=== This is the CP2K Performance-Test === Updating 979a205f9..92831d4de Fast-forward tests/QS/regtest-gw-cubic/TEST_FILES | 2 +- tools/docker/Dockerfile.test_hip_cuda_A100 | 4 ++++ tools/docker/Dockerfile.test_hip_cuda_P100 | 4 ++++ tools/docker/Dockerfile.test_hip_cuda_V100 | 4 ++++ tools/docker/generate_dockerfiles.py | 4 ++++ 5 files changed, 17 insertions(+), 1 deletion(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 92831d4de10a6dd11ff469e14bc401d403fbb9d7 ################# 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/01 job id: 43111562 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/02 job id: 43111563 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/03 job id: 43111564 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/04 job id: 43111565 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/05 job id: 43111566 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/06 job id: 43111567 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/07 job id: 43111569 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/08 job id: 43111570 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/09 job id: 43111571 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/10 job id: 43111572 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/11 job id: 43111573 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/12 job id: 43111574 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/13 job id: 43111576 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/14 job id: 43111577 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/15 job id: 43111578 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/16 job id: 43111580 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/17 job id: 43111581 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/18 job id: 43111582 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/19 job id: 43111583 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/20 job id: 43111584 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/21 job id: 43111585 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/22 job id: 43111586 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/23 job id: 43111587 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/24 job id: 43111588 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/25 job id: 43111589 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/26 job id: 43111590 --- 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/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.035 132.128 132.129 farming_run 1 2.0 131.476 131.481 132.102 132.106 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.455120E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.024 114.894 114.894 qs_energies 1 2.0 0.000 0.000 114.706 114.707 mp2_main 1 3.0 0.000 0.000 112.887 112.888 mp2_gpw_main 1 4.0 0.019 0.026 112.097 112.098 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 93.297 93.795 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.534 56.030 mp2_eri_3c_integrate_gpw 272 7.0 0.156 0.171 41.861 47.286 get_2c_integrals 1 6.0 0.000 0.000 36.985 37.592 integrate_v_rspace 273 8.0 0.437 0.453 25.173 30.300 pw_transfer 6555 10.6 0.377 0.386 27.405 28.014 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.069 26.695 grid_integrate_task_list 273 9.0 20.976 26.608 20.976 26.608 fft_wrap_pw1pw2_100 2178 12.4 1.186 1.302 23.634 24.262 compute_2c_integrals 1 7.0 0.002 0.002 19.312 19.313 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.884 19.028 mp2_eri_2c_integrate_gpw 1 9.0 2.387 2.436 18.881 19.025 rpa_ri_compute_en 1 5.0 0.000 0.000 18.692 18.943 cp_fm_cholesky_decompose 12 8.2 17.610 18.201 17.610 18.201 cholesky_decomp 1 7.0 0.000 0.000 16.518 17.116 fft3d_s 5443 13.4 16.198 16.781 16.220 16.802 ao_to_mo_and_store_B_mult_1 272 7.0 10.847 15.563 10.847 15.563 calculate_wavefunction 272 8.0 5.450 5.591 12.604 13.167 rpa_num_int 1 6.0 0.000 0.000 10.541 10.541 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.519 10.538 calc_mat_Q 8 8.0 0.000 0.000 9.381 9.488 contract_S_to_Q 8 9.0 0.000 0.000 8.799 8.905 calc_potential_gpw 544 9.5 0.006 0.006 8.246 8.642 parallel_gemm_fm 14 9.1 0.000 0.000 8.385 8.474 parallel_gemm_fm_cosma 14 10.1 8.385 8.474 8.385 8.474 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.217 8.433 potential_pw2rs 545 10.0 0.107 0.108 7.707 8.376 collocate_single_gaussian 272 10.0 0.040 0.042 7.460 7.633 create_integ_mat 1 6.0 0.006 0.008 7.558 7.558 array2fm 1 7.0 0.000 0.000 6.577 7.012 pw_scatter_s 2720 13.7 4.437 4.715 4.437 4.715 pw_gather_s 2722 13.2 3.881 4.339 3.881 4.339 array2fm_buffer_send 1 8.0 2.910 3.179 2.910 3.179 pw_poisson_solve 545 10.5 1.127 1.183 2.167 2.335 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.096985, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 0.038 394.898 394.899 farming_run 1 2.0 394.065 394.071 394.859 394.861 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.206727E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 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 0.049 209.589 209.590 qs_energies 1 2.0 0.000 0.000 209.342 209.355 scf_env_do_scf 1 3.0 0.000 0.000 106.553 106.554 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.699 105.707 rebuild_ks_matrix 4 6.0 0.000 0.000 105.697 105.705 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 105.697 105.705 hfx_ks_matrix 4 8.0 0.001 0.001 105.322 105.325 integrate_four_center 4 9.0 0.144 0.457 105.321 105.324 mp2_main 1 3.0 0.000 0.000 102.504 102.517 mp2_gpw_main 1 4.0 0.031 0.045 101.631 101.643 integrate_four_center_main 4 10.0 0.096 0.605 96.537 99.377 integrate_four_center_bin 262 11.0 96.441 99.372 96.441 99.372 init_scf_loop 1 4.0 0.000 0.000 92.395 92.395 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.826 75.854 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.414 55.441 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 41.963 47.276 integrate_v_rspace 95 8.0 0.395 0.572 28.341 33.461 pw_transfer 2240 10.6 0.146 0.163 29.831 30.253 fft_wrap_pw1pw2 1868 11.4 0.017 0.020 28.852 29.300 grid_integrate_task_list 95 9.0 23.628 28.948 23.628 28.948 mp2_ri_gpw_compute_en 1 5.0 0.061 0.074 26.659 28.387 fft_wrap_pw1pw2_100 730 12.4 1.261 1.423 26.591 26.975 ao_to_mo_and_store_B_mult_1 91 7.0 10.778 26.528 10.778 26.528 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.899 24.989 25.000 get_2c_integrals 1 6.0 0.000 0.000 20.305 20.347 compute_2c_integrals 1 7.0 0.002 0.003 19.293 19.295 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.805 19.167 mp2_eri_2c_integrate_gpw 1 9.0 1.761 1.941 18.803 19.166 fft3d_s 1823 13.4 18.380 18.665 18.393 18.678 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.156 14.156 calculate_wavefunction 91 8.0 2.022 2.062 9.742 10.012 mp2_ri_gpw_compute_en_expansio 172 7.0 0.555 0.588 8.859 9.563 potential_pw2rs 186 10.0 0.033 0.035 8.598 9.162 local_gemm 172 8.0 8.305 8.999 8.305 8.999 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.200 8.519 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.533 7.906 8.246 calc_potential_gpw 182 9.5 0.002 0.002 7.907 8.158 collocate_single_gaussian 91 10.0 0.017 0.024 7.843 8.020 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.420 6.346 6.420 mp_sendrecv_dm3 2068 8.0 5.926 6.252 5.926 6.252 mp_sync 38 10.4 3.474 6.023 3.474 6.023 pw_gather_s 912 13.2 4.910 5.344 4.910 5.344 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.629724, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1493.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 451.121152E+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 1697096. 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 0.014 0.040 52.229 52.230 qs_mol_dyn_low 1 2.0 0.003 0.005 51.953 51.962 qs_forces 11 3.9 0.004 0.022 51.881 51.881 qs_energies 11 4.9 0.003 0.009 50.425 50.437 scf_env_do_scf 11 5.9 0.000 0.001 44.297 44.298 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.193 42.194 dbcsr_multiply_generic 2286 12.5 0.093 0.096 33.957 34.474 qs_scf_new_mos 108 7.5 0.000 0.001 32.338 32.638 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.337 32.637 ot_scf_mini 108 9.5 0.002 0.003 30.665 30.861 multiply_cannon 2286 13.5 0.186 0.193 26.292 27.953 multiply_cannon_loop 2286 14.5 1.508 1.578 25.596 27.249 velocity_verlet 10 3.0 0.002 0.007 25.834 25.834 ot_mini 108 10.5 0.001 0.001 19.613 19.866 qs_ot_get_derivative 108 11.5 0.001 0.002 16.565 16.774 mp_waitall_1 267858 16.1 8.825 15.367 8.825 15.367 multiply_cannon_metrocomm3 54864 15.5 0.066 0.073 6.118 13.502 multiply_cannon_multrec 54864 15.5 4.211 6.570 7.783 11.230 rebuild_ks_matrix 119 8.3 0.000 0.000 7.797 7.942 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.016 7.796 7.941 multiply_cannon_sync_h2d 54864 15.5 5.860 7.613 5.860 7.613 mp_sum_l 7207 12.9 5.733 7.399 5.733 7.399 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.875 7.010 qs_ot_get_p 119 10.4 0.001 0.002 6.431 6.757 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.518 5.989 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.424 5.538 init_scf_run 11 5.9 0.000 0.001 4.885 4.885 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.885 4.885 dbcsr_mm_accdrv_process 76910 16.1 1.184 1.831 3.495 4.846 sum_up_and_integrate 119 10.3 0.013 0.021 4.460 4.467 integrate_v_rspace 119 11.3 0.002 0.003 4.447 4.456 qs_rho_update_rho_low 119 7.7 0.001 0.004 4.028 4.138 calculate_rho_elec 119 8.7 0.011 0.017 4.028 4.134 qs_ot_p2m_diag 50 11.0 0.005 0.012 3.480 3.568 multiply_cannon_metrocomm1 54864 15.5 0.051 0.056 1.849 3.499 calculate_dm_sparse 119 9.5 0.001 0.001 3.007 3.155 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.905 3.110 apply_single 119 13.6 0.000 0.000 2.905 3.110 jit_kernel_multiply 13 15.8 2.248 2.938 2.248 2.938 rs_pw_transfer 974 11.9 0.011 0.012 2.737 2.846 calculate_first_density_matrix 1 7.0 0.000 0.002 2.744 2.748 ot_diis_step 108 11.5 0.006 0.009 2.708 2.708 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.583 2.584 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.417 2.515 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.378 2.380 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.379 2.380 cp_fm_redistribute_end 50 14.0 2.159 2.357 2.163 2.358 cp_fm_diag_elpa_base 50 14.0 0.194 2.325 0.194 2.334 density_rs2pw 119 9.7 0.004 0.004 2.138 2.252 acc_transpose_blocks 54864 15.5 0.232 0.247 1.780 2.216 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.131 2.174 grid_integrate_task_list 119 12.3 2.019 2.103 2.019 2.103 mp_sum_d 4125 12.0 1.473 2.099 1.473 2.099 init_scf_loop 11 6.9 0.000 0.002 2.086 2.088 wfi_extrapolate 11 7.9 0.001 0.002 2.073 2.073 potential_pw2rs 119 12.3 0.004 0.004 1.834 1.850 pw_transfer 1439 11.6 0.051 0.056 1.614 1.692 make_m2s 4572 13.5 0.053 0.055 1.622 1.666 fft_wrap_pw1pw2 1201 12.6 0.006 0.007 1.539 1.620 make_images 4572 14.5 0.133 0.139 1.541 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.495 1.512 mp_alltoall_d11v 2130 13.8 1.286 1.439 1.286 1.439 mp_waitany 12084 13.8 1.249 1.427 1.249 1.427 fft3d_ps 1201 14.6 0.359 0.465 1.318 1.392 grid_collocate_task_list 119 9.7 1.292 1.367 1.292 1.367 dbcsr_dot_sd 1205 11.9 0.047 0.057 0.827 1.267 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.186 1.266 acc_transpose_blocks_kernels 54864 16.5 0.253 0.384 0.787 1.067 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.230000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.096320E+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 1031941. 