109 3.376 3.109 3.376 ls_scf_post 1 4.0 0.000 0.000 3.263 3.269 dbcsr_special_finalize 555 9.7 0.006 0.007 2.857 3.203 dbcsr_merge_single_wm 555 10.7 0.538 0.659 2.848 3.195 ls_scf_store_result 1 5.0 0.000 0.000 3.019 3.065 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.911 2.974 dbcsr_data_release 10477 10.7 1.577 2.394 1.577 2.394 dbcsr_finalize 304 7.8 0.049 0.061 1.803 1.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.395000, 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/a3b6180a81f48dbca65be969b5f13520da8e081a_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 4.629840E+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 0.041 0.056 99.493 99.493 qs_energies 1 2.0 0.000 0.000 98.683 98.688 ls_scf 1 3.0 0.000 0.000 96.765 96.770 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.403 78.623 ls_scf_main 1 4.0 0.000 0.000 62.106 62.107 multiply_cannon 111 7.7 0.093 0.194 55.739 61.034 density_matrix_trs4 2 5.0 0.002 0.003 55.068 55.156 multiply_cannon_loop 111 8.7 0.069 0.076 51.182 52.735 mp_waitall_1 5481 11.0 26.667 32.108 26.667 32.108 ls_scf_init_scf 1 4.0 0.000 0.000 31.060 31.061 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.877 29.918 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.467 27.479 multiply_cannon_multrec 444 9.7 14.117 16.568 21.312 23.781 make_m2s 222 7.7 0.004 0.005 17.714 20.244 make_images 222 8.7 3.724 4.426 17.653 20.184 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.222 16.585 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.305 15.861 make_images_data 222 9.7 0.003 0.004 9.933 12.478 hybrid_alltoall_any 227 10.6 0.787 3.773 9.745 12.286 dbcsr_mm_accdrv_process 3003 10.4 0.197 0.485 6.896 8.089 multiply_cannon_sync_h2d 444 9.7 6.516 7.834 6.516 7.834 dbcsr_mm_accdrv_process_sort 3003 11.4 6.486 7.579 6.486 7.579 mp_allgather_i34 111 8.7 2.792 6.965 2.792 6.965 arnoldi_extremal 4 6.8 0.000 0.000 5.952 5.966 arnoldi_normal_ev 4 7.8 0.002 0.005 5.952 5.966 build_subspace 16 8.4 0.015 0.019 5.557 5.570 mp_sum_l 807 5.4 2.995 5.331 2.995 5.331 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.901 5.043 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.259 4.447 dbcsr_matrix_vector_mult_local 304 10.0 3.741 4.227 3.743 4.229 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.984 3.989 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.625 3.927 mp_irecv_dv 1241 11.2 1.605 3.907 1.605 3.907 calculate_norms 792 9.8 3.564 3.756 3.564 3.756 ls_scf_post 1 4.0 0.000 0.000 3.600 3.605 make_images_sizes 222 9.7 0.000 0.000 1.139 3.588 mp_alltoall_i44 222 10.7 1.139 3.587 1.139 3.587 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.380 3.495 ls_scf_store_result 1 5.0 0.000 0.000 3.378 3.428 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.314 dbcsr_merge_all 275 8.9 0.473 0.527 2.050 2.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.493000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3617.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/a3b6180a81f48dbca65be969b5f13520da8e081a_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 8.732869E+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 0.083 0.155 108.119 108.119 qs_energies 1 2.0 0.000 0.000 106.643 106.658 ls_scf 1 3.0 0.000 0.000 103.776 103.791 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.721 77.861 ls_scf_main 1 4.0 0.000 0.000 66.018 66.019 density_matrix_trs4 2 5.0 0.002 0.003 57.179 57.251 multiply_cannon 111 7.7 0.117 0.175 50.001 51.711 multiply_cannon_loop 111 8.7 0.067 0.069 46.526 47.920 ls_scf_init_scf 1 4.0 0.000 0.000 34.095 34.095 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.764 32.776 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.942 29.949 mp_waitall_1 4569 11.1 22.258 26.433 22.258 26.433 make_m2s 222 7.7 0.005 0.005 23.857 24.838 make_images 222 8.7 4.604 5.021 23.751 24.730 multiply_cannon_multrec 444 9.7 17.829 18.572 22.490 23.296 make_images_data 222 9.7 0.003 0.003 13.080 15.491 hybrid_alltoall_any 227 10.6 1.663 3.633 12.904 15.475 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.502 11.380 multiply_cannon_sync_h2d 444 9.7 8.855 8.890 8.855 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.528 7.538 arnoldi_normal_ev 4 7.8 0.003 0.009 7.528 7.538 build_subspace 16 8.4 0.026 0.036 6.965 6.977 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.608 5.755 dbcsr_matrix_vector_mult_local 304 10.0 5.164 5.484 5.166 5.486 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.979 5.231 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.824 4.913 dbcsr_mm_accdrv_process 1814 10.4 0.171 0.360 4.490 4.623 dbcsr_mm_accdrv_process_sort 1814 11.4 4.171 4.304 4.171 4.304 ls_scf_post 1 4.0 0.000 0.000 3.663 3.677 mp_allgather_i34 111 8.7 1.124 3.538 1.124 3.538 make_images_sizes 222 9.7 0.000 0.000 1.432 3.486 mp_alltoall_i44 222 10.7 1.432 3.486 1.432 3.486 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.430 calculate_norms 792 9.8 3.238 3.272 3.238 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.153 dbcsr_merge_all 275 8.9 0.893 0.923 2.874 2.933 dbcsr_complete_redistribute 5 7.6 1.438 1.472 2.787 2.913 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.837 2.837 dbcsr_data_release 12724 10.6 2.319 2.829 2.319 2.829 mp_sum_l 807 5.4 1.553 2.683 1.553 2.683 matrix_ls_to_qs 2 6.0 0.000 0.000 2.439 2.585 dbcsr_sort_data 325 11.1 2.444 2.519 2.444 2.519 dbcsr_new_transposed 4 7.5 0.243 0.252 2.339 2.355 dbcsr_frobenius_norm 74 6.6 2.056 2.138 2.217 2.269 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.208 dbcsr_add_anytype 103 7.2 0.859 0.892 2.134 2.208 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.182 2.184 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.119000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6919.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: a3b6180a81f48dbca65be969b5f13520da8e081a Summary: empty Status: OK