=== This is the CP2K Performance-Test === Updating b6737578d..b57866618 Fast-forward src/emd/rt_propagation_output.F | 1 - src/emd/rt_propagation_utils.F | 2 + src/input_cp2k_mp2.F | 12 ++ src/mp2_ri_gpw.F | 186 +++++++++++---------- src/mp2_setup.F | 1 + src/mp2_types.F | 2 +- .../H2O_grad_gpw_single_group.inp | 99 +++++++++++ tests/QS/regtest-mp2-grad-1/TEST_FILES | 1 + .../CH_dyn_screen_single_group.inp | 109 ++++++++++++ tests/QS/regtest-mp2-grad-2/TEST_FILES | 1 + .../QS/regtest-ri-mp2/RI_MP2_CH3_single_group.inp | 83 +++++++++ tests/QS/regtest-ri-mp2/TEST_FILES | 1 + 12 files changed, 408 insertions(+), 90 deletions(-) create mode 100644 tests/QS/regtest-mp2-grad-1/H2O_grad_gpw_single_group.inp create mode 100644 tests/QS/regtest-mp2-grad-2/CH_dyn_screen_single_group.inp create mode 100644 tests/QS/regtest-ri-mp2/RI_MP2_CH3_single_group.inp Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: b5786661874f8455978168cb43dd147fdac8cf4f ################# 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 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, 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 (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/01 job id: 44268501 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/02 job id: 44268502 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/03 job id: 44268503 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/04 job id: 44268504 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/05 job id: 44268505 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/06 job id: 44268506 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/07 job id: 44268508 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/08 job id: 44268509 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/09 job id: 44268510 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/10 job id: 44268511 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/11 job id: 44268512 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/12 job id: 44268513 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/13 job id: 44268514 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/14 job id: 44268515 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/15 job id: 44268516 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/16 job id: 44268517 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/17 job id: 44268518 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/18 job id: 44268519 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/19 job id: 44268520 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/20 job id: 44268521 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/21 job id: 44268522 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/22 job id: 44268523 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/23 job id: 44268524 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/24 job id: 44268525 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/25 job id: 44268526 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/26 job id: 44268528 --- 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/b5786661874f8455978168cb43dd147fdac8cf4f_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 0.030 0.035 133.552 133.553 farming_run 1 2.0 132.978 132.982 133.515 133.517 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.462026E+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 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 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 0.011 0.034 115.228 115.229 qs_energies 1 2.0 0.000 0.000 114.936 114.938 mp2_main 1 3.0 0.000 0.000 112.966 112.969 mp2_gpw_main 1 4.0 0.020 0.026 112.104 112.107 mp2_ri_gpw_compute_in 1 5.0 0.172 0.175 93.100 93.499 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.479 55.879 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.168 41.787 47.126 get_2c_integrals 1 6.0 0.001 0.009 36.913 37.445 integrate_v_rspace 273 8.0 0.437 0.449 25.165 30.250 pw_transfer 6555 10.6 0.378 0.389 27.371 27.797 grid_integrate_task_list 273 9.0 20.967 26.535 20.967 26.535 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.049 26.481 fft_wrap_pw1pw2_100 2178 12.4 1.165 1.238 23.569 24.016 compute_2c_integrals 1 7.0 0.003 0.003 19.228 19.228 rpa_ri_compute_en 1 5.0 0.001 0.016 18.899 19.141 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.793 18.953 mp2_eri_2c_integrate_gpw 1 9.0 2.390 2.426 18.790 18.949 cp_fm_cholesky_decompose 12 8.2 17.644 18.206 17.644 18.206 cholesky_decomp 1 7.0 0.000 0.000 16.539 17.084 fft3d_s 5443 13.4 16.133 16.580 16.155 16.602 ao_to_mo_and_store_B_mult_1 272 7.0 10.841 15.548 10.841 15.548 calculate_wavefunction 272 8.0 5.402 5.563 12.542 13.168 rpa_num_int 1 6.0 0.000 0.001 10.589 10.589 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.500 10.517 calc_mat_Q 8 8.0 0.000 0.000 9.347 9.436 contract_S_to_Q 8 9.0 0.000 0.000 8.771 8.855 calc_potential_gpw 544 9.5 0.005 0.006 8.226 8.597 parallel_gemm_fm 14 9.1 0.000 0.000 8.348 8.467 parallel_gemm_fm_cosma 14 10.1 8.348 8.467 8.348 8.467 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.166 8.437 potential_pw2rs 545 10.0 0.107 0.109 7.679 8.347 create_integ_mat 1 6.0 0.022 0.027 7.804 7.804 collocate_single_gaussian 272 10.0 0.040 0.042 7.418 7.627 array2fm 1 7.0 0.000 0.000 6.605 7.245 pw_scatter_s 2720 13.7 4.431 4.678 4.431 4.678 pw_gather_s 2722 13.2 3.907 4.319 3.907 4.319 array2fm_buffer_send 1 8.0 2.973 3.168 2.973 3.168 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.106890, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2733.