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 0.028 0.054 37.838 37.839 qs_mol_dyn_low 1 2.0 0.003 0.003 37.589 37.596 qs_forces 11 3.9 0.002 0.002 37.531 37.532 qs_energies 11 4.9 0.008 0.013 35.866 35.868 scf_env_do_scf 11 5.9 0.000 0.001 30.696 30.696 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.249 28.249 dbcsr_multiply_generic 2286 12.5 0.099 0.105 21.099 21.463 qs_scf_new_mos 108 7.5 0.001 0.001 19.499 19.750 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.498 19.749 ot_scf_mini 108 9.5 0.002 0.003 18.606 18.787 velocity_verlet 10 3.0 0.001 0.001 17.893 17.894 multiply_cannon 2286 13.5 0.208 0.216 16.192 17.843 multiply_cannon_loop 2286 14.5 0.901 0.968 15.099 16.496 ot_mini 108 10.5 0.001 0.001 11.631 11.871 mp_waitall_1 217478 16.2 5.828 10.899 5.828 10.899 qs_ot_get_derivative 108 11.5 0.001 0.001 9.182 9.363 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.109 9.347 multiply_cannon_multrec 27432 15.5 1.980 4.582 5.852 8.826 rebuild_ks_matrix 119 8.3 0.000 0.000 7.015 7.156 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.015 7.156 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.198 6.327 dbcsr_mm_accdrv_process 47894 16.0 3.026 5.178 3.802 5.623 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.539 4.386 qs_ot_get_p 119 10.4 0.001 0.001 4.107 4.330 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.043 4.122 apply_single 119 13.6 0.000 0.000 3.043 4.122 sum_up_and_integrate 119 10.3 0.025 0.027 4.065 4.070 integrate_v_rspace 119 11.3 0.002 0.002 4.040 4.047 mp_sum_l 7207 12.9 2.038 4.021 2.038 4.021 init_scf_run 11 5.9 0.000 0.001 3.857 3.857 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.857 3.857 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.517 3.554 calculate_rho_elec 119 8.7 0.021 0.024 3.517 3.554 multiply_cannon_sync_h2d 27432 15.5 2.186 2.799 2.186 2.799 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.699 2.718 make_m2s 4572 13.5 0.052 0.058 2.441 2.652 rs_pw_transfer 974 11.9 0.010 0.011 2.479 2.578 make_images 4572 14.5 0.200 0.235 2.352 2.563 init_scf_loop 11 6.9 0.001 0.007 2.426 2.427 ot_diis_step 108 11.5 0.011 0.011 2.400 2.400 calculate_first_density_matrix 1 7.0 0.000 0.002 2.360 2.362 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.270 2.270 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.097 2.193 calculate_dm_sparse 119 9.5 0.000 0.000 2.030 2.099 density_rs2pw 119 9.7 0.004 0.004 1.880 1.956 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.953 1.954 grid_integrate_task_list 119 12.3 1.841 1.944 1.841 1.944 cp_fm_redistribute_end 50 14.0 1.604 1.925 1.608 1.926 cp_fm_diag_elpa_base 50 14.0 0.310 1.874 0.317 1.905 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.898 1.900 potential_pw2rs 119 12.3 0.006 0.006 1.812 1.823 pw_transfer 1439 11.6 0.063 0.067 1.728 1.755 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.690 1.728 jit_kernel_multiply 9 16.1 0.724 1.682 0.724 1.682 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.638 1.668 prepare_preconditioner 11 7.9 0.000 0.000 1.522 1.548 make_preconditioner 11 8.9 0.000 0.001 1.522 1.548 make_images_data 4572 15.5 0.045 0.051 1.133 1.523 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.428 1.482 acc_transpose_blocks 27432 15.5 0.108 0.113 1.184 1.482 wfi_extrapolate 11 7.9 0.001 0.001 1.435 1.435 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.988 1.428 fft3d_ps 1201 14.6 0.499 0.552 1.346 1.373 grid_collocate_task_list 119 9.7 1.240 1.316 1.240 1.316 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.285 1.313 mp_allgather_i34 2286 14.5 0.536 1.286 0.536 1.286 mp_alltoall_d11v 2130 13.8 1.172 1.272 1.172 1.272 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.242 1.249 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.157 1.203 mp_sum_d 4125 12.0 0.605 1.050 0.605 1.050 qs_energies_init_hamiltonians 11 5.9 0.002 0.011 1.044 1.045 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.877 0.890 acc_transpose_blocks_kernels 27432 16.5 0.181 0.269 0.654 0.859 rs_pw_transfer_PW2RS_50 119 14.3 0.592 0.613 0.731 0.776 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.692 0.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=37.839000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.636364, yerr=1.493789 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 520.650752E+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 0.013 0.029 31.000 31.001 qs_mol_dyn_low 1 2.0 0.003 0.003 30.784 30.791 qs_forces 11 3.9 0.002 0.002 30.697 30.698 qs_energies 11 4.9 0.002 0.003 29.138 29.140 scf_env_do_scf 11 5.9 0.000 0.001 24.567 24.567 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.057 22.058 dbcsr_multiply_generic 2286 12.5 0.092 0.096 16.034 16.100 velocity_verlet 10 3.0 0.001 0.001 14.798 14.800 qs_scf_new_mos 108 7.5 0.001 0.001 14.324 14.347 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.323 14.346 multiply_cannon 2286 13.5 0.195 0.200 12.919 13.730 ot_scf_mini 108 9.5 0.002 0.003 13.613 13.628 multiply_cannon_loop 2286 14.5 0.634 0.664 12.140 12.952 ot_mini 108 10.5 0.001 0.001 8.501 8.511 qs_ot_get_derivative 108 11.5 0.001 0.001 7.026 7.040 multiply_cannon_multrec 18288 15.5 1.951 2.875 6.741 7.013 rebuild_ks_matrix 119 8.3 0.000 0.000 6.193 6.208 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.193 6.207 dbcsr_mm_accdrv_process 38222 16.0 4.009 5.307 4.706 5.474 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.457 5.470 sum_up_and_integrate 119 10.3 0.030 0.031 3.841 3.849 integrate_v_rspace 119 11.3 0.002 0.002 3.810 3.822 mp_waitall_1 169478 16.3 2.665 3.479 2.665 3.479 init_scf_run 11 5.9 0.000 0.001 3.400 3.400 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.400 3.400 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.629 3.267 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.126 3.138 calculate_rho_elec 119 8.7 0.030 0.031 3.126 3.138 qs_ot_get_p 119 10.4 0.001 0.004 2.997 3.016 init_scf_loop 11 6.9 0.001 0.005 2.494 2.501 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.040 2.340 apply_single 119 13.6 0.000 0.000 2.040 2.340 calculate_first_density_matrix 1 7.0 0.000 0.001 2.232 2.233 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.376 2.209 rs_pw_transfer 974 11.9 0.009 0.010 2.113 2.202 qs_ot_p2m_diag 50 11.0 0.012 0.013 1.959 1.963 grid_integrate_task_list 119 12.3 1.804 1.921 1.804 1.921 make_m2s 4572 13.5 0.044 0.045 1.687 1.836 calculate_dm_sparse 119 9.5 0.000 0.001 1.796 1.807 density_rs2pw 119 9.7 0.004 0.004 1.710 1.783 make_images 4572 14.5 0.189 0.201 1.603 1.750 prepare_preconditioner 11 7.9 0.000 0.000 1.703 1.706 make_preconditioner 11 8.9 0.000 0.001 1.703 1.706 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.704 1.705 pw_transfer 1439 11.6 0.063 0.066 1.663 1.673 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.656 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.557 1.641 potential_pw2rs 119 12.3 0.007 0.008 1.632 1.640 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.572 1.584 multiply_cannon_sync_h2d 18288 15.5 1.422 1.541 1.422 1.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.521 1.527 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.498 1.505 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.497 1.498 cp_fm_redistribute_end 50 14.0 1.111 1.480 1.112 1.480 cp_fm_diag_elpa_base 50 14.0 0.352 1.429 0.366 1.466 ot_diis_step 108 11.5 0.011 0.011 1.457 1.458 mp_sum_l 7207 12.9 1.047 1.410 1.047 1.410 grid_collocate_task_list 119 9.7 1.202 1.304 1.202 1.304 fft_wrap_pw1pw2_140 487 13.2 0.087 0.091 1.273 1.286 jit_kernel_multiply 7 16.3 0.646 1.273 0.646 1.273 fft3d_ps 1201 14.6 0.505 0.522 1.259 1.272 acc_transpose_blocks 18288 15.5 0.075 0.077 1.220 1.245 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 wfi_extrapolate 11 7.9 0.001 0.001 1.124 1.124 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 0.947 0.947 make_images_data 4572 15.5 0.045 0.048 0.765 0.926 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.852 0.872 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.651 0.844 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.810 acc_transpose_blocks_kernels 18288 16.5 0.209 0.217 0.781 0.793 mp_alltoall_d11v 2130 13.8 0.657 0.728 0.657 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.661 0.720 cp_fm_cholesky_invert 11 10.9 0.697 0.701 0.697 0.701 mp_alltoall_z22v 1201 16.6 0.608 0.673 0.608 0.673 mp_waitany 9880 13.7 0.519 0.628 0.519 0.628 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.001000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.272727, yerr=1.420045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 544.911360E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.029 34.127 34.128 qs_mol_dyn_low 1 2.0 0.003 0.003 33.871 33.878 qs_forces 11 3.9 0.010 0.016 33.815 33.816 qs_energies 11 4.9 0.005 0.012 32.117 32.126 scf_env_do_scf 11 5.9 0.000 0.001 26.952 26.953 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.504 23.504 dbcsr_multiply_generic 2286 12.5 0.099 0.103 17.757 17.886 velocity_verlet 10 3.0 0.001 0.001 17.272 17.273 qs_scf_new_mos 108 7.5 0.001 0.001 15.588 15.642 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.588 15.641 multiply_cannon 2286 13.5 0.223 0.248 14.337 14.716 ot_scf_mini 108 9.5 0.002 0.003 14.661 14.714 multiply_cannon_loop 2286 14.5 0.939 0.972 13.415 13.775 ot_mini 108 10.5 0.001 0.001 8.999 9.068 multiply_cannon_multrec 27432 15.5 2.336 2.990 8.577 8.931 dbcsr_mm_accdrv_process 47916 15.9 5.275 6.915 6.144 7.325 qs_ot_get_derivative 108 11.5 0.001 0.001 7.205 7.259 rebuild_ks_matrix 119 8.3 0.000 0.000 6.371 6.426 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.371 6.425 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.691 5.744 sum_up_and_integrate 119 10.3 0.035 0.037 3.629 3.636 init_scf_run 11 5.9 0.000 0.001 3.604 3.604 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.604 3.604 integrate_v_rspace 119 11.3 0.002 0.002 3.594 3.602 init_scf_loop 11 6.9 0.005 0.039 3.430 3.430 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.166 3.198 calculate_rho_elec 119 8.7 0.040 0.046 3.165 3.198 qs_ot_get_p 119 10.4 0.001 0.001 3.080 3.159 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.706 3.110 prepare_preconditioner 11 7.9 0.000 0.000 2.530 2.538 make_preconditioner 11 8.9 0.000 0.000 2.530 2.538 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.138 2.455 mp_waitall_1 145218 16.4 1.857 2.429 1.857 2.429 calculate_first_density_matrix 1 7.0 0.000 0.002 2.283 2.284 make_m2s 4572 13.5 0.054 0.055 2.083 2.218 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.025 2.150 apply_single 119 13.6 0.000 0.000 2.024 2.149 make_images 4572 14.5 0.271 0.334 1.976 2.110 calculate_dm_sparse 119 9.5 0.000 0.001 2.030 2.089 rs_pw_transfer 974 11.9 0.009 0.009 1.861 1.958 grid_integrate_task_list 119 12.3 1.829 1.892 1.829 1.892 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.859 1.886 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.873 1.881 jit_kernel_multiply 10 15.9 0.809 1.816 0.809 1.816 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.808 1.810 density_rs2pw 119 9.7 0.003 0.004 1.658 1.768 ot_diis_step 108 11.5 0.012 0.012 1.750 1.751 pw_transfer 1439 11.6 0.063 0.067 1.709 1.738 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.619 1.651 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.637 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.567 1.568 acc_transpose_blocks 27432 15.5 0.110 0.113 1.453 1.479 potential_pw2rs 119 12.3 0.008 0.010 1.428 1.430 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.335 1.369 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.354 1.355 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.756 1.347 cp_fm_redistribute_end 50 14.0 0.892 1.334 0.893 1.334 fft3d_ps 1201 14.6 0.533 0.586 1.300 1.324 cp_fm_diag_elpa_base 50 14.0 0.420 1.279 0.439 1.319 qs_energies_init_hamiltonians 11 5.9 0.037 0.044 1.297 1.306 grid_collocate_task_list 119 9.7 1.221 1.290 1.221 1.290 wfi_extrapolate 11 7.9 0.001 0.001 1.275 1.275 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.221 1.232 mp_sum_l 7207 12.9 0.921 1.179 0.921 1.179 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.078 1.099 cp_fm_upper_to_full 72 13.5 0.790 1.092 0.790 1.092 multiply_cannon_sync_h2d 27432 15.5 1.003 1.084 1.003 1.084 dbcsr_complete_redistribute 329 12.2 0.132 0.168 0.764 1.035 make_images_data 4572 15.5 0.045 0.048 0.809 0.945 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.694 0.894 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.831 0.851 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.578 0.838 mp_alltoall_d11v 2130 13.8 0.706 0.808 0.706 0.808 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.800 0.805 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.669 0.778 mp_sum_d 4123 12.0 0.591 0.753 0.591 0.753 cp_fm_cholesky_invert 11 10.9 0.724 0.727 0.724 0.727 mp_alltoall_i22 627 13.8 0.423 0.701 0.423 0.701 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.128000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=518.090909, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 598.925312E+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 0.011 0.028 27.617 27.618 qs_mol_dyn_low 1 2.0 0.003 0.003 27.378 27.386 qs_forces 11 3.9 0.002 0.002 27.320 27.321 qs_energies 11 4.9 0.001 0.001 25.616 25.619 scf_env_do_scf 11 5.9 0.000 0.001 20.998 20.998 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.532 18.532 velocity_verlet 10 3.0 0.001 0.001 14.125 14.127 dbcsr_multiply_generic 2286 12.5 0.092 0.095 12.068 12.207 qs_scf_new_mos 108 7.5 0.001 0.001 10.957 10.978 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.957 10.977 ot_scf_mini 108 9.5 0.002 0.002 10.281 10.309 multiply_cannon 2286 13.5 0.230 0.237 9.557 9.872 multiply_cannon_loop 2286 14.5 0.330 0.338 8.665 8.886 multiply_cannon_multrec 9144 15.5 1.590 1.826 5.746 5.945 rebuild_ks_matrix 119 8.3 0.000 0.000 5.811 5.836 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.811 5.836 ot_mini 108 10.5 0.001 0.001 5.785 5.819 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.178 5.201 qs_ot_get_derivative 108 11.5 0.001 0.001 4.494 4.522 dbcsr_mm_accdrv_process 12550 15.8 3.166 3.727 4.054 4.121 sum_up_and_integrate 119 10.3 0.037 0.041 3.565 3.570 integrate_v_rspace 119 11.3 0.002 0.003 3.527 3.533 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.216 3.226 calculate_rho_elec 119 8.7 0.059 0.061 3.216 3.225 init_scf_run 11 5.9 0.000 0.001 3.190 3.190 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.189 3.189 qs_ot_get_p 119 10.4 0.001 0.001 2.726 2.755 init_scf_loop 11 6.9 0.000 0.000 2.446 2.448 mp_waitall_1 121218 16.5 1.727 2.261 1.727 2.261 calculate_first_density_matrix 1 7.0 0.000 0.000 2.081 2.083 grid_integrate_task_list 119 12.3 1.842 1.930 1.842 1.930 make_m2s 4572 13.5 0.034 0.035 1.735 1.897 make_images 4572 14.5 0.266 0.298 1.647 1.808 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.803 1.806 jit_kernel_multiply 10 15.9 0.850 1.746 0.850 1.746 pw_transfer 1439 11.6 0.063 0.065 1.728 1.738 calculate_dm_sparse 119 9.5 0.000 0.000 1.706 1.727 prepare_preconditioner 11 7.9 0.000 0.000 1.715 1.720 make_preconditioner 11 8.9 0.000 0.000 1.715 1.720 rs_pw_transfer 974 11.9 0.008 0.008 1.649 1.711 density_rs2pw 119 9.7 0.003 0.003 1.613 1.665 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.638 1.646 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.604 1.631 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.606 1.607 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.520 1.535 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.504 1.505 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.345 1.354 fft_wrap_pw1pw2_140 487 13.2 0.082 0.085 1.333 1.343 potential_pw2rs 119 12.3 0.010 0.011 1.337 1.342 grid_collocate_task_list 119 9.7 1.272 1.340 1.272 1.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.317 1.317 fft3d_ps 1201 14.6 0.538 0.550 1.306 1.314 cp_fm_redistribute_end 50 14.0 0.652 1.296 0.653 1.296 cp_fm_diag_elpa_base 50 14.0 0.599 1.228 0.642 1.284 ot_diis_step 108 11.5 0.012 0.013 1.279 1.279 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.216 1.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.150 1.175 apply_single 119 13.6 0.000 0.000 1.150 1.175 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.134 1.148 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.124 1.129 wfi_extrapolate 11 7.9 0.001 0.001 1.063 1.063 hybrid_alltoall_any 4725 16.4 0.062 0.178 0.785 1.041 make_images_data 4572 15.5 0.039 0.042 0.800 1.005 mp_alltoall_d11v 2130 13.8 0.831 0.935 0.831 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.924 acc_transpose_blocks 9144 15.5 0.038 0.039 0.887 0.896 cp_fm_cholesky_invert 11 10.9 0.854 0.857 0.854 0.857 multiply_cannon_sync_h2d 9144 15.5 0.711 0.787 0.711 0.787 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.390 0.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.755 0.758 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.715 0.724 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.673 0.721 mp_alltoall_z22v 1201 16.6 0.642 0.674 0.642 0.674 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.648 0.652 mp_allgather_i34 2286 14.5 0.228 0.620 0.228 0.620 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.315 0.600 qs_create_task_list 11 7.9 0.000 0.001 0.553 0.575 generate_qs_task_list 11 8.9 0.190 0.213 0.552 0.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.618000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.454545, yerr=3.985511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 758.935552E+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 0.016 0.033 39.578 39.578 qs_mol_dyn_low 1 2.0 0.003 0.003 39.379 39.387 qs_forces 11 3.9 0.001 0.002 39.320 39.321 qs_energies 11 4.9 0.001 0.001 37.412 37.415 scf_env_do_scf 11 5.9 0.000 0.001 31.949 31.949 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.226 24.229 velocity_verlet 10 3.0 0.001 0.001 22.523 22.529 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.228 17.466 qs_scf_new_mos 108 7.5 0.001 0.001 15.550 15.652 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.549 15.652 ot_scf_mini 108 9.5 0.002 0.002 14.483 14.586 multiply_cannon 2286 13.5 0.298 0.301 13.489 14.405 multiply_cannon_loop 2286 14.5 0.342 0.352 12.249 13.195 ot_mini 108 10.5 0.001 0.001 8.724 8.842 multiply_cannon_multrec 9144 15.5 3.369 4.795 8.505 8.585 init_scf_loop 11 6.9 0.000 0.000 7.696 7.702 rebuild_ks_matrix 119 8.3 0.000 0.000 6.710 6.853 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.710 6.852 qs_ot_get_derivative 108 11.5 0.001 0.001 6.721 6.824 prepare_preconditioner 11 7.9 0.000 0.000 6.776 6.790 make_preconditioner 11 8.9 0.000 0.000 6.776 6.790 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.334 6.667 dbcsr_mm_accdrv_process 12550 15.8 4.120 5.561 5.011 6.313 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.069 6.199 cp_fm_upper_to_full 72 14.2 3.159 4.539 3.159 4.539 sum_up_and_integrate 119 10.3 0.064 0.065 3.581 3.587 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.580 3.583 calculate_rho_elec 119 8.7 0.118 0.121 3.579 3.582 integrate_v_rspace 119 11.3 0.003 0.003 3.517 3.523 init_scf_run 11 5.9 0.000 0.001 3.462 3.462 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.462 3.462 mp_waitall_1 97218 16.6 2.401 3.369 2.401 3.369 qs_ot_get_p 119 10.4 0.001 0.001 3.062 3.200 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.444 2.853 dbcsr_complete_redistribute 329 12.2 0.287 0.291 1.962 2.791 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.664 2.477 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.156 2.439 apply_single 119 13.6 0.000 0.000 2.156 2.439 make_m2s 4572 13.5 0.037 0.038 2.174 2.335 mp_alltoall_i22 627 13.8 1.413 2.258 1.413 2.258 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.437 2.244 make_images 4572 14.5 0.351 0.384 2.055 2.216 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 1.302 2.199 calculate_dm_sparse 119 9.5 0.000 0.000 2.140 2.154 calculate_first_density_matrix 1 7.0 0.000 0.000 2.080 2.082 grid_integrate_task_list 119 12.3 1.988 2.004 1.988 2.004 ot_diis_step 108 11.5 0.014 0.014 1.979 1.979 pw_transfer 1439 11.6 0.066 0.067 1.975 1.979 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.879 1.884 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.846 1.847 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.791 1.792 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.743 1.744 density_rs2pw 119 9.7 0.003 0.004 1.717 1.736 mp_sum_l 7207 12.9 1.030 1.723 1.030 1.723 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.647 1.699 fft_wrap_pw1pw2_140 487 13.2 0.087 0.089 1.582 1.586 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.536 1.575 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.570 1.570 fft3d_ps 1201 14.6 0.565 0.573 1.514 1.520 grid_collocate_task_list 119 9.7 1.434 1.450 1.434 1.450 cp_fm_cholesky_invert 11 10.9 1.404 1.407 1.404 1.407 rs_pw_transfer 974 11.9 0.009 0.009 1.319 1.349 wfi_extrapolate 11 7.9 0.001 0.001 1.323 1.323 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.285 1.285 cp_fm_diag_elpa_base 50 14.0 1.144 1.194 1.284 1.284 hybrid_alltoall_any 4725 16.4 0.086 0.147 1.062 1.253 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.219 1.223 make_images_data 4572 15.5 0.043 0.046 0.998 1.204 potential_pw2rs 119 12.3 0.014 0.014 1.159 1.161 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.120 1.139 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.084 1.101 mp_alltoall_d11v 2130 13.8 1.081 1.088 1.081 1.088 multiply_cannon_sync_h2d 9144 15.5 1.042 1.047 1.042 1.047 jit_kernel_multiply 6 15.7 0.864 1.021 0.864 1.021 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.983 1.019 qs_create_task_list 11 7.9 0.000 0.001 0.941 0.954 generate_qs_task_list 11 8.9 0.372 0.391 0.941 0.953 acc_transpose_blocks 9144 15.5 0.037 0.038 0.906 0.921 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.852 0.866 mp_alltoall_z22v 1201 16.6 0.813 0.831 0.813 0.831 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=39.578000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.727273, yerr=14.245196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 505.438208E+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 1771933. 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 0.028 0.059 81.948 81.963 qs_mol_dyn_low 1 2.0 0.003 0.003 81.599 81.673 qs_forces 11 3.9 0.002 0.003 81.531 81.532 qs_energies 11 4.9 0.001 0.002 78.756 78.767 scf_env_do_scf 11 5.9 0.000 0.001 70.163 70.165 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.680 64.680 dbcsr_multiply_generic 2055 12.4 0.105 0.110 51.951 52.135 qs_scf_new_mos 99 7.5 0.000 0.001 47.913 48.047 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.912 48.046 ot_scf_mini 99 9.5 0.002 0.002 45.488 45.610 multiply_cannon 2055 13.4 0.180 0.187 42.493 43.255 multiply_cannon_loop 2055 14.4 1.560 1.601 41.537 42.292 velocity_verlet 10 3.0 0.001 0.001 41.898 41.898 ot_mini 99 10.5 0.001 0.001 27.672 27.764 qs_ot_get_derivative 99 11.5 0.001 0.001 20.831 20.933 multiply_cannon_multrec 49320 15.4 12.154 12.893 17.161 17.729 rebuild_ks_matrix 110 8.3 0.000 0.000 14.274 14.392 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.274 14.392 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.547 12.653 mp_waitall_1 241148 16.1 11.646 12.351 11.646 12.351 multiply_cannon_sync_h2d 49320 15.4 9.930 10.506 9.930 10.506 qs_ot_get_p 110 10.4 0.001 0.001 9.286 9.397 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.011 8.047 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.168 7.677 apply_single 110 13.6 0.000 0.001 7.168 7.677 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.125 7.663 sum_up_and_integrate 110 10.3 0.036 0.044 6.904 6.919 integrate_v_rspace 110 11.3 0.002 0.003 6.867 6.892 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.607 6.653 init_scf_run 11 5.9 0.000 0.001 6.607 6.607 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.607 6.607 ot_diis_step 99 11.5 0.005 0.006 6.558 6.558 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.115 6.249 calculate_rho_elec 110 8.6 0.020 0.024 6.115 6.249 qs_ot_p2m_diag 48 11.0 0.012 0.019 5.979 6.015 mp_sum_l 6514 12.8 5.208 5.957 5.208 5.957 init_scf_loop 11 6.9 0.000 0.000 5.457 5.457 dbcsr_mm_accdrv_process 87628 16.1 2.066 2.155 4.886 5.168 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.012 5.013 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.552 4.553 cp_fm_redistribute_end 48 14.0 3.961 4.528 3.965 4.530 cp_fm_diag_elpa_base 48 14.0 0.558 4.469 0.561 4.493 wfi_extrapolate 11 7.9 0.001 0.001 3.955 3.955 make_m2s 4110 13.4 0.060 0.064 3.806 3.908 calculate_dm_sparse 110 9.5 0.000 0.001 3.788 3.878 make_images 4110 14.4 0.177 0.192 3.712 3.820 rs_pw_transfer 902 11.9 0.011 0.012 3.641 3.792 multiply_cannon_metrocomm1 49320 15.4 0.057 0.061 2.630 3.731 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.501 3.505 grid_integrate_task_list 110 12.3 3.236 3.391 3.236 3.391 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.289 3.343 prepare_preconditioner 11 7.9 0.000 0.000 3.318 3.335 make_preconditioner 11 8.9 0.000 0.000 3.318 3.335 density_rs2pw 110 9.6 0.004 0.004 3.167 3.305 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.192 3.228 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.114 3.154 pw_transfer 1331 11.6 0.053 0.067 2.759 2.816 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.672 2.731 calculate_first_density_matrix 1 7.0 0.000 0.000 2.569 2.571 jit_kernel_multiply 13 15.9 2.538 2.552 2.538 2.552 potential_pw2rs 110 12.3 0.006 0.007 2.521 2.541 mp_alltoall_d11v 2046 13.8 2.103 2.470 2.103 2.470 fft_wrap_pw1pw2_140 451 13.1 0.170 0.191 2.252 2.313 fft3d_ps 1111 14.6 0.751 0.842 2.209 2.260 acc_transpose_blocks 49320 15.4 0.222 0.231 2.175 2.237 grid_collocate_task_list 110 9.6 2.087 2.197 2.087 2.197 mp_waitany 14300 13.8 1.797 2.083 1.797 2.083 mp_sum_d 3879 11.9 1.468 1.957 1.468 1.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.867 1.885 make_images_data 4110 15.4 0.043 0.046 1.742 1.874 cp_fm_cholesky_invert 11 10.9 1.844 1.847 1.844 1.847 hybrid_alltoall_any 4261 16.3 0.082 0.478 1.524 1.775 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.629 1.649 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.963000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.909091, yerr=3.728381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.637312E+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 2914018. 