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 0.037 396.917 396.919 farming_run 1 2.0 395.713 395.717 396.873 396.876 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.220833E+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 27402025. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.047 209.813 209.813 qs_energies 1 2.0 0.000 0.000 209.562 209.568 scf_env_do_scf 1 3.0 0.000 0.000 106.720 106.720 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.844 105.854 rebuild_ks_matrix 4 6.0 0.000 0.000 105.843 105.852 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.065 105.843 105.852 hfx_ks_matrix 4 8.0 0.001 0.001 105.455 105.458 integrate_four_center 4 9.0 0.144 0.449 105.454 105.457 mp2_main 1 3.0 0.000 0.000 102.548 102.554 mp2_gpw_main 1 4.0 0.033 0.044 101.698 101.708 integrate_four_center_main 4 10.0 0.116 0.527 96.903 99.790 integrate_four_center_bin 267 11.0 96.787 99.769 96.787 99.769 init_scf_loop 1 4.0 0.000 0.000 92.517 92.517 mp2_ri_gpw_compute_in 1 5.0 0.067 0.077 74.871 75.970 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.536 55.635 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.156 42.084 47.405 integrate_v_rspace 95 8.0 0.397 0.567 28.461 33.604 pw_transfer 2240 10.6 0.145 0.165 29.899 30.260 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.926 29.290 grid_integrate_task_list 95 9.0 23.750 29.098 23.750 29.098 mp2_ri_gpw_compute_en 1 5.0 0.056 0.067 26.679 28.461 fft_wrap_pw1pw2_100 730 12.4 1.250 1.392 26.627 26.994 ao_to_mo_and_store_B_mult_1 91 7.0 10.764 26.191 10.764 26.191 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.840 1.893 24.935 24.947 get_2c_integrals 1 6.0 0.000 0.000 20.242 20.271 compute_2c_integrals 1 7.0 0.007 0.021 19.220 19.225 compute_2c_integrals_loop_lm 1 8.0 0.003 0.008 18.793 19.061 mp2_eri_2c_integrate_gpw 1 9.0 1.733 1.827 18.790 19.059 fft3d_s 1823 13.4 18.417 18.724 18.431 18.738 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.201 14.201 calculate_wavefunction 91 8.0 2.015 2.048 9.743 10.005 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.588 8.800 9.611 potential_pw2rs 186 10.0 0.033 0.035 8.618 9.054 local_gemm 172 8.0 8.244 9.033 8.244 9.033 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.517 7.914 8.571 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.205 8.489 calc_potential_gpw 182 9.5 0.002 0.002 7.896 8.106 collocate_single_gaussian 91 10.0 0.017 0.021 7.847 8.050 mp_sendrecv_dm3 2068 8.0 5.947 6.603 5.947 6.603 mp2_ri_gpw_compute_en_ener 172 7.0 6.336 6.469 6.336 6.469 mp_sync 38 10.4 3.218 6.437 3.218 6.437 pw_gather_s 912 13.2 4.928 5.302 4.928 5.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.697165, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 451.850240E+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 579164. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 52.512 52.512 qs_mol_dyn_low 1 2.0 0.003 0.003 52.262 52.270 qs_forces 11 3.9 0.002 0.002 52.199 52.200 qs_energies 11 4.9 0.001 0.002 50.715 50.724 scf_env_do_scf 11 5.9 0.001 0.001 44.675 44.675 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.635 42.635 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.808 34.285 qs_scf_new_mos 108 7.5 0.000 0.001 32.381 32.679 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.380 32.678 ot_scf_mini 108 9.5 0.002 0.002 30.728 30.913 multiply_cannon 2286 13.5 0.186 0.195 26.187 27.805 multiply_cannon_loop 2286 14.5 1.504 1.573 25.485 27.093 velocity_verlet 10 3.0 0.002 0.005 25.773 25.774 ot_mini 108 10.5 0.001 0.001 19.468 19.717 qs_ot_get_derivative 108 11.5 0.001 0.001 16.518 16.729 mp_waitall_1 245248 16.5 8.527 15.174 8.527 15.174 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 6.048 13.177 multiply_cannon_multrec 54864 15.5 4.200 6.569 7.737 11.154 rebuild_ks_matrix 119 8.3 0.000 0.000 7.964 8.099 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 7.964 8.099 multiply_cannon_sync_h2d 54864 15.5 5.883 7.583 5.883 7.583 mp_sum_l 7207 12.9 5.621 7.259 5.621 7.259 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.012 7.132 qs_ot_get_p 119 10.4 0.001 0.001 6.504 6.825 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.591 6.029 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.478 5.586 init_scf_run 11 5.9 0.000 0.001 4.825 4.825 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.825 4.825 dbcsr_mm_accdrv_process 76910 16.1 1.191 1.831 3.458 4.797 sum_up_and_integrate 119 10.3 0.012 0.014 4.555 4.562 integrate_v_rspace 119 11.3 0.002 0.002 4.543 4.551 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.271 4.395 calculate_rho_elec 119 8.7 0.011 0.017 4.271 4.395 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.776 3.838 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.797 3.217 calculate_dm_sparse 119 9.5 0.000 0.001 2.971 3.112 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.883 3.063 apply_single 119 13.6 0.000 0.000 2.883 3.063 jit_kernel_multiply 13 15.8 2.205 2.900 2.205 2.900 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.869 2.870 rs_pw_transfer 974 11.9 0.011 0.012 2.761 2.867 calculate_first_density_matrix 1 7.0 0.000 0.001 2.692 2.698 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.656 2.657 ot_diis_step 108 11.5 0.006 0.006 2.642 2.642 cp_fm_redistribute_end 50 14.0 2.416 2.633 2.421 2.634 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.555 2.632 cp_fm_diag_elpa_base 50 14.0 0.211 2.554 0.212 2.562 density_rs2pw 119 9.7 0.004 0.005 2.313 2.436 