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 0.025 0.042 69.703 69.704 qs_mol_dyn_low 1 2.0 0.003 0.003 69.395 69.404 qs_forces 11 3.9 0.002 0.002 69.103 69.105 qs_energies 11 4.9 0.004 0.010 65.794 65.799 scf_env_do_scf 11 5.9 0.000 0.001 57.214 57.218 scf_env_do_scf_inner_loop 99 6.5 0.003 0.013 49.312 49.313 dbcsr_multiply_generic 2055 12.4 0.112 0.119 37.910 38.105 velocity_verlet 10 3.0 0.001 0.001 36.741 36.747 qs_scf_new_mos 99 7.5 0.001 0.001 33.160 33.297 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.160 33.296 multiply_cannon 2055 13.4 0.222 0.244 30.974 32.207 ot_scf_mini 99 9.5 0.003 0.004 31.475 31.604 multiply_cannon_loop 2055 14.4 0.926 0.949 29.635 30.391 ot_mini 99 10.5 0.001 0.001 18.446 18.589 multiply_cannon_multrec 24660 15.4 7.562 10.088 13.747 16.051 rebuild_ks_matrix 110 8.3 0.000 0.000 13.770 13.897 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.769 13.896 qs_ot_get_derivative 99 11.5 0.001 0.001 12.608 12.741 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.150 12.261 mp_waitall_1 186928 16.3 8.163 10.785 8.163 10.785 multiply_cannon_sync_h2d 24660 15.4 6.957 8.718 6.957 8.718 multiply_cannon_metrocomm3 24660 15.4 0.068 0.071 5.248 8.365 init_scf_loop 11 6.9 0.002 0.018 7.864 7.864 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.529 7.271 apply_single 110 13.6 0.000 0.001 6.529 7.271 sum_up_and_integrate 110 10.3 0.053 0.059 6.652 6.662 integrate_v_rspace 110 11.3 0.002 0.002 6.599 6.612 qs_ot_get_p 110 10.4 0.001 0.001 6.374 6.535 dbcsr_mm_accdrv_process 52282 16.1 4.677 5.605 6.028 6.384 init_scf_run 11 5.9 0.000 0.001 6.180 6.181 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.180 6.180 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.948 5.958 calculate_rho_elec 110 8.6 0.039 0.047 5.948 5.958 prepare_preconditioner 11 7.9 0.000 0.000 5.808 5.827 make_preconditioner 11 8.9 0.000 0.001 5.808 5.827 ot_diis_step 99 11.5 0.010 0.010 5.778 5.779 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.368 5.526 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.707 5.442 make_m2s 4110 13.4 0.056 0.059 4.420 4.902 make_images 4110 14.4 0.395 0.438 4.312 4.789 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.402 4.423 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.962 3.963 pw_transfer 1331 11.6 0.066 0.070 3.470 3.611 wfi_extrapolate 11 7.9 0.001 0.001 3.523 3.523 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.364 3.502 density_rs2pw 110 9.6 0.004 0.004 3.203 3.425 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.389 3.391 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.297 3.382 cp_fm_redistribute_end 48 14.0 2.521 3.359 2.524 3.359 cp_fm_diag_elpa_base 48 14.0 0.800 3.243 0.832 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.317 3.318 grid_integrate_task_list 110 12.3 3.155 3.315 3.155 3.315 rs_pw_transfer 902 11.9 0.012 0.014 3.042 3.194 calculate_dm_sparse 110 9.5 0.001 0.001 2.991 3.029 make_images_data 4110 15.4 0.047 0.051 2.500 3.004 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.931 3.001 fft_wrap_pw1pw2_140 451 13.1 0.199 0.218 2.788 2.923 hybrid_alltoall_any 4261 16.3 0.102 0.445 2.176 2.854 fft3d_ps 1111 14.6 1.073 1.267 2.722 2.850 cp_fm_cholesky_invert 11 10.9 2.699 2.708 2.699 2.708 mp_sum_l 6514 12.8 1.900 2.615 1.900 2.615 calculate_first_density_matrix 1 7.0 0.000 0.002 2.562 2.564 potential_pw2rs 110 12.3 0.008 0.009 2.476 2.493 jit_kernel_multiply 13 16.2 1.001 2.468 1.001 2.468 grid_collocate_task_list 110 9.6 2.121 2.271 2.121 2.271 mp_alltoall_d11v 2046 13.8 1.935 2.148 1.935 2.148 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.985 2.014 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.803 1.879 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.869 1.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.784 1.801 mp_allgather_i34 2055 14.4 0.650 1.724 0.650 1.724 mp_irecv_dv 57340 16.2 0.681 1.670 0.681 1.670 dbcsr_complete_redistribute 325 12.2 0.263 0.325 1.294 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.548 1.560 acc_transpose_blocks 24660 15.4 0.109 0.112 1.523 1.539 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.479 mp_alltoall_z22v 1111 16.6 1.378 1.460 1.378 1.460 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.704000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.181818, yerr=6.925817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 660.422656E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.031 60.798 60.799 qs_mol_dyn_low 1 2.0 0.003 0.003 60.445 60.456 qs_forces 11 3.9 0.002 0.002 60.072 60.073 qs_energies 11 4.9 0.004 0.016 56.894 56.898 scf_env_do_scf 11 5.9 0.001 0.001 48.919 48.919 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.087 40.088 velocity_verlet 10 3.0 0.001 0.001 33.063 33.075 dbcsr_multiply_generic 2055 12.4 0.107 0.109 28.570 28.866 qs_scf_new_mos 99 7.5 0.001 0.001 25.582 25.682 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.581 25.681 ot_scf_mini 99 9.5 0.003 0.003 24.345 24.460 multiply_cannon 2055 13.4 0.213 0.225 22.151 23.453 multiply_cannon_loop 2055 14.4 0.616 0.632 20.899 22.174 ot_mini 99 10.5 0.001 0.001 13.798 13.917 rebuild_ks_matrix 110 8.3 0.000 0.000 12.184 12.318 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.183 12.317 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.747 10.870 multiply_cannon_multrec 16440 15.4 3.878 5.007 9.596 10.717 mp_waitall_1 146766 16.3 7.269 10.186 7.269 10.186 qs_ot_get_derivative 99 11.5 0.001 0.001 9.296 9.414 init_scf_loop 11 6.9 0.001 0.003 8.793 8.799 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.411 7.230 prepare_preconditioner 11 7.9 0.000 0.000 7.018 7.032 make_preconditioner 11 8.9 0.000 0.000 7.018 7.032 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.340 6.702 sum_up_and_integrate 110 10.3 0.061 0.062 6.457 6.471 integrate_v_rspace 110 11.3 0.002 0.002 6.395 6.409 qs_ot_get_p 110 10.4 0.001 0.001 5.748 5.895 dbcsr_mm_accdrv_process 34862 16.1 4.744 5.251 5.573 5.713 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.596 5.604 calculate_rho_elec 110 8.6 0.058 0.058 5.596 5.604 init_scf_run 11 5.9 0.000 0.001 5.519 5.519 scf_env_initial_rho_setup 11 6.9 0.002 0.006 5.518 5.519 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.015 5.461 apply_single 110 13.6 0.000 0.000 5.014 5.461 make_m2s 4110 13.4 0.049 0.051 4.187 4.574 make_images 4110 14.4 0.388 0.506 4.073 4.459 ot_diis_step 99 11.5 0.010 0.011 4.458 4.459 multiply_cannon_sync_h2d 16440 15.4 3.686 4.356 3.686 4.356 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.056 4.061 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.179 3.809 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.715 3.715 grid_integrate_task_list 110 12.3 3.179 3.374 3.179 3.374 pw_transfer 1331 11.6 0.064 0.072 3.187 3.198 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.165 3.167 cp_fm_redistribute_end 48 14.0 1.961 3.131 1.963 3.132 cp_fm_diag_elpa_base 48 14.0 1.104 2.997 1.163 3.106 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.081 3.093 density_rs2pw 110 9.6 0.004 0.004 2.871 3.005 wfi_extrapolate 11 7.9 0.001 0.001 2.993 2.993 rs_pw_transfer 902 11.9 0.010 0.011 2.768 2.923 make_images_data 4110 15.4 0.043 0.047 2.395 2.864 cp_fm_cholesky_invert 11 10.9 2.843 2.850 2.843 2.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.772 2.773 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.112 2.751 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.596 2.612 calculate_dm_sparse 110 9.5 0.001 0.001 2.512 2.543 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.417 2.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.428 2.478 multiply_cannon_metrocomm4 14385 15.4 0.043 0.047 0.887 2.420 fft3d_ps 1111 14.6 1.064 1.077 2.396 2.406 calculate_first_density_matrix 1 7.0 0.000 0.002 2.373 2.376 grid_collocate_task_list 110 9.6 2.190 2.337 2.190 2.337 mp_irecv_dv 48980 15.7 0.819 2.302 0.819 2.302 potential_pw2rs 110 12.3 0.010 0.010 2.279 2.289 mp_sum_l 6514 12.8 1.622 2.222 1.622 2.222 mp_alltoall_d11v 2046 13.8 1.808 2.106 1.808 2.106 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.966 1.969 dbcsr_complete_redistribute 325 12.2 0.322 0.358 1.499 1.957 cp_fm_upper_to_full 70 13.6 1.395 1.845 1.395 1.845 cp_fm_cholesky_decompose 22 10.9 1.651 1.675 1.651 1.675 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.639 1.656 mp_allgather_i34 2055 14.4 0.576 1.582 0.576 1.582 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.375 1.485 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.465 1.477 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.998 1.447 mp_waitany 17072 13.8 1.215 1.405 1.215 1.405 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.280 1.287 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.799000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.727273, yerr=8.634928 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 722.055168E+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 0.029 0.067 64.857 64.858 qs_mol_dyn_low 1 2.0 0.003 0.003 64.519 64.528 qs_forces 11 3.9 0.002 0.002 64.090 64.091 qs_energies 11 4.9 0.003 0.006 60.741 60.744 scf_env_do_scf 11 5.9 0.001 0.001 52.480 52.483 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.744 40.746 velocity_verlet 10 3.0 0.001 0.001 36.923 36.941 dbcsr_multiply_generic 2055 12.4 0.114 0.119 29.461 29.764 qs_scf_new_mos 99 7.5 0.001 0.001 26.454 26.551 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.454 26.551 ot_scf_mini 99 9.5 0.003 0.003 24.823 24.935 multiply_cannon 2055 13.4 0.243 0.260 22.448 23.481 multiply_cannon_loop 2055 14.4 0.887 0.905 21.097 21.760 ot_mini 99 10.5 0.001 0.001 14.273 14.385 multiply_cannon_multrec 24660 15.4 4.226 6.799 12.627 13.723 rebuild_ks_matrix 110 8.3 0.000 0.000 11.840 11.942 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.839 11.942 init_scf_loop 11 6.9 0.001 0.003 11.695 11.697 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.482 10.573 qs_ot_get_derivative 99 11.5 0.001 0.001 10.118 10.221 prepare_preconditioner 11 7.9 0.000 0.000 9.958 9.973 make_preconditioner 11 8.9 0.000 0.001 9.958 9.973 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.208 9.652 dbcsr_mm_accdrv_process 52304 16.0 6.794 8.411 8.252 9.157 mp_waitall_1 126806 16.4 4.576 6.658 4.576 6.658 sum_up_and_integrate 110 10.3 0.067 0.071 6.296 6.312 integrate_v_rspace 110 11.3 0.002 0.003 6.229 6.244 qs_ot_get_p 110 10.4 0.001 0.001 5.584 5.729 make_m2s 4110 13.4 0.059 0.061 5.396 5.633 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.576 5.588 calculate_rho_elec 110 8.6 0.077 0.081 5.576 5.587 init_scf_run 11 5.9 0.000 0.001 5.491 5.491 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.490 5.491 make_images 4110 14.4 0.573 0.694 5.257 5.490 cp_fm_upper_to_full 70 13.8 3.252 4.626 3.252 4.626 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.017 4.136 apply_single 110 13.6 0.000 0.000 4.017 4.136 ot_diis_step 99 11.5 0.011 0.011 4.114 4.115 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.783 3.798 dbcsr_complete_redistribute 325 12.2 0.425 0.484 2.660 3.774 grid_integrate_task_list 110 12.3 3.273 3.430 3.273 3.430 multiply_cannon_sync_h2d 24660 15.4 3.180 3.353 3.180 3.353 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.290 3.338 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.531 3.314 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.306 3.307 pw_transfer 1331 11.6 0.064 0.071 3.247 3.269 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.116 3.214 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.141 3.167 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.035 3.089 make_images_data 4110 15.4 0.046 0.050 2.730 3.023 wfi_extrapolate 11 7.9 0.001 0.001 2.950 2.950 calculate_dm_sparse 110 9.5 0.001 0.001 2.921 2.950 hybrid_alltoall_any 4261 16.3 0.119 0.455 2.387 2.940 density_rs2pw 110 9.6 0.004 0.004 2.732 2.882 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.741 2.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.805 2.807 cp_fm_cholesky_invert 11 10.9 2.793 2.803 2.793 2.803 mp_alltoall_i22 605 13.7 1.630 2.788 1.630 2.788 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.763 2.765 cp_fm_redistribute_end 48 14.0 1.374 2.736 1.375 2.737 cp_fm_diag_elpa_base 48 14.0 1.280 2.610 1.359 2.720 fft_wrap_pw1pw2_140 451 13.1 0.202 0.212 2.666 2.692 rs_pw_transfer 902 11.9 0.010 0.011 2.385 2.527 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.432 2.465 fft3d_ps 1111 14.6 1.064 1.093 2.444 2.461 calculate_first_density_matrix 1 7.0 0.001 0.002 2.447 2.451 grid_collocate_task_list 110 9.6 2.223 2.417 2.223 2.417 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.269 2.270 jit_kernel_multiply 12 15.8 1.128 2.191 1.128 2.191 potential_pw2rs 110 12.3 0.012 0.013 2.079 2.093 mp_alltoall_d11v 2046 13.8 1.826 2.001 1.826 2.001 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.717 1.747 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.600 1.701 cp_fm_cholesky_decompose 22 10.9 1.658 1.689 1.658 1.689 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.604 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.528 1.544 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.843 1.542 mp_sum_l 6514 12.8 0.940 1.513 0.940 1.513 acc_transpose_blocks 24660 15.4 0.103 0.104 1.490 1.508 mp_irecv_dv 62702 16.1 0.745 1.466 0.745 1.466 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.219 1.