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.378 2.380 acc_transpose_blocks 54864 15.5 0.238 0.257 1.816 2.255 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.109 2.156 grid_integrate_task_list 119 12.3 2.028 2.120 2.028 2.120 wfi_extrapolate 11 7.9 0.001 0.001 2.074 2.074 init_scf_loop 11 6.9 0.000 0.000 2.018 2.018 pw_transfer 1439 11.6 0.053 0.060 1.875 1.948 potential_pw2rs 119 12.3 0.004 0.004 1.898 1.911 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.798 1.874 mp_sum_d 4125 12.0 1.287 1.856 1.287 1.856 make_m2s 4572 13.5 0.053 0.057 1.686 1.738 make_images 4572 14.5 0.133 0.138 1.603 1.656 fft3d_ps 1201 14.6 0.372 0.478 1.569 1.640 mp_alltoall_d11v 2130 13.8 1.272 1.504 1.272 1.504 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.479 1.496 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.378 1.457 mp_waitany 12084 13.8 1.245 1.419 1.245 1.419 grid_collocate_task_list 119 9.7 1.287 1.342 1.287 1.342 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.726 1.116 acc_transpose_blocks_kernels 54864 16.5 0.253 0.391 0.808 1.098 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.512000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 489.762816E+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 1030362. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 39.098 39.099 qs_mol_dyn_low 1 2.0 0.003 0.004 38.831 38.839 qs_forces 11 3.9 0.001 0.002 38.771 38.772 qs_energies 11 4.9 0.003 0.013 37.067 37.079 scf_env_do_scf 11 5.9 0.000 0.001 31.726 31.727 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.231 29.231 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.483 21.852 qs_scf_new_mos 108 7.5 0.001 0.001 19.945 20.192 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.945 20.192 ot_scf_mini 108 9.5 0.002 0.003 19.061 19.239 velocity_verlet 10 3.0 0.001 0.002 18.345 18.346 multiply_cannon 2286 13.5 0.208 0.219 16.502 18.022 multiply_cannon_loop 2286 14.5 0.900 0.973 15.399 16.823 ot_mini 108 10.5 0.001 0.001 11.785 12.025 mp_waitall_1 200699 16.5 5.682 11.045 5.682 11.045 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.181 9.824 qs_ot_get_derivative 108 11.5 0.001 0.001 9.327 9.508 multiply_cannon_multrec 27432 15.5 1.967 4.484 6.016 8.803 rebuild_ks_matrix 119 8.3 0.000 0.001 7.260 7.401 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.259 7.401 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.421 6.551 dbcsr_mm_accdrv_process 47894 16.0 3.226 5.518 3.978 5.844 qs_ot_get_p 119 10.4 0.001 0.001 4.386 4.615 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.627 4.481 sum_up_and_integrate 119 10.3 0.024 0.026 4.235 4.240 integrate_v_rspace 119 11.3 0.002 0.002 4.211 4.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.050 4.118 apply_single 119 13.6 0.000 0.000 3.050 4.118 init_scf_run 11 5.9 0.000 0.001 4.063 4.063 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.062 4.063 mp_sum_l 7207 12.9 2.044 4.006 2.044 4.006 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.844 3.881 calculate_rho_elec 119 8.7 0.021 0.024 3.844 3.880 rs_pw_transfer 974 11.9 0.010 0.011 2.621 3.033 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.941 2.960 multiply_cannon_sync_h2d 27432 15.5 2.189 2.880 2.189 2.880 make_m2s 4572 13.5 0.052 0.054 2.495 2.729 make_images 4572 14.5 0.202 0.238 2.406 2.640 density_rs2pw 119 9.7 0.004 0.004 2.174 2.603 calculate_first_density_matrix 1 7.0 0.000 0.002 2.534 2.535 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.505 2.506 init_scf_loop 11 6.9 0.002 0.012 2.473 2.473 ot_diis_step 108 11.5 0.011 0.011 2.409 2.410 calculate_dm_sparse 119 9.5 0.000 0.001 2.160 2.234 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.135 2.228 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.160 2.161 cp_fm_redistribute_end 50 14.0 1.787 2.134 1.791 2.135 cp_fm_diag_elpa_base 50 14.0 0.330 2.021 0.343 2.076 pw_transfer 1439 11.6 0.066 0.072 2.028 2.065 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.935 1.976 potential_pw2rs 119 12.3 0.006 0.006 1.954 1.965 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.952 1.954 grid_integrate_task_list 119 12.3 1.842 1.947 1.842 1.947 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.752 1.792 fft3d_ps 1201 14.6 0.513 0.567 1.637 1.670 jit_kernel_multiply 8 16.1 0.700 1.643 0.700 1.643 prepare_preconditioner 11 7.9 0.000 0.001 1.559 1.587 make_preconditioner 11 8.9 0.000 0.000 1.559 1.587 acc_transpose_blocks 27432 15.5 0.112 0.117 1.242 1.557 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.505 1.545 make_images_data 4572 15.5 0.045 0.052 1.141 1.543 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.458 1.513 wfi_extrapolate 11 7.9 0.001 0.001 1.477 1.477 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.999 1.459 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.385 1.391 grid_collocate_task_list 119 9.7 1.228 1.368 1.228 1.368 mp_alltoall_d11v 2130 13.8 1.192 1.311 1.192 1.311 mp_allgather_i34 2286 14.5 0.545 1.309 0.545 1.309 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.171 1.218 mp_sum_d 4125 12.0 0.610 1.063 0.610 1.063 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.556 0.977 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.966 0.967 mp_waitany 5720 13.7 0.530 0.966 0.530 0.966 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.701 0.929 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.918 mp_alltoall_z22v 1201 16.6 0.718 0.817 0.718 0.817 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 0.731 0.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.099000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.454545, yerr=2.061052 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 522.539008E+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 2325109. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.021 0.041 32.102 32.107 qs_mol_dyn_low 1 2.0 0.003 0.003 31.888 31.896 qs_forces 11 3.9 0.002 0.002 31.828 31.828 qs_energies 11 4.9 0.001 0.002 30.246 30.248 scf_env_do_scf 11 5.9 0.000 0.001 25.544 25.545 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.026 23.026 dbcsr_multiply_generic 2286 12.5 0.093 0.096 16.223 16.315 velocity_verlet 10 3.0 0.001 0.002 15.253 15.255 qs_scf_new_mos 108 7.5 0.001 0.001 14.726 14.742 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.726 14.741 ot_scf_mini 108 9.5 0.002 0.002 14.008 14.019 multiply_cannon 2286 13.5 0.197 0.203 12.990 13.816 multiply_cannon_loop 2286 14.5 0.637 0.672 12.195 13.044 ot_mini 108 10.5 0.001 0.001 8.659 8.673 qs_ot_get_derivative 108 11.5 0.001 0.001 7.158 7.169 multiply_cannon_multrec 18288 15.5 1.963 2.833 6.788 6.970 rebuild_ks_matrix 119 8.3 0.000 0.001 6.422 6.439 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.422 6.439 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.670 5.686 dbcsr_mm_accdrv_process 38222 16.0 4.119 5.339 4.740 5.528 sum_up_and_integrate 119 10.3 0.029 0.030 3.999 4.004 integrate_v_rspace 119 11.3 0.002 0.003 3.969 3.978 mp_waitall_1 158411 16.6 2.515 3.673 2.515 3.673 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.516 3.520 calculate_rho_elec 119 8.7 0.030 0.031 3.515 3.520 init_scf_run 11 5.9 0.000 0.001 3.436 3.436 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.435 3.436 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.677 3.282 qs_ot_get_p 119 10.4 0.001 0.001 3.177 3.195 rs_pw_transfer 974 11.9 0.009 0.010 2.312 2.592 init_scf_loop 11 6.9 0.000 0.000 2.500 2.501 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.384 2.447 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.030 2.352 apply_single 119 13.6 0.000 0.000 2.030 2.352 density_rs2pw 119 9.7 0.004 0.004 2.035 2.294 calculate_first_density_matrix 1 7.0 0.000 0.000 2.223 2.225 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.125 2.131 pw_transfer 1439 11.6 0.066 0.072 1.977 1.987 make_m2s 4572 13.5 0.044 0.046 1.777 1.923 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.883 1.895 grid_integrate_task_list 119 12.3 1.805 1.895 1.805 1.895 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.861 1.861 make_images 4572 14.5 0.192 0.204 1.692 1.836 calculate_dm_sparse 119 9.5 0.000 0.000 1.802 1.810 potential_pw2rs 119 12.3 0.007 0.008 1.766 1.772 prepare_preconditioner 11 7.9 0.000 0.000 1.717 1.719 make_preconditioner 11 8.9 0.000 0.000 1.717 1.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.670 1.672 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.569 1.654 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.635 1.636 cp_fm_redistribute_end 50 14.0 1.221 1.616 1.222 1.617 cp_fm_diag_elpa_base 50 14.0 0.379 1.535 0.393 1.581 fft3d_ps 1201 14.6 0.524 0.544 1.563 1.577 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.565 1.571 multiply_cannon_sync_h2d 18288 15.5 1.407 1.559 1.407 1.559 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.510 1.517 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.495 1.507 ot_diis_step 108 11.5 0.011 0.011 1.482 1.482 mp_sum_l 7207 12.9 1.071 1.434 1.071 1.434 grid_collocate_task_list 119 9.7 1.214 1.363 1.214 1.363 jit_kernel_multiply 7 16.1 0.569 1.327 0.569 1.327 acc_transpose_blocks 18288 15.5 0.078 0.080 1.233 1.263 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.176 1.182 wfi_extrapolate 11 7.9 0.001 0.001 1.166 1.166 qs_energies_init_hamiltonians 11 5.9 0.001 0.020 0.966 0.985 make_images_data 4572 15.5 0.045 0.049 0.781 0.946 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.877 0.897 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.669 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.823 0.825 mp_waitany 9880 13.7 0.513 0.808 0.513 0.808 acc_transpose_blocks_kernels 18288 16.5 0.210 0.219 0.783 0.797 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.121 0.501 0.780 mp_alltoall_z22v 1201 16.6 0.638 0.745 0.638 0.745 mp_alltoall_d11v 2130 13.8 0.647 0.742 0.647 0.742 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.647 0.730 cp_fm_cholesky_invert 11 10.9 0.724 0.728 0.724 0.728 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.107000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=1.966664 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 557.723648E+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_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.010 0.027 35.374 35.374 qs_mol_dyn_low 1 2.0 0.003 0.003 35.150 35.158 qs_forces 11 3.9 0.002 0.002 35.089 35.090 qs_energies 11 4.9 0.001 0.001 33.366 33.369 scf_env_do_scf 11 5.9 0.000 0.001 28.223 28.224 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.825 24.825 dbcsr_multiply_generic 2286 12.5 0.099 0.100 18.486 18.676 velocity_verlet 10 3.0 0.002 0.002 17.730 17.732 qs_scf_new_mos 108 7.5 0.001 0.001 16.319 16.363 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.319 16.362 multiply_cannon 2286 13.5 0.229 0.262 15.022 15.921 ot_scf_mini 108 9.5 0.002 0.003 15.382 15.432 multiply_cannon_loop 2286 14.5 0.940 0.970 14.084 14.866 ot_mini 108 10.5 0.001 0.001 9.545 9.608 multiply_cannon_multrec 27432 15.5 2.394 3.238 8.889 9.312 qs_ot_get_derivative 108 11.5 0.001 0.001 7.737 