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.858000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.000000, yerr=6.619530 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 829.542400E+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 0.036 0.039 57.913 57.914 qs_mol_dyn_low 1 2.0 0.003 0.003 56.372 57.642 qs_forces 11 3.9 0.002 0.002 55.177 55.178 qs_energies 11 4.9 0.001 0.002 51.552 51.557 scf_env_do_scf 11 5.9 0.000 0.001 43.348 43.348 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.355 35.356 velocity_verlet 10 3.0 0.001 0.001 32.129 32.197 dbcsr_multiply_generic 2055 12.4 0.104 0.107 23.036 23.181 qs_scf_new_mos 99 7.5 0.001 0.001 21.107 21.159 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.106 21.158 ot_scf_mini 99 9.5 0.002 0.002 19.851 19.886 multiply_cannon 2055 13.4 0.246 0.259 17.429 18.491 multiply_cannon_loop 2055 14.4 0.323 0.335 16.052 16.243 rebuild_ks_matrix 110 8.3 0.000 0.000 11.577 11.605 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.576 11.604 ot_mini 99 10.5 0.001 0.001 10.614 10.643 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.311 10.337 multiply_cannon_multrec 8220 15.4 3.198 4.423 7.403 8.421 mp_waitall_1 106626 16.5 6.480 8.283 6.480 8.283 init_scf_loop 11 6.9 0.000 0.000 7.939 7.940 qs_ot_get_derivative 99 11.5 0.001 0.001 6.736 6.770 sum_up_and_integrate 110 10.3 0.080 0.081 6.275 6.288 prepare_preconditioner 11 7.9 0.000 0.000 6.282 6.285 make_preconditioner 11 8.9 0.000 0.000 6.282 6.285 integrate_v_rspace 110 11.3 0.003 0.003 6.195 6.208 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.867 5.935 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.694 5.708 calculate_rho_elec 110 8.6 0.115 0.115 5.693 5.708 qs_ot_get_p 110 10.4 0.001 0.001 5.259 5.314 init_scf_run 11 5.9 0.000 0.001 5.083 5.083 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.083 5.083 dbcsr_mm_accdrv_process 17442 15.9 2.902 3.918 4.074 4.982 make_m2s 4110 13.4 0.038 0.040 4.285 4.562 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.215 4.559 make_images 4110 14.4 0.632 0.687 4.156 4.432 ot_diis_step 99 11.5 0.012 0.012 3.852 3.852 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.838 3.842 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.767 3.810 apply_single 110 13.6 0.000 0.000 3.767 3.810 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.525 3.526 grid_integrate_task_list 110 12.3 3.364 3.488 3.364 3.488 pw_transfer 1331 11.6 0.065 0.070 3.343 3.356 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.236 3.251 cp_fm_cholesky_invert 11 10.9 3.131 3.135 3.131 3.135 multiply_cannon_sync_h2d 8220 15.4 2.907 3.058 2.907 3.058 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.965 2.966 cp_fm_redistribute_end 48 14.0 0.746 2.920 0.752 2.922 density_rs2pw 110 9.6 0.004 0.004 2.745 2.911 make_images_data 4110 15.4 0.039 0.043 2.452 2.898 cp_fm_diag_elpa_base 48 14.0 1.978 2.716 2.160 2.885 fft_wrap_pw1pw2_140 451 13.1 0.211 0.214 2.743 2.771 wfi_extrapolate 11 7.9 0.001 0.001 2.759 2.759 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.365 2.744 qs_energies_init_hamiltonians 11 5.9 0.004 0.008 2.660 2.661 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.526 2.527 calculate_dm_sparse 110 9.5 0.001 0.001 2.479 2.512 fft3d_ps 1111 14.6 1.126 1.161 2.496 2.506 grid_collocate_task_list 110 9.6 2.323 2.446 2.323 2.446 md_write_output 11 3.9 0.003 0.035 0.154 2.442 md_output 10 3.0 0.000 0.000 0.156 2.440 rs_pw_transfer 902 11.9 0.010 0.010 2.263 2.430 write_trajectory 44 4.9 0.001 0.007 0.151 2.406 write_particle_coordinates 11 5.9 0.150 2.398 0.150 2.398 calculate_first_density_matrix 1 7.0 0.000 0.000 2.233 2.234 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.147 2.163 potential_pw2rs 110 12.3 0.015 0.015 2.007 2.015 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.983 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.933 1.952 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.841 1.857 mp_alltoall_d11v 2046 13.8 1.710 1.845 1.710 1.845 cp_fm_cholesky_decompose 22 10.9 1.706 1.723 1.706 1.723 mp_allgather_i34 2055 14.4 0.564 1.693 0.564 1.693 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.502 1.620 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.598 1.603 dbcsr_complete_redistribute 325 12.2 0.553 0.585 1.503 1.581 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.426 1.439 mp_max_l 33 2.8 1.271 1.356 1.271 1.356 qs_create_task_list 11 7.9 0.001 0.001 1.230 1.332 generate_qs_task_list 11 8.9 0.379 0.448 1.230 1.331 mp_waitany 9240 13.8 1.067 1.267 1.067 1.267 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.186 1.226 multiply_cannon_metrocomm4 6165 15.4 0.017 0.018 0.474 1.197 jit_kernel_multiply 7 15.7 0.867 1.174 0.867 1.174 mp_irecv_dv 24056 15.7 0.451 1.161 0.451 1.161 mp_alltoall_z22v 1111 16.6 1.136 1.158 1.136 1.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=57.914000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.454545, yerr=11.420585 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 1.322684E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.035 88.576 88.577 qs_mol_dyn_low 1 2.0 0.003 0.003 88.225 88.255 qs_forces 11 3.9 0.002 0.002 88.128 88.131 qs_energies 11 4.9 0.001 0.002 83.992 83.996 scf_env_do_scf 11 5.9 0.000 0.001 73.901 73.901 velocity_verlet 10 3.0 0.001 0.001 56.726 56.735 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.200 44.201 init_scf_loop 11 6.9 0.000 0.000 29.622 29.624 dbcsr_multiply_generic 2055 12.4 0.119 0.121 29.400 29.521 qs_scf_new_mos 99 7.5 0.001 0.001 27.539 27.625 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.538 27.624 prepare_preconditioner 11 7.9 0.000 0.000 27.566 27.579 make_preconditioner 11 8.9 0.000 0.000 27.566 27.579 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.646 27.023 ot_scf_mini 99 9.5 0.002 0.002 25.752 25.820 multiply_cannon 2055 13.4 0.349 0.369 21.944 22.743 multiply_cannon_loop 2055 14.4 0.342 0.346 20.043 20.379 cp_fm_upper_to_full 70 14.2 12.763 18.392 12.763 18.392 ot_mini 99 10.5 0.001 0.001 14.142 14.210 rebuild_ks_matrix 110 8.3 0.001 0.001 13.398 13.475 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.398 13.475 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.162 12.232 dbcsr_complete_redistribute 325 12.2 1.013 1.035 7.624 10.871 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.525 9.772 multiply_cannon_multrec 8220 15.4 4.370 4.567 9.559 9.646 mp_waitall_1 87304 16.6 8.434 9.550 8.434 9.550 qs_ot_get_derivative 99 11.5 0.001 0.001 9.368 9.436 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.904 9.128 mp_alltoall_i22 605 13.7 5.534 8.791 5.534 8.791 sum_up_and_integrate 110 10.3 0.150 0.152 6.748 6.767 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.617 6.654 calculate_rho_elec 110 8.6 0.227 0.227 6.617 6.653 integrate_v_rspace 110 11.3 0.003 0.003 6.598 6.618 qs_ot_get_p 110 10.4 0.001 0.001 6.310 6.394 make_m2s 4110 13.4 0.043 0.043 5.524 6.114 cp_fm_cholesky_invert 11 10.9 6.094 6.099 6.094 6.099 make_images 4110 14.4 0.873 0.919 5.337 5.925 init_scf_run 11 5.9 0.000 0.001 5.798 5.799 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.798 5.798 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.123 5.428 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.846 5.382 apply_single 110 13.6 0.000 0.000 4.846 5.382 dbcsr_mm_accdrv_process 11614 15.7 3.188 3.701 5.046 5.327 ot_diis_step 99 11.5 0.015 0.016 4.747 4.747 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.576 4.582 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.114 4.115 multiply_cannon_sync_h2d 8220 15.4 3.952 3.956 3.952 3.956 pw_transfer 1331 11.6 0.074 0.074 3.923 3.927 hybrid_alltoall_any 4261 16.3 0.255 0.552 3.079 3.855 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.807 3.810 make_images_data 4110 15.4 0.042 0.044 3.075 3.795 grid_integrate_task_list 110 12.3 3.658 3.717 3.658 3.717 qs_energies_init_hamiltonians 11 5.9 0.010 0.011 3.670 3.671 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.468 3.468 cp_fm_diag_elpa_base 48 14.0 2.933 3.133 3.466 3.466 wfi_extrapolate 11 7.9 0.001 0.001 3.444 3.444 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.910 3.378 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.285 3.291 calculate_dm_sparse 110 9.5 0.001 0.001 3.169 3.189 density_rs2pw 110 9.6 0.004 0.004 3.084 3.098 fft3d_ps 1111 14.6 1.268 1.278 3.015 3.021 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.956 2.957 grid_collocate_task_list 110 9.6 2.624 2.645 2.624 2.645 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.578 2.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.395 2.418 mp_alltoall_d11v 2046 13.8 2.257 2.314 2.257 2.314 rs_pw_transfer 902 11.9 0.010 0.010 2.216 2.257 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.198 2.246 calculate_first_density_matrix 1 7.0 0.000 0.000 2.244 2.245 cp_fm_cholesky_decompose 22 10.9 2.220 2.240 2.220 2.240 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.217 potential_pw2rs 110 12.3 0.021 0.022 2.155 2.166 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.909 1.955 qs_create_task_list 11 7.9 0.001 0.001 1.906 1.952 generate_qs_task_list 11 8.9 0.739 0.791 1.906 1.951 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.841 1.850 mp_allgather_i34 2055 14.4 0.682 1.801 0.682 1.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.577000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1192.363636, yerr=55.075605 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 628.092928E+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 3864914. 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 0.024 0.047 204.850 204.860 qs_mol_dyn_low 1 2.0 0.003 0.003 204.088 204.327 qs_forces 11 3.9 0.005 0.006 203.861 203.862 qs_energies 11 4.9 0.003 0.008 198.336 198.350 scf_env_do_scf 11 5.9 0.001 0.002 181.687 181.692 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 160.748 160.750 dbcsr_multiply_generic 2507 12.6 0.177 0.183 124.487 124.859 velocity_verlet 10 3.0 0.001 0.001 123.456 123.459 qs_scf_new_mos 117 7.6 0.001 0.001 121.588 121.751 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.587 121.750 ot_scf_mini 117 9.6 0.003 0.004 114.941 115.118 multiply_cannon 2507 13.6 0.236 0.242 100.643 102.101 multiply_cannon_loop 2507 14.6 2.149 2.233 98.422 99.863 ot_mini 117 10.6 0.001 0.001 65.920 66.072 multiply_cannon_multrec 60168 15.6 32.994 35.008 41.321 43.199 qs_ot_get_derivative 117 11.6 0.001 0.001 41.293 41.436 rebuild_ks_matrix 128 8.3 0.001 0.001 33.392 33.617 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.018 33.391 33.616 mp_waitall_1 291448 16.2 29.062 31.706 29.062 31.706 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.035 30.227 multiply_cannon_sync_h2d 60168 15.6 27.169 28.493 27.169 28.493 qs_ot_get_p 128 10.4 0.001 0.001 27.114 27.360 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.199 24.738 apply_single 128 13.6 0.001 0.001 24.198 24.737 ot_diis_step 117 11.6 0.007 0.008 24.274 24.276 init_scf_loop 11 6.9 0.001 0.003 20.858 20.859 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.356 20.454 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.242 19.362 multiply_cannon_metrocomm3 60168 15.6 0.116 0.120 15.835 17.857 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.755 17.756 prepare_preconditioner 11 7.9 0.000 0.000 16.251 16.307 make_preconditioner 11 8.9 0.000 0.003 16.251 16.307 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.466 15.672 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.790 14.798 cp_fm_redistribute_end 83 14.4 11.653 14.711 11.666 14.714 cp_fm_diag_elpa_base 83 14.4 3.005 14.466 3.037 14.586 make_m2s 5014 13.6 0.103 0.110 14.019 14.338 make_images 5014 14.6 0.402 0.422 13.840 14.163 sum_up_and_integrate 128 10.3 0.089 0.107 14.121 14.140 integrate_v_rspace 128 11.3 0.003 0.004 14.032 14.054 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.007 13.115 calculate_rho_elec 128 8.7 0.044 0.063 13.006 13.115 init_scf_run 11 5.9 0.000 0.001 12.459 12.459 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.458 12.459 mp_sum_l 7870 13.0 8.588 10.030 8.588 10.030 cp_fm_cholesky_invert 11 10.9 9.301 9.309 9.301 9.309 wfi_extrapolate 11 7.9 0.001 0.001 9.160 9.161 calculate_dm_sparse 128 9.5 0.001 0.001 8.465 8.537 dbcsr_mm_accdrv_process 124484 16.2 3.294 3.434 7.894 8.409 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.047 8.227 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.910 8.050 make_images_data 5014 15.6 0.067 0.074 6.911 7.810 multiply_cannon_metrocomm1 60168 15.6 0.083 0.089 5.984 7.664 grid_integrate_task_list 128 12.3 7.126 7.506 7.126 7.506 density_rs2pw 128 9.7 0.006 0.007 6.762 7.319 pw_transfer 1547 11.6 0.075 0.106 7.036 7.310 hybrid_alltoall_any 5200 16.5 0.293 2.273 6.030 7.251 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.832 7.081 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.660 6.669 rs_pw_transfer 1046 11.9 0.016 0.018 5.916 6.428 fft_wrap_pw1pw2_140 523 13.2 0.443 0.504 5.803 5.987 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.785 5.867 fft3d_ps 1291 14.7 2.092 2.569 5.626 5.817 mp_alltoall_d11v 2415 14.1 4.327 5.816 4.327 5.816 grid_collocate_task_list 128 9.7 4.741 5.034 4.741 5.034 cp_fm_cholesky_decompose 22 10.9 4.719 4.734 4.719 4.734 potential_pw2rs 128 12.3 0.009 0.010 4.576 4.603 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.860000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.727273, yerr=6.823634 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 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 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 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 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 825.368576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56635. MP_Allreduce 11357 942. MP_Sync 172 MP_Alltoall 1983 5062318. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 84 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 0.035 0.070 191.644 191.645 qs_mol_dyn_low 1 2.0 0.003 0.003 191.108 191.218 qs_forces 11 3.9 0.003 0.003 190.897 190.898 qs_energies 11 4.9 0.005 0.028 184.192 184.204 scf_env_do_scf 11 5.9 0.001 0.001 167.685 167.695 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 134.661 134.664 velocity_verlet 10 3.0 0.001 0.001 120.830 120.834 dbcsr_multiply_generic 2529 12.6 0.187 0.197 98.126 99.330 qs_scf_new_mos 118 7.6 0.001 0.001 95.563 96.042 qs_scf_loop_do_ot 118 8.6 0.001 0.001 95.562 96.042 ot_scf_mini 118 9.6 0.004 0.005 90.721 91.250 multiply_cannon 2529 13.6 0.482 0.532 77.853 82.121 multiply_cannon_loop 2529 14.6 1.261 1.303 74.549 77.212 ot_mini 118 10.6 0.001 0.001 50.301 50.786 mp_waitall_1 228724 16.4 25.385 39.325 25.385 39.325 multiply_cannon_multrec 30348 15.6 22.027 26.622 31.672 36.476 rebuild_ks_matrix 129 8.3 0.001 0.001 33.004 33.488 qs_ks_build_kohn_sham_matrix 129 9.3 0.027 0.057 33.003 33.487 init_scf_loop 11 6.9 0.001 0.004 32.932 32.933 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.792 30.252 prepare_preconditioner 11 7.9 0.000 0.000 28.508 28.578 make_preconditioner 11 8.9 0.000 0.001 28.508 28.578 qs_ot_get_derivative 118 11.6 0.001 0.002 28.046 28.572 multiply_cannon_metrocomm3 30348 15.6 0.093 0.099 16.025 28.499 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.196 27.712 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.290 23.376 apply_single 129 13.6 0.001 0.001 22.289 23.376 qs_ot_get_p 129 10.4 0.001 0.001 22.037 22.650 ot_diis_step 118 11.6 0.015 0.015 22.090 22.092 multiply_cannon_sync_h2d 30348 15.6 19.321 21.795 19.321 21.795 qs_ot_p2m_diag 84 11.4 0.190 0.219 17.240 17.281 cp_fm_cholesky_invert 11 10.9 16.589 16.602 16.589 16.602 cp_dbcsr_syevd 84 12.4 0.005 0.006 16.061 16.062 make_m2s 5058 13.6 0.090 0.094 14.507 16.052 make_images 5058 14.6 1.152 1.348 14.296 15.843 sum_up_and_integrate 129 10.3 0.118 0.134 14.446 14.475 integrate_v_rspace 129 11.3 0.003 0.003 14.328 14.364 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.000 13.036 calculate_rho_elec 129 8.7 0.088 0.106 13.000 13.035 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.925 12.938 cp_fm_redistribute_end 84 14.4 7.513 12.857 7.523 12.859 cp_fm_diag_elpa_base 84 14.4 5.095 12.393 5.313 12.760 init_scf_run 11 5.9 0.000 0.001 11.584 11.588 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.584 11.588 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.118 11.489 multiply_cannon_metrocomm4 27819 15.6 0.098 0.113 3.825 10.743 make_images_data 5058 15.6 0.065 0.073 8.857 10.637 mp_irecv_dv 70084 16.3 3.629 10.354 3.629 10.354 hybrid_alltoall_any 5245 16.5 0.345 1.511 7.591 9.884 dbcsr_mm_accdrv_process 62758 16.2 4.497 5.304 9.103 9.705 wfi_extrapolate 11 7.9 0.001 0.001 8.373 8.373 pw_transfer 1559 11.6 0.085 0.098 7.853 7.917 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.627 7.688 grid_integrate_task_list 129 12.3 7.232 7.621 7.232 7.621 cp_fm_cholesky_decompose 22 10.9 7.043 7.129 7.043 7.129 density_rs2pw 129 9.7 0.006 0.006 6.750 7.086 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.228 6.968 fft_wrap_pw1pw2_140 527 13.2 0.473 0.521 6.665 6.714 calculate_dm_sparse 129 9.5 0.001 0.002 6.491 6.616 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.253 6.263 rs_pw_transfer 1054 12.0 0.015 0.017 5.652 6.043 fft3d_ps 1301 14.7 2.796 2.941 5.986 6.026 mp_sum_l 7936 13.1 4.009 5.969 4.009 5.969 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.394 5.447 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.256 5.428 grid_collocate_task_list 129 9.7 4.948 5.334 4.948 5.334 mp_allgather_i34 2529 14.6 1.867 5.212 1.867 5.212 potential_pw2rs 129 12.3 0.014 0.016 4.860 4.881 mp_alltoall_d11v 2429 14.1 4.160 4.601 4.160 4.601 mp_sum_d 4499 12.2 2.679 4.203 2.679 4.203 dbcsr_complete_redistribute 397 12.7 0.816 0.955 3.269 4.120 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.645000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.272727, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+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 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 947.585024E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.089 175.716 175.741 qs_mol_dyn_low 1 2.0 0.003 0.003 175.226 175.240 qs_forces 11 3.9 0.026 0.049 174.854 174.861 qs_energies 11 4.9 0.003 0.014 168.335 168.366 scf_env_do_scf 11 5.9 0.003 0.023 152.872 152.873 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.716 117.720 velocity_verlet 10 3.0 0.001 0.001 112.664 112.673 dbcsr_multiply_generic 2507 12.6 0.181 0.184 81.625 82.717 qs_scf_new_mos 117 7.6 0.001 0.001 81.188 81.576 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.187 81.575 ot_scf_mini 117 9.6 0.007 0.030 77.076 77.471 multiply_cannon 2507 13.6 0.499 0.520 61.615 66.185 multiply_cannon_loop 2507 14.6 0.859 0.887 58.764 61.284 ot_mini 117 10.6 0.001 0.001 42.456 42.842 mp_waitall_1 178456 16.5 25.844 35.367 25.844 35.367 init_scf_loop 11 6.9 0.002 0.017 35.053 35.054 prepare_preconditioner 11 7.9 0.000 0.000 31.067 31.118 make_preconditioner 11 8.9 0.000 0.001 31.067 31.118 rebuild_ks_matrix 128 8.3 0.001 0.001 30.027 30.577 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.026 30.576 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.624 30.055 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.109 27.607 multiply_cannon_multrec 20056 15.6 13.380 16.480 21.997 25.128 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.579 24.944 qs_ot_get_derivative 117 11.6 0.001 0.002 22.534 22.919 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.065 21.048 apply_single 128 13.6 0.001 0.001 20.065 21.048 qs_ot_get_p 128 10.4 0.001 0.001 19.989 20.419 ot_diis_step 117 11.6 0.018 0.018 19.820 19.820 multiply_cannon_sync_h2d 20056 15.6 14.182 16.011 14.182 16.011 make_m2s 5014 13.6 0.081 0.085 15.056 15.853 qs_ot_p2m_diag 83 11.4 0.266 0.273 15.825 15.834 make_images 5014 14.6 1.160 1.265 14.823 15.620 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.826 14.827 cp_fm_cholesky_invert 11 10.9 14.379 14.388 14.379 14.388 sum_up_and_integrate 128 10.3 0.133 0.148 14.069 14.097 integrate_v_rspace 128 11.3 0.003 0.004 13.935 13.970 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.883 12.912 calculate_rho_elec 128 8.7 0.131 0.146 12.883 12.912 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.682 11.683 cp_fm_redistribute_end 83 14.4 4.389 11.619 4.406 11.625 cp_fm_diag_elpa_base 83 14.4 6.790 11.056 7.199 11.531 make_images_data 5014 15.6 0.060 0.067 9.456 10.649 init_scf_run 11 5.9 0.000 0.001 10.399 10.399 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.399 10.399 hybrid_alltoall_any 5200 16.5 0.432 1.977 8.110 9.823 multiply_cannon_metrocomm4 17549 15.6 0.059 0.069 3.519 9.490 mp_irecv_dv 50230 16.2 3.398 9.242 3.398 9.242 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.848 9.141 dbcsr_mm_accdrv_process 41502 16.2 4.403 5.195 8.072 8.182 pw_transfer 1547 11.6 0.084 0.104 7.709 7.811 grid_integrate_task_list 128 12.3 7.305 7.635 7.305 7.635 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.485 7.590 cp_fm_upper_to_full 105 14.5 5.851 7.533 5.851 7.533 cp_fm_cholesky_decompose 22 10.9 7.333 7.399 7.333 7.399 wfi_extrapolate 11 7.9 0.001 0.001 7.344 7.344 density_rs2pw 128 9.7 0.005 0.006 6.378 6.808 fft_wrap_pw1pw2_140 523 13.2 0.478 0.531 6.596 6.706 dbcsr_complete_redistribute 395 12.7 1.172 1.206 4.676 6.466 fft3d_ps 1291 14.7 2.695 2.898 5.749 5.805 calculate_dm_sparse 128 9.5 0.001 0.001 5.700 5.781 grid_collocate_task_list 128 9.7 5.076 5.552 5.076 5.552 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.531 5.535 rs_pw_transfer 1046 11.9 0.014 0.014 4.948 5.380 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.493 5.279 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.520 5.169 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.627 4.778 mp_alltoall_d11v 2415 14.1 4.254 4.691 4.254 4.691 mp_sum_l 7870 13.0 3.165 4.541 3.165 4.541 mp_allgather_i34 2507 14.6 1.377 4.511 1.377 4.511 potential_pw2rs 128 12.3 0.020 0.022 4.461 4.476 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.424 4.188 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.022 4.054 mp_alltoall_i22 716 14.1 1.980 3.914 1.980 3.914 qs_energies_init_hamiltonians 11 5.9 0.010 0.017 3.806 3.829 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.741000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=900.545455, yerr=5.532772 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.130230E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57194. MP_Allreduce 11251 1067. MP_Sync 170 MP_Alltoall 1712 12503107. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.043 185.236 185.237 qs_mol_dyn_low 1 2.0 0.003 0.003 184.782 184.895 qs_forces 11 3.9 0.003 0.003 184.609 184.615 qs_energies 11 4.9 0.002 0.002 177.613 177.623 scf_env_do_scf 11 5.9 0.001 0.001 161.212 161.225 velocity_verlet 10 3.0 0.001 0.001 123.028 123.033 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.157 115.158 qs_scf_new_mos 117 7.6 0.001 0.001 79.736 80.090 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.735 80.089 dbcsr_multiply_generic 2507 12.6 0.189 0.194 78.917 79.668 ot_scf_mini 117 9.6 0.003 0.003 75.262 75.555 multiply_cannon 2507 13.6 0.555 0.588 54.399 57.044 multiply_cannon_loop 2507 14.6 1.185 1.212 50.813 52.605 init_scf_loop 11 6.9 0.000 0.000 45.935 45.935 ot_mini 117 10.6 0.001 0.001 42.000 42.308 prepare_preconditioner 11 7.9 0.000 0.000 41.852 41.885 make_preconditioner 11 8.9 0.000 0.000 41.852 41.885 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.462 40.448 multiply_cannon_multrec 30084 15.6 13.973 18.727 25.844 30.548 rebuild_ks_matrix 128 8.3 0.001 0.001 28.872 29.147 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 28.871 29.147 mp_waitall_1 153770 16.5 17.558 27.529 17.558 27.529 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.101 26.339 qs_ot_get_derivative 117 11.6 0.001 0.002 22.319 22.627 make_m2s 5014 13.6 0.096 0.099 20.260 21.301 make_images 5014 14.6 1.962 2.234 19.955 20.996 qs_ot_get_p 128 10.4 0.001 0.001 19.515 19.871 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.131 19.751 apply_single 128 13.6 0.001 0.001 19.131 19.750 ot_diis_step 117 11.6 0.017 0.018 19.564 19.565 cp_fm_upper_to_full 105 14.7 10.977 16.259 10.977 16.259 cp_fm_cholesky_invert 11 10.9 15.977 15.986 15.977 15.986 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.508 15.560 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.589 15.018 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.229 14.231 sum_up_and_integrate 128 10.3 0.140 0.153 13.947 13.976 integrate_v_rspace 128 11.3 0.003 0.004 13.806 13.840 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.990 13.028 calculate_rho_elec 128 8.7 0.175 0.190 12.990 13.027 dbcsr_complete_redistribute 395 12.7 1.522 1.633 8.969 12.773 make_images_data 5014 15.6 0.063 0.067 10.891 12.640 multiply_cannon_sync_h2d 30084 15.6 11.622 12.526 11.622 12.526 dbcsr_mm_accdrv_process 62264 16.2 7.289 8.351 11.444 11.910 hybrid_alltoall_any 5200 16.5 0.527 2.219 9.859 11.763 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.618 11.421 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.150 11.152 cp_fm_redistribute_end 83 14.4 1.897 11.073 1.910 11.079 cp_fm_diag_elpa_base 83 14.4 8.564 10.481 9.134 10.970 init_scf_run 11 5.9 0.000 0.001 10.535 10.536 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.534 10.536 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.367 10.030 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.250 9.469 mp_alltoall_i22 716 14.1 5.561 9.278 5.561 9.278 pw_transfer 1547 11.6 0.085 0.101 7.818 7.896 grid_integrate_task_list 128 12.3 7.496 7.822 7.496 7.822 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.593 7.674 wfi_extrapolate 11 7.9 0.001 0.001 7.501 7.501 cp_fm_cholesky_decompose 22 10.9 7.357 7.459 7.357 7.459 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.849 7.451 mp_irecv_dv 76098 16.2 2.703 7.192 2.703 7.192 fft_wrap_pw1pw2_140 523 13.2 0.483 0.488 6.724 6.811 density_rs2pw 128 9.7 0.005 0.006 6.170 6.451 calculate_dm_sparse 128 9.5 0.001 0.001 6.132 6.199 fft3d_ps 1291 14.7 2.801 2.868 5.825 5.893 grid_collocate_task_list 128 9.7 5.209 5.614 5.209 5.614 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.379 5.423 mp_alltoall_d11v 2415 14.1 4.739 5.225 4.739 5.225 rs_pw_transfer 1046 11.9 0.013 0.015 4.484 4.