7.787 dbcsr_mm_accdrv_process 47916 15.9 5.518 7.108 6.400 7.668 rebuild_ks_matrix 119 8.3 0.000 0.000 6.632 6.691 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.632 6.690 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.892 5.942 sum_up_and_integrate 119 10.3 0.035 0.038 3.825 3.832 init_scf_run 11 5.9 0.000 0.001 3.812 3.813 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.812 3.812 integrate_v_rspace 119 11.3 0.002 0.002 3.789 3.797 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.568 3.597 calculate_rho_elec 119 8.7 0.040 0.046 3.568 3.596 init_scf_loop 11 6.9 0.000 0.000 3.377 3.377 qs_ot_get_p 119 10.4 0.001 0.001 3.240 3.318 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.893 3.211 mp_waitall_1 137007 16.6 2.007 2.591 2.007 2.591 prepare_preconditioner 11 7.9 0.000 0.000 2.503 2.512 make_preconditioner 11 8.9 0.000 0.000 2.503 2.512 calculate_first_density_matrix 1 7.0 0.000 0.000 2.455 2.457 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.115 2.442 make_m2s 4572 13.5 0.054 0.055 2.147 2.255 calculate_dm_sparse 119 9.5 0.000 0.000 2.164 2.213 rs_pw_transfer 974 11.9 0.009 0.009 2.003 2.190 density_rs2pw 119 9.7 0.004 0.004 1.995 2.184 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.968 2.147 make_images 4572 14.5 0.273 0.334 2.039 2.147 apply_single 119 13.6 0.000 0.000 1.967 2.147 pw_transfer 1439 11.6 0.066 0.072 2.095 2.133 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.001 2.045 jit_kernel_multiply 10 15.8 0.822 2.015 0.822 2.015 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.992 2.001 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.914 1.940 grid_integrate_task_list 119 12.3 1.814 1.925 1.814 1.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.890 1.891 ot_diis_step 108 11.5 0.012 0.012 1.770 1.770 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.717 1.733 fft3d_ps 1201 14.6 0.558 0.609 1.676 1.714 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.647 1.693 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.680 1.681 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 1.035 1.650 potential_pw2rs 119 12.3 0.008 0.009 1.591 1.597 acc_transpose_blocks 27432 15.5 0.116 0.119 1.543 1.581 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.438 1.440 cp_fm_redistribute_end 50 14.0 0.953 1.416 0.954 1.416 cp_fm_diag_elpa_base 50 14.0 0.440 1.348 0.460 1.389 grid_collocate_task_list 119 9.7 1.221 1.356 1.221 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.346 1.353 wfi_extrapolate 11 7.9 0.001 0.001 1.306 1.306 mp_sum_l 7207 12.9 0.894 1.261 0.894 1.261 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.095 1.116 cp_fm_upper_to_full 72 13.5 0.796 1.104 0.796 1.104 multiply_cannon_sync_h2d 27432 15.5 0.991 1.093 0.991 1.093 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.080 1.081 dbcsr_complete_redistribute 329 12.2 0.123 0.153 0.766 1.036 make_images_data 4572 15.5 0.045 0.049 0.816 0.936 acc_transpose_blocks_kernels 27432 16.5 0.268 0.276 0.906 0.932 mp_alltoall_d11v 2130 13.8 0.701 0.926 0.701 0.926 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.874 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.702 0.862 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.831 0.837 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.567 0.832 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.176 0.815 cp_fm_cholesky_invert 11 10.9 0.741 0.744 0.741 0.744 mp_alltoall_z22v 1201 16.6 0.700 0.737 0.700 0.737 mp_alltoall_i22 627 13.8 0.416 0.712 0.416 0.712 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.374000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.090909, yerr=6.082083 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 598.736896E+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_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.021 0.047 28.575 28.577 qs_mol_dyn_low 1 2.0 0.003 0.003 28.358 28.367 qs_forces 11 3.9 0.002 0.002 28.298 28.298 qs_energies 11 4.9 0.001 0.002 26.578 26.581 scf_env_do_scf 11 5.9 0.000 0.001 21.877 21.877 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.365 19.366 velocity_verlet 10 3.0 0.002 0.002 14.513 14.516 dbcsr_multiply_generic 2286 12.5 0.093 0.095 12.181 12.312 qs_scf_new_mos 108 7.5 0.001 0.001 11.093 11.119 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.092 11.118 ot_scf_mini 108 9.5 0.002 0.002 10.418 10.444 multiply_cannon 2286 13.5 0.230 0.241 9.646 10.159 multiply_cannon_loop 2286 14.5 0.330 0.341 8.729 8.949 rebuild_ks_matrix 119 8.3 0.000 0.000 6.064 6.082 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.064 6.081 multiply_cannon_multrec 9144 15.5 1.594 1.830 5.736 5.955 ot_mini 108 10.5 0.001 0.001 5.803 5.834 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.405 5.422 qs_ot_get_derivative 108 11.5 0.001 0.001 4.502 4.530 dbcsr_mm_accdrv_process 12550 15.8 3.121 4.059 4.041 4.099 sum_up_and_integrate 119 10.3 0.038 0.040 3.721 3.726 integrate_v_rspace 119 11.3 0.002 0.003 3.683 3.688 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.628 3.654 calculate_rho_elec 119 8.7 0.060 0.061 3.628 3.653 init_scf_run 11 5.9 0.000 0.001 3.271 3.271 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.271 3.271 qs_ot_get_p 119 10.4 0.001 0.001 2.828 2.869 init_scf_loop 11 6.9 0.000 0.000 2.491 2.492 pw_transfer 1439 11.6 0.067 0.070 2.194 2.205 mp_waitall_1 115863 16.7 1.608 2.166 1.608 2.166 calculate_first_density_matrix 1 7.0 0.000 0.000 2.123 2.126 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.100 2.113 density_rs2pw 119 9.7 0.004 0.004 1.940 2.080 grid_integrate_task_list 119 12.3 1.847 1.922 1.847 1.922 make_m2s 4572 13.5 0.034 0.035 1.759 1.906 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.904 1.906 rs_pw_transfer 974 11.9 0.008 0.008 1.706 1.831 make_images 4572 14.5 0.269 0.306 1.670 1.816 fft3d_ps 1201 14.6 0.562 0.574 1.759 1.770 prepare_preconditioner 11 7.9 0.000 0.000 1.740 1.744 make_preconditioner 11 8.9 0.000 0.000 1.740 1.744 calculate_dm_sparse 119 9.5 0.000 0.000 1.717 1.739 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.727 1.738 jit_kernel_multiply 10 15.7 0.881 1.735 0.881 1.735 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.707 1.707 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.629 1.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.571 1.573 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.501 1.517 potential_pw2rs 119 12.3 0.010 0.010 1.446 1.449 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.427 1.428 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.412 1.419 cp_fm_redistribute_end 50 14.0 0.713 1.404 0.713 1.405 grid_collocate_task_list 119 9.7 1.272 1.387 1.272 1.387 cp_fm_diag_elpa_base 50 14.0 0.646 1.330 0.690 1.387 ot_diis_step 108 11.5 0.012 0.013 1.289 1.289 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.217 1.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.167 1.185 apply_single 119 13.6 0.000 0.000 1.167 1.185 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.136 1.148 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.138 1.145 wfi_extrapolate 11 7.9 0.001 0.001 1.095 1.095 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.785 1.026 make_images_data 4572 15.5 0.039 0.042 0.827 1.013 acc_transpose_blocks 9144 15.5 0.040 0.040 0.904 0.948 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.918 mp_alltoall_d11v 2130 13.8 0.790 0.912 0.790 0.912 cp_fm_cholesky_invert 11 10.9 0.883 0.886 0.883 0.886 multiply_cannon_sync_h2d 9144 15.5 0.713 0.789 0.713 0.789 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.784 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.456 0.766 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.724 0.731 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.672 0.725 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.658 0.702 mp_alltoall_z22v 1201 16.6 0.644 0.699 0.644 0.699 mp_allgather_i34 2286 14.5 0.253 0.681 0.253 0.681 yz_to_x 606 15.1 0.265 0.276 0.602 0.619 x_to_yz 595 16.2 0.275 0.287 0.583 0.592 jit_kernel_transpose 5 15.6 0.541 0.585 0.541 0.585 mp_waitany 5200 13.7 0.450 0.574 0.450 0.574 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.577000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.545455, yerr=4.097974 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 765.431808E+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_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.080 42.047 42.055 qs_mol_dyn_low 1 2.0 0.003 0.003 41.618 41.736 qs_forces 11 3.9 0.002 0.002 41.486 41.487 qs_energies 11 4.9 0.001 0.001 39.499 39.502 scf_env_do_scf 11 5.9 0.001 0.001 33.637 33.637 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.787 25.789 velocity_verlet 10 3.0 0.057 0.064 23.541 23.547 dbcsr_multiply_generic 2286 12.5 0.100 0.102 17.644 17.833 qs_scf_new_mos 108 7.5 0.001 0.001 15.901 15.997 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.900 15.996 ot_scf_mini 108 9.5 0.002 0.002 14.812 14.915 multiply_cannon 2286 13.5 0.302 0.309 13.779 14.718 multiply_cannon_loop 2286 14.5 0.343 0.349 12.512 13.489 ot_mini 108 10.5 0.001 0.001 8.854 8.971 multiply_cannon_multrec 9144 15.5 3.431 4.839 8.667 8.765 init_scf_loop 11 6.9 0.000 0.000 7.823 7.829 rebuild_ks_matrix 119 8.3 0.000 0.000 7.390 7.534 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.390 7.534 qs_ot_get_derivative 108 11.5 0.001 0.001 6.806 6.909 prepare_preconditioner 11 7.9 0.000 0.000 6.836 6.851 make_preconditioner 11 8.9 0.000 0.000 6.836 6.851 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.682 6.813 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.405 6.732 dbcsr_mm_accdrv_process 12550 15.8 3.947 5.671 5.113 6.472 cp_fm_upper_to_full 72 14.2 3.179 4.551 3.179 4.551 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.286 4.292 calculate_rho_elec 119 8.7 0.118 0.121 4.285 4.291 sum_up_and_integrate 119 10.3 0.065 0.067 4.059 4.065 integrate_v_rspace 119 11.3 0.003 0.003 3.993 4.000 init_scf_run 11 5.9 0.000 0.001 3.827 3.827 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.827 3.827 qs_ot_get_p 119 10.4 0.001 0.001 3.229 3.365 mp_waitall_1 94719 16.7 2.332 3.349 2.332 3.349 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.497 2.941 pw_transfer 1439 11.6 0.069 0.070 2.835 2.841 dbcsr_complete_redistribute 329 12.2 0.285 0.289 1.961 2.819 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.736 2.742 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.663 2.503 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.229 2.489 apply_single 119 13.6 0.000 0.000 2.229 2.489 make_m2s 4572 13.5 0.038 0.038 2.277 2.460 calculate_first_density_matrix 1 7.0 0.000 0.000 2.364 2.366 fft3d_ps 1201 14.6 0.595 0.604 2.360 2.365 make_images 4572 14.5 0.353 0.383 2.158 2.340 mp_alltoall_i22 627 13.8 1.458 2.318 1.458 2.318 fft_wrap_pw1pw2_140 487 13.2 0.095 0.098 2.306 2.312 calculate_dm_sparse 119 9.5 0.000 0.000 2.250 2.271 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.426 2.260 density_rs2pw 119 9.7 0.004 0.004 2.202 2.219 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.296 