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.335 4.397 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.384 4.384 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.287 4.372 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.177 4.230 potential_pw2rs 128 12.3 0.022 0.023 4.196 4.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.237000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1069.363636, yerr=13.425029 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 5.910120E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.497076E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592466352 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57602. MP_Allreduce 11109 1171. MP_Sync 88 MP_Alltoall 1724 18848034. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.046 171.951 171.952 qs_mol_dyn_low 1 2.0 0.003 0.003 171.437 171.520 qs_forces 11 3.9 0.003 0.003 171.143 171.147 qs_energies 11 4.9 0.001 0.002 163.880 163.890 scf_env_do_scf 11 5.9 0.001 0.001 146.896 146.908 velocity_verlet 10 3.0 0.001 0.001 112.857 112.869 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 111.210 111.211 dbcsr_multiply_generic 2529 12.6 0.184 0.190 74.900 75.494 qs_scf_new_mos 118 7.6 0.001 0.001 75.360 75.482 qs_scf_loop_do_ot 118 8.6 0.001 0.001 75.359 75.482 ot_scf_mini 118 9.6 0.004 0.004 70.921 71.047 multiply_cannon 2529 13.6 0.589 0.622 54.905 59.099 multiply_cannon_loop 2529 14.6 0.452 0.469 50.324 51.600 ot_mini 118 10.6 0.001 0.001 40.081 40.207 init_scf_loop 11 6.9 0.000 0.000 35.528 35.529 mp_waitall_1 130746 16.6 26.723 34.909 26.723 34.909 prepare_preconditioner 11 7.9 0.000 0.000 31.709 31.733 make_preconditioner 11 8.9 0.000 0.000 31.709 31.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.577 29.855 rebuild_ks_matrix 129 8.3 0.001 0.001 28.661 28.844 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 28.661 28.844 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.225 26.391 multiply_cannon_multrec 10116 15.6 10.467 14.774 17.774 22.610 multiply_cannon_metrocomm3 10116 15.6 0.023 0.024 12.745 20.189 qs_ot_get_derivative 118 11.6 0.002 0.002 19.961 20.086 ot_diis_step 118 11.6 0.020 0.021 20.052 20.052 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.584 19.939 apply_single 129 13.6 0.001 0.001 19.584 19.939 make_m2s 5058 13.6 0.067 0.072 16.276 18.461 cp_fm_cholesky_invert 11 10.9 18.182 18.188 18.182 18.188 make_images 5058 14.6 2.297 2.658 15.968 18.154 qs_ot_get_p 129 10.4 0.001 0.001 17.786 17.929 sum_up_and_integrate 129 10.3 0.184 0.194 14.041 14.094 qs_ot_p2m_diag 84 11.4 0.502 0.507 14.006 14.023 integrate_v_rspace 129 11.3 0.004 0.004 13.857 13.908 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.134 13.183 calculate_rho_elec 129 8.7 0.260 0.271 13.134 13.182 cp_dbcsr_syevd 84 12.4 0.005 0.005 12.848 12.850 make_images_data 5058 15.6 0.053 0.061 9.978 12.329 multiply_cannon_sync_h2d 10116 15.6 11.571 12.249 11.571 12.249 hybrid_alltoall_any 5245 16.5 0.854 3.864 9.782 12.248 init_scf_run 11 5.9 0.000 0.001 10.246 10.246 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.246 10.246 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.813 9.815 cp_fm_diag_elpa_base 84 14.4 9.555 9.633 9.800 9.800 multiply_cannon_metrocomm1 10116 15.6 0.028 0.029 4.579 8.537 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.089 8.168 grid_integrate_task_list 129 12.3 7.817 8.155 7.817 8.155 cp_fm_cholesky_decompose 22 10.9 7.844 7.939 7.844 7.939 pw_transfer 1559 11.6 0.084 0.092 7.700 7.727 dbcsr_mm_accdrv_process 20934 16.1 2.616 3.453 6.938 7.569 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.478 7.502 mp_allgather_i34 2529 14.6 2.777 7.287 2.777 7.287 wfi_extrapolate 11 7.9 0.001 0.001 7.247 7.247 fft_wrap_pw1pw2_140 527 13.2 0.507 0.528 6.604 6.631 density_rs2pw 129 9.7 0.005 0.006 5.902 6.268 calculate_dm_sparse 129 9.5 0.001 0.001 6.088 6.193 multiply_cannon_metrocomm4 7587 15.6 0.023 0.026 1.887 6.050 mp_irecv_dv 29102 15.9 1.850 5.957 1.850 5.957 grid_collocate_task_list 129 9.7 5.528 5.777 5.528 5.777 dbcsr_complete_redistribute 397 12.7 2.170 2.268 5.384 5.758 fft3d_ps 1301 14.7 2.743 2.813 5.625 5.660 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.263 5.270 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 5.211 5.212 mp_alltoall_d11v 2429 14.1 4.793 5.208 4.793 5.208 rs_pw_transfer 1054 12.0 0.013 0.013 4.008 4.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.073 4.106 potential_pw2rs 129 12.3 0.027 0.028 3.982 3.998 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.882 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.518 3.841 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.739 3.773 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.588 3.667 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.465 3.504 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=171.952000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1392.363636, yerr=43.187272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.084866E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3949 59908. MP_Allreduce 10870 1518. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.049 288.409 288.409 qs_mol_dyn_low 1 2.0 0.003 0.003 287.889 287.901 qs_forces 11 3.9 0.003 0.003 287.544 287.549 qs_energies 11 4.9 0.002 0.002 278.990 278.997 scf_env_do_scf 11 5.9 0.001 0.001 257.284 257.299 velocity_verlet 10 3.0 0.001 0.001 208.795 208.834 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.429 132.430 init_scf_loop 11 6.9 0.000 0.000 124.593 124.594 prepare_preconditioner 11 7.9 0.000 0.000 119.938 119.976 make_preconditioner 11 8.9 0.000 0.000 119.938 119.976 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.617 117.075 qs_scf_new_mos 116 7.6 0.001 0.001 90.447 90.580 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.446 90.579 ot_scf_mini 116 9.6 0.004 0.004 85.641 85.733 dbcsr_multiply_generic 2485 12.5 0.210 0.218 83.160 83.933 cp_fm_upper_to_full 104 14.8 53.096 76.359 53.096 76.359 multiply_cannon 2485 13.5 0.697 0.755 58.542 59.438 multiply_cannon_loop 2485 14.5 0.467 0.480 54.800 56.791 ot_mini 116 10.6 0.001 0.001 44.629 44.738 dbcsr_complete_redistribute 393 12.7 3.989 4.068 30.019 43.303 copy_fm_to_dbcsr 208 11.6 0.002 0.002 26.609 39.883 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.278 37.433 mp_alltoall_i22 712 14.1 22.110 35.204 22.110 35.204 rebuild_ks_matrix 127 8.3 0.001 0.001 33.320 33.463 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.319 33.463 cp_fm_cholesky_invert 11 10.9 33.401 33.407 33.401 33.407 mp_waitall_1 104546 16.7 28.706 33.176 28.706 33.176 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.067 31.201 qs_ot_get_p 127 10.4 0.001 0.001 25.642 25.750 qs_ot_get_derivative 116 11.6 0.002 0.002 23.930 24.025 qs_ot_p2m_diag 82 11.4 0.868 0.873 21.619 21.648 make_m2s 4970 13.5 0.075 0.078 19.895 20.978 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 19.034 20.847 ot_diis_step 116 11.6 0.022 0.022 20.658 20.658 make_images 4970 14.5 3.723 3.871 19.418 20.507 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.822 20.223 apply_single 127 13.6 0.001 0.001 19.822 20.223 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.925 19.926 multiply_cannon_multrec 9940 15.5 10.363 12.077 17.694 17.807 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.746 16.747 cp_fm_diag_elpa_base 82 14.4 12.440 14.011 16.742 16.742 sum_up_and_integrate 127 10.3 0.320 0.321 15.547 15.667 multiply_cannon_sync_h2d 9940 15.5 15.541 15.557 15.541 15.557 integrate_v_rspace 127 11.3 0.004 0.004 15.227 15.346 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.788 14.802 calculate_rho_elec 127 8.7 0.479 0.480 14.787 14.802 hybrid_alltoall_any 5155 16.4 1.293 3.026 10.943 12.859 make_images_data 4970 15.5 0.059 0.063 10.752 12.610 init_scf_run 11 5.9 0.000 0.001 11.757 11.757 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.756 11.757 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.297 9.366 dbcsr_mm_accdrv_process 20590 16.0 3.815 5.676 7.098 8.903 cp_fm_cholesky_decompose 22 10.9 8.797 8.831 8.797 8.831 wfi_extrapolate 11 7.9 0.001 0.001 8.706 8.706 grid_integrate_task_list 127 12.3 8.492 8.663 8.492 8.663 pw_transfer 1535 11.6 0.090 0.091 8.189 8.197 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.957 7.963 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.936 7.937 fft_wrap_pw1pw2_140 519 13.2 0.536 0.544 7.066 7.080 mp_alltoall_d11v 2401 14.1 6.890 7.031 6.890 7.031 calculate_dm_sparse 127 9.5 0.001 0.001 6.641 6.695 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.394 6.508 grid_collocate_task_list 127 9.7 6.288 6.316 6.288 6.316 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.102 6.186 fft3d_ps 1281 14.7 2.736 2.753 6.016 6.036 density_rs2pw 127 9.7 0.005 0.005 5.930 5.962 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=288.409000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2737.727273, yerr=177.268718 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 1.261101E+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 18848822. 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 0.026 0.093 84.673 84.674 qs_energies 1 2.0 0.000 0.000 84.184 84.190 ls_scf 1 3.0 0.001 0.012 83.289 83.295 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.433 72.614 multiply_cannon 111 7.7 0.017 0.020 55.835 57.229 multiply_cannon_loop 111 8.7 0.211 0.223 52.409 54.060 ls_scf_main 1 4.0 0.000 0.003 52.120 52.121 density_matrix_trs4 2 5.0 0.002 0.003 46.679 46.756 ls_scf_init_scf 1 4.0 0.000 0.001 28.124 28.126 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.073 27.120 mp_waitall_1 11316 10.9 22.413 25.022 22.413 25.022 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.980 25.000 multiply_cannon_multrec 2664 9.7 8.155 8.894 15.452 17.156 multiply_cannon_sync_h2d 2664 9.7 13.585 15.802 13.585 15.802 make_m2s 222 7.7 0.008 0.011 12.969 13.508 make_images 222 8.7 0.099 0.110 12.946 13.488 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.694 11.606 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.497 8.530 make_images_data 222 9.7 0.004 0.005 7.549 8.080 dbcsr_mm_accdrv_process 4760 10.4 0.507 0.607 6.916 7.887 hybrid_alltoall_any 227 10.6 0.215 1.827 6.477 7.820 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.094 6.209 7.094 calculate_norms 4752 9.8 5.514 6.174 5.514 6.174 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.061 5.189 mp_sum_l 807 5.4 3.132 4.552 3.132 4.552 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.074 3.715 mp_irecv_dv 6231 10.9 2.056 3.686 2.056 3.686 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.315 3.389 make_images_sizes 222 9.7 0.000 0.000 0.675 3.321 mp_alltoall_i44 222 10.7 0.674 3.320 0.674 3.320 arnoldi_extremal 4 6.8 0.000 0.000 3.181 3.217 arnoldi_normal_ev 4 7.8 0.001 0.003 3.181 3.217 build_subspace 16 8.4 0.009 0.012 3.086 3.088 ls_scf_post 1 4.0 0.000 0.002 3.042 3.050 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.907 dbcsr_special_finalize 555 9.7 0.005 0.006 2.392 2.801 dbcsr_merge_single_wm 555 10.7 0.461 0.601 2.384 2.793 make_images_pack 222 9.7 2.206 2.629 2.208 2.631 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.318 2.552 dbcsr_sort_data 658 11.4 2.180 2.520 2.180 2.520 dbcsr_matrix_vector_mult_local 304 10.0 2.068 2.459 2.070 2.461 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.186 2.253 buffer_matrices_ensure_size 222 8.7 1.750 2.066 1.750 2.066 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.721 1.725 rebuild_ks_matrix 3 7.3 0.000 0.000 1.712 1.715 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.712 1.715 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.674000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 2.104111E+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 0.056 0.103 90.854 90.857 qs_energies 1 2.0 0.000 0.000 90.264 90.290 ls_scf 1 3.0 0.000 0.001 88.863 88.891 dbcsr_multiply_generic 111 6.7 0.016 0.020 74.868 75.204 multiply_cannon 111 7.7 0.028 0.039 52.945 56.837 ls_scf_main 1 4.0 0.002 0.028 54.418 54.428 multiply_cannon_loop 111 8.7 0.117 0.124 49.686 52.856 density_matrix_trs4 2 5.0 0.002 0.003 48.780 48.990 ls_scf_init_scf 1 4.0 0.000 0.002 30.902 30.903 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.782 29.856 mp_waitall_1 9246 10.9 20.795 29.448 20.795 29.448 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.162 27.176 multiply_cannon_multrec 1332 9.7 13.131 16.578 22.325 26.896 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.510 20.054 make_m2s 222 7.7 0.006 0.008 15.282 15.849 make_images 222 8.7 1.578 1.965 15.252 15.820 dbcsr_mm_accdrv_process 4041 10.4 0.274 0.454 8.792 10.335 dbcsr_mm_accdrv_process_sort 4041 11.4 8.386 9.886 8.386 9.886 make_images_data 222 9.7 0.004 0.006 8.710 9.536 hybrid_alltoall_any 227 10.6 0.533 2.489 8.243 9.026 mp_sum_l 807 5.4 5.332 7.948 5.332 7.948 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.199 7.709 mp_irecv_dv 3311 11.0 3.179 7.647 3.179 7.647 calculate_norms 2376 9.8 6.024 6.734 6.024 6.734 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.153 6.633 multiply_cannon_sync_h2d 1332 9.7 4.849 5.944 4.849 5.944 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.223 arnoldi_extremal 4 6.8 0.000 0.002 4.611 4.634 arnoldi_normal_ev 4 7.8 0.001 0.007 4.611 4.634 build_subspace 16 8.4 0.014 0.021 4.349 4.353 ls_scf_post 1 4.0 0.002 0.012 3.543 3.570 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.123 3.349 ls_scf_store_result 1 5.0 0.000 0.000 3.225 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.735 3.215 2.737 3.217 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.207 2.626 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.508 2.599 make_images_pack 222 9.7 2.027 2.407 2.030 2.409 mp_allgather_i34 111 8.7 0.958 2.279 0.958 2.279 dbcsr_sort_data 436 11.2 1.846 2.107 1.846 2.107 dbcsr_data_new 4174 10.1 1.618 1.836 1.618 1.836 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.822 1.824 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.857000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1709.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 2.727334E+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 0.032 0.048 93.453 93.455 qs_energies 1 2.0 0.000 0.000 92.863 92.866 ls_scf 1 3.0 0.000 0.001 91.475 91.479 dbcsr_multiply_generic 111 6.7 0.015 0.018 76.353 76.651 ls_scf_main 1 4.0 0.000 0.002 57.367 57.373 multiply_cannon 111 7.7 0.044 0.109 52.608 56.662 multiply_cannon_loop 111 8.7 0.100 0.106 48.978 52.471 density_matrix_trs4 2 5.0 0.002 0.003 51.435 51.569 mp_waitall_1 7374 11.0 23.634 33.311 23.634 33.311 ls_scf_init_scf 1 4.0 0.000 0.001 30.484 30.489 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.374 29.429 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.004 27.019 multiply_cannon_multrec 888 9.7 12.651 15.839 21.142 24.925 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.905 22.177 make_m2s 222 7.7 0.006 0.007 17.092 18.385 make_images 222 8.7 1.971 2.270 17.054 18.345 make_images_data 222 9.7 0.003 0.004 9.773 10.889 hybrid_alltoall_any 227 10.6 0.619 2.866 9.388 10.697 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.415 8.026 9.238 dbcsr_mm_accdrv_process_sort 3754 11.4 7.657 8.823 7.657 8.823 mp_sum_l 807 5.4 5.091 8.710 5.091 8.710 multiply_cannon_sync_h2d 888 9.7 6.067 7.169 6.067 7.169 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.412 7.039 mp_irecv_dv 2335 11.1 2.397 6.996 2.397 6.996 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.902 6.877 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.574 6.549 arnoldi_extremal 4 6.8 0.000 0.000 5.083 5.103 arnoldi_normal_ev 4 7.8 0.002 0.007 5.083 5.103 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.917 5.077 build_subspace 16 8.4 0.014 0.020 4.774 4.779 calculate_norms 1584 9.8 4.265 4.613 4.265 4.613 mp_allgather_i34 111 8.7 1.474 3.897 1.474 3.897 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.433 3.766 ls_scf_post 1 4.0 0.000 0.001 3.624 3.628 dbcsr_matrix_vector_mult_local 304 10.0 3.009 3.572 3.011 3.574 ls_scf_store_result 1 5.0 0.000 0.000 3.346 3.450 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.801 2.908 make_images_sizes 222 9.7 0.000 0.000 0.995 2.343 mp_alltoall_i44 222 10.7 0.994 2.343 0.994 2.343 dbcsr_sort_data 325 11.1 1.878 2.168 1.878 2.168 make_images_pack 222 9.7 1.819 2.133 1.821 2.135 dbcsr_data_release 9322 10.9 1.305 1.999 1.305 1.999 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.884 1.889 rebuild_ks_matrix 3 7.3 0.000 0.000 1.865 1.870 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.010 1.865 1.870 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.455000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2180.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 3.316388E+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 0.035 0.052 96.954 96.955 qs_energies 1 2.0 0.000 0.000 96.369 96.374 ls_scf 1 3.0 0.000 0.000 94.735 94.741 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.662 78.898 ls_scf_main 1 4.0 0.000 0.000 58.942 58.942 multiply_cannon 111 7.7 0.051 0.142 51.657 56.127 density_matrix_trs4 2 5.0 0.002 0.003 52.878 53.002 multiply_cannon_loop 111 8.7 0.115 0.127 46.581 50.092 ls_scf_init_scf 1 4.0 0.000 0.000 32.555 32.556 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.397 31.487 mp_waitall_1 6438 11.0 22.830 29.518 22.830 29.518 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.944 28.961 multiply_cannon_multrec 1332 9.7 14.227 17.395 22.063 25.086 make_m2s 222 7.7 0.007 0.008 21.189 22.625 make_images 222 8.7 3.131 3.592 21.138 22.576 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.210 17.079 make_images_data 222 9.7 0.004 0.004 11.864 13.622 hybrid_alltoall_any 227 10.6 0.799 3.817 11.173 12.942 dbcsr_mm_accdrv_process 3641 10.4 0.211 0.378 7.468 8.992 dbcsr_mm_accdrv_process_sort 3641 11.4 7.095 8.574 7.095 8.574 mp_sum_l 807 5.4 4.228 7.446 4.228 7.446 multiply_cannon_sync_h2d 1332 9.7 5.516 6.376 5.516 6.376 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.072 6.092 mp_irecv_dv 3229 10.9 2.050 6.018 2.050 6.018 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.208 5.846 arnoldi_extremal 4 6.8 0.000 0.000 5.212 5.230 arnoldi_normal_ev 4 7.8 0.001 0.004 5.212 5.229 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.584 5.211 build_subspace 16 8.4 0.014 0.021 4.876 4.884 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.579 4.757 mp_allgather_i34 111 8.7 2.216 4.580 2.216 4.580 calculate_norms 2376 9.8 4.193 4.563 4.193 4.563 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.586 3.867 dbcsr_matrix_vector_mult_local 304 10.0 3.177 3.679 3.179 3.681 dbcsr_sort_data 658 11.4 3.031 3.491 3.031 3.491 ls_scf_post 1 4.0 0.000 0.000 3.239 3.244 dbcsr_special_finalize 555 9.7 0.006 0.007 2.766 3.228 dbcsr_merge_single_wm 555 10.7 0.530 0.668 2.758 3.219 ls_scf_store_result 1 5.0 0.000 0.000 2.991 3.043 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.898 2.958 dbcsr_data_release 10477 10.7 1.587 2.461 1.587 2.461 dbcsr_finalize 304 7.8 0.049 0.061 1.804 1.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.955000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2672.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 4.637786E+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 0.040 0.057 99.157 99.158 qs_energies 1 2.0 0.000 0.000 98.411 98.415 ls_scf 1 3.0 0.000 0.000 96.498 96.506 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.175 78.390 ls_scf_main 1 4.0 0.000 0.000 62.154 62.154 multiply_cannon 111 7.7 0.080 0.220 55.351 60.657 density_matrix_trs4 2 5.0 0.002 0.003 55.100 55.185 multiply_cannon_loop 111 8.7 0.070 0.083 50.787 52.233 mp_waitall_1 5481 11.0 26.749 32.391 26.749 32.391 ls_scf_init_scf 1 4.0 0.000 0.000 30.745 30.750 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.575 29.614 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.413 27.422 multiply_cannon_multrec 444 9.7 14.040 16.512 21.057 24.531 make_m2s 222 7.7 0.004 0.005 17.823 20.330 make_images 222 8.7 3.718 4.434 17.762 20.269 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.265 16.452 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.146 14.734 make_images_data 222 9.7 0.003 0.004 10.138 12.529 hybrid_alltoall_any 227 10.6 0.789 3.788 9.815 12.333 multiply_cannon_sync_h2d 444 9.7 6.530 9.010 6.530 9.010 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.345 6.716 7.821 dbcsr_mm_accdrv_process_sort 3003 11.4 6.402 7.476 6.402 7.476 mp_allgather_i34 111 8.7 2.802 7.058 2.802 7.058 arnoldi_extremal 4 6.8 0.000 0.000 5.863 5.873 arnoldi_normal_ev 4 7.8 0.002 0.005 5.863 5.873 build_subspace 16 8.4 0.015 0.019 5.467 5.476 mp_sum_l 807 5.4 3.038 5.400 3.038 5.400 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.646 4.811 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.171 4.397 dbcsr_matrix_vector_mult_local 304 10.0 3.675 4.140 3.677 4.142 multiply_cannon_metrocomm4 333 9.7 0.001 0.001 1.611 4.140 mp_irecv_dv 1241 11.2 1.590 4.114 1.590 4.114 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.045 3.996 calculate_norms 792 9.8 3.542 3.661 3.542 3.661 make_images_sizes 222 9.7 0.000 0.000 1.044 3.618 mp_alltoall_i44 222 10.7 1.044 3.618 1.044 3.618 ls_scf_post 1 4.0 0.000 0.000 3.600 3.603 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.396 3.491 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.428 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.326 dbcsr_merge_all 275 8.9 0.474 0.531 2.057 2.160 dbcsr_data_release 10123 10.8 1.330 2.004 1.330 2.004 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.158000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3569.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/92831d4de10a6dd11ff469e14bc401d403fbb9d7_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 8.718397E+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 0.078 0.094 108.072 108.073 qs_energies 1 2.0 0.000 0.000 106.695 106.707 ls_scf 1 3.0 0.000 0.000 103.829 103.840 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.039 78.143 ls_scf_main 1 4.0 0.000 0.000 66.143 66.144 density_matrix_trs4 2 5.0 0.002 0.003 57.237 57.300 multiply_cannon 111 7.7 0.131 0.185 50.427 52.258 multiply_cannon_loop 111 8.7 0.069 0.075 46.946 47.762 ls_scf_init_scf 1 4.0 0.000 0.000 33.980 33.981 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.662 32.673 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.840 29.849 mp_waitall_1 4569 11.1 22.828 26.566 22.828 26.566 make_m2s 222 7.7 0.005 0.005 24.012 24.986 make_images 222 8.7 4.602 5.008 23.901 24.836 multiply_cannon_multrec 444 9.7 17.884 18.632 22.463 23.152 hybrid_alltoall_any 227 10.6 1.667 3.644 12.992 15.843 make_images_data 222 9.7 0.003 0.003 13.207 15.589 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.033 11.833 multiply_cannon_sync_h2d 444 9.7 8.838 8.891 8.838 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.327 7.340 arnoldi_normal_ev 4 7.8 0.003 0.009 7.327 7.340 build_subspace 16 8.4 0.026 0.036 6.783 6.794 dbcsr_matrix_vector_mult 304 9.0 0.009 0.026 5.454 5.589 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.303 dbcsr_matrix_vector_mult_local 304 10.0 5.013 5.296 5.015 5.299 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.837 4.925 dbcsr_mm_accdrv_process 1814 10.4 0.184 0.322 4.413 4.544 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.246 4.114 4.246 ls_scf_post 1 4.0 0.000 0.000 3.706 3.717 make_images_sizes 222 9.7 0.000 0.000 1.448 3.579 mp_alltoall_i44 222 10.7 1.448 3.579 1.448 3.579 ls_scf_store_result 1 5.0 0.000 0.000 3.451 3.460 mp_allgather_i34 111 8.7 1.107 3.405 1.107 3.405 calculate_norms 792 9.8 3.250 3.295 3.250 3.295 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.154 dbcsr_merge_all 275 8.9 0.891 0.921 2.871 2.934 dbcsr_complete_redistribute 5 7.6 1.448 1.492 2.777 2.925 dbcsr_data_release 12724 10.6 2.331 2.863 2.331 2.863 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.835 2.835 matrix_ls_to_qs 2 6.0 0.000 0.000 2.415 2.592 dbcsr_sort_data 325 11.1 2.451 2.526 2.451 2.526 dbcsr_new_transposed 4 7.5 0.243 0.251 2.304 2.319 dbcsr_frobenius_norm 74 6.6 2.061 2.135 2.192 2.224 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.207 dbcsr_add_anytype 103 7.2 0.860 0.891 2.133 2.207 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.195 2.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.073000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6930.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 92831d4de10a6dd11ff469e14bc401d403fbb9d7 Summary: empty Status: OK