2.209 grid_integrate_task_list 119 12.3 2.061 2.083 2.061 2.083 ot_diis_step 108 11.5 0.014 0.014 2.022 2.022 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.996 1.998 qs_ot_p2m_diag 50 11.0 0.044 0.044 1.979 1.980 jit_kernel_multiply 10 15.4 1.139 1.866 1.139 1.866 mp_sum_l 7207 12.9 1.078 1.786 1.078 1.786 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.771 1.772 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.677 1.727 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.693 1.711 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.706 1.707 grid_collocate_task_list 119 9.7 1.492 1.517 1.492 1.517 cp_fm_cholesky_invert 11 10.9 1.461 1.465 1.461 1.465 rs_pw_transfer 974 11.9 0.009 0.009 1.439 1.456 potential_pw2rs 119 12.3 0.014 0.015 1.449 1.452 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.418 1.419 cp_fm_diag_elpa_base 50 14.0 1.276 1.327 1.417 1.417 wfi_extrapolate 11 7.9 0.001 0.001 1.395 1.395 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.105 1.368 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.308 1.316 make_images_data 4572 15.5 0.043 0.046 1.068 1.309 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.128 1.147 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.106 1.118 mp_alltoall_d11v 2130 13.8 1.077 1.102 1.077 1.102 multiply_cannon_sync_h2d 9144 15.5 1.039 1.042 1.039 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.991 1.028 acc_transpose_blocks 9144 15.5 0.039 0.039 1.010 1.023 qs_create_task_list 11 7.9 0.000 0.000 0.938 0.954 yz_to_x 606 15.1 0.460 0.470 0.946 0.954 generate_qs_task_list 11 8.9 0.370 0.389 0.937 0.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.919 0.933 mp_alltoall_z22v 1201 16.6 0.813 0.845 0.813 0.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.055000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.454545, yerr=11.919149 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 500.027392E+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 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.039 0.089 84.042 84.043 qs_mol_dyn_low 1 2.0 0.003 0.004 83.705 83.715 qs_forces 11 3.9 0.002 0.003 83.443 83.445 qs_energies 11 4.9 0.007 0.049 80.506 80.523 scf_env_do_scf 11 5.9 0.001 0.001 71.540 71.543 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.916 65.917 dbcsr_multiply_generic 2055 12.4 0.107 0.112 52.179 52.411 qs_scf_new_mos 99 7.5 0.001 0.001 48.404 48.531 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.404 48.530 ot_scf_mini 99 9.5 0.002 0.003 46.012 46.112 multiply_cannon 2055 13.4 0.179 0.186 42.619 43.478 multiply_cannon_loop 2055 14.4 1.551 1.600 41.675 42.577 velocity_verlet 10 3.0 0.001 0.001 42.468 42.471 ot_mini 99 10.5 0.001 0.001 27.770 27.871 qs_ot_get_derivative 99 11.5 0.001 0.001 20.905 20.997 multiply_cannon_multrec 49320 15.4 12.146 12.959 17.160 17.969 rebuild_ks_matrix 110 8.3 0.000 0.001 14.709 14.881 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.017 14.709 14.880 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.882 13.035 mp_waitall_1 220248 16.4 11.512 12.618 11.512 12.618 multiply_cannon_sync_h2d 49320 15.4 9.973 10.546 9.973 10.546 qs_ot_get_p 110 10.4 0.001 0.001 9.765 9.911 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 7.012 8.429 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.172 7.692 apply_single 110 13.6 0.000 0.001 7.172 7.691 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.081 7.616 sum_up_and_integrate 110 10.3 0.036 0.043 7.162 7.181 integrate_v_rspace 110 11.3 0.003 0.004 7.125 7.153 init_scf_run 11 5.9 0.000 0.001 6.829 6.830 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.829 6.829 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.689 6.825 calculate_rho_elec 110 8.6 0.021 0.026 6.688 6.825 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.727 6.806 ot_diis_step 99 11.5 0.006 0.006 6.537 6.537 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.352 6.376 mp_sum_l 6514 12.8 5.338 6.145 5.338 6.145 init_scf_loop 11 6.9 0.001 0.005 5.594 5.596 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.307 5.307 dbcsr_mm_accdrv_process 87628 16.1 2.063 2.208 4.892 5.242 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.820 4.821 cp_fm_redistribute_end 48 14.0 4.207 4.794 4.210 4.795 cp_fm_diag_elpa_base 48 14.0 0.579 4.649 0.583 4.681 rs_pw_transfer 902 11.9 0.012 0.014 3.712 4.394 density_rs2pw 110 9.6 0.004 0.005 3.464 4.103 wfi_extrapolate 11 7.9 0.001 0.001 4.029 4.030 make_m2s 4110 13.4 0.061 0.064 3.803 3.898 calculate_dm_sparse 110 9.5 0.001 0.001 3.757 3.853 make_images 4110 14.4 0.178 0.191 3.707 3.805 multiply_cannon_metrocomm1 49320 15.4 0.061 0.066 2.692 3.726 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.614 3.618 grid_integrate_task_list 110 12.3 3.233 3.408 3.233 3.408 pw_transfer 1331 11.6 0.055 0.067 3.333 3.406 prepare_preconditioner 11 7.9 0.000 0.001 3.376 3.394 make_preconditioner 11 8.9 0.000 0.002 3.376 3.394 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.245 3.321 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.266 3.318 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.233 3.278 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.164 3.209 fft3d_ps 1111 14.6 0.784 0.891 2.774 2.837 fft_wrap_pw1pw2_140 451 13.1 0.171 0.192 2.735 2.820 potential_pw2rs 110 12.3 0.006 0.007 2.726 2.755 calculate_first_density_matrix 1 7.0 0.000 0.003 2.701 2.706 jit_kernel_multiply 13 15.9 2.545 2.609 2.545 2.609 mp_waitany 14300 13.8 1.774 2.464 1.774 2.464 mp_alltoall_d11v 2046 13.8 2.058 2.443 2.058 2.443 acc_transpose_blocks 49320 15.4 0.228 0.242 2.215 2.316 grid_collocate_task_list 110 9.6 2.088 2.295 2.088 2.295 mp_sum_d 3879 11.9 1.679 2.193 1.679 2.193 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.970 1.990 cp_fm_cholesky_invert 11 10.9 1.844 1.848 1.844 1.848 make_images_data 4110 15.4 0.043 0.046 1.701 1.805 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.479 1.681 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.043000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=1.477098 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 588.496896E+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 2462973. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.022 0.036 70.117 70.118 qs_mol_dyn_low 1 2.0 0.003 0.004 69.818 69.828 qs_forces 11 3.9 0.002 0.002 69.750 69.751 qs_energies 11 4.9 0.001 0.002 66.417 66.421 scf_env_do_scf 11 5.9 0.000 0.001 56.398 56.402 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.772 48.772 dbcsr_multiply_generic 2055 12.4 0.111 0.117 37.642 37.852 velocity_verlet 10 3.0 0.001 0.002 36.032 36.033 qs_scf_new_mos 99 7.5 0.001 0.001 32.354 32.487 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.353 32.486 multiply_cannon 2055 13.4 0.223 0.245 31.174 32.221 multiply_cannon_loop 2055 14.4 0.928 0.951 29.928 30.881 ot_scf_mini 99 9.5 0.003 0.003 30.709 30.823 ot_mini 99 10.5 0.001 0.001 18.037 18.159 multiply_cannon_multrec 24660 15.4 7.640 10.081 14.024 16.215 rebuild_ks_matrix 110 8.3 0.000 0.001 13.968 14.104 qs_ks_build_kohn_sham_matrix 110 9.3 0.026 0.069 13.968 14.104 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.316 12.441 qs_ot_get_derivative 99 11.5 0.001 0.001 12.267 12.388 mp_waitall_1 176588 16.5 7.593 9.852 7.593 9.852 multiply_cannon_sync_h2d 24660 15.4 7.034 8.499 7.034 8.499 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.159 7.745 init_scf_run 11 5.9 0.000 0.001 7.667 7.668 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.667 7.668 init_scf_loop 11 6.9 0.000 0.000 7.591 7.592 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.476 7.192 apply_single 110 13.6 0.000 0.001 6.476 7.192 sum_up_and_integrate 110 10.3 0.051 0.057 6.651 6.663 integrate_v_rspace 110 11.3 0.002 0.002 6.600 6.615 dbcsr_mm_accdrv_process 52282 16.1 4.803 5.799 6.223 6.511 qs_rho_update_rho_low 110 7.6 0.011 0.089 6.369 6.379 calculate_rho_elec 110 8.6 0.047 0.106 6.357 6.379 qs_ot_get_p 110 10.4 0.001 0.001 6.061 6.228 ot_diis_step 99 11.5 0.010 0.010 5.723 5.723 prepare_preconditioner 11 7.9 0.000 0.000 5.439 5.455 make_preconditioner 11 8.9 0.000 0.000 5.439 5.455 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.659 5.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.020 5.173 make_m2s 4110 13.4 0.057 0.060 4.115 4.561 make_images 4110 14.4 0.405 0.449 4.007 4.449 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.218 4.239 calculate_first_density_matrix 1 7.0 0.000 0.000 3.912 3.914 pw_transfer 1331 11.6 0.067 0.074 3.747 3.894 density_rs2pw 110 9.6 0.004 0.005 3.328 3.841 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.796 3.796 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.640 3.789 wfi_extrapolate 11 7.9 0.001 0.001 3.519 3.519 rs_pw_transfer 902 11.9 0.012 0.013 2.957 3.475 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.406 3.407 grid_integrate_task_list 110 12.3 3.165 3.339 3.165 3.339 fft_wrap_pw1pw2_140 451 13.1 0.203 0.221 3.122 3.276 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.274 3.276 cp_fm_redistribute_end 48 14.0 2.448 3.248 2.450 3.249 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.133 3.198 cp_fm_diag_elpa_base 48 14.0 0.764 3.098 0.795 3.193 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.077 3.134 fft3d_ps 1111 14.6 1.102 1.312 2.989 3.124 calculate_dm_sparse 110 9.5 0.001 0.001 3.015 3.042 hybrid_alltoall_any 4261 16.3 0.103 0.446 2.044 2.804 make_images_data 4110 15.4 0.047 0.051 2.330 2.778 cp_fm_cholesky_invert 11 10.9 2.516 2.523 2.516 2.523 grid_collocate_task_list 110 9.6 2.056 2.462 2.056 2.462 potential_pw2rs 110 12.3 0.008 0.009 2.442 2.455 jit_kernel_multiply 11 16.3 1.066 2.321 1.066 2.321 mp_sum_l 6514 12.8 1.735 2.306 1.735 2.306 mp_alltoall_d11v 2046 13.8 1.868 2.163 1.868 2.163 make_basis_sm 11 9.8 0.001 0.004 2.115 2.118 parallel_gemm_fm 81 9.0 0.000 0.000 2.077 2.083 parallel_gemm_fm_cosma 81 10.0 2.077 2.083 2.077 2.083 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.927 1.951 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.854 1.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.815 1.828 mp_waitany 10164 13.8 1.238 1.803 1.238 1.803 multiply_cannon_metrocomm4 22605 15.4 0.072 0.074 0.780 1.685 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.601 1.612 acc_transpose_blocks 24660 15.4 0.114 0.117 1.568 1.595 mp_allgather_i34 2055 14.4 0.575 1.583 0.575 1.583 mp_irecv_dv 57340 16.2 0.659 1.568 0.659 1.568 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.215 0.970 1.492 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.370 1.477 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.118000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.818182, yerr=7.565690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 660.680704E+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_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.040 0.067 61.833 61.836 qs_mol_dyn_low 1 2.0 0.003 0.003 61.477 61.488 qs_forces 11 3.9 0.002 0.003 61.389 61.389 qs_energies 11 4.9 0.001 0.002 58.173 58.176 scf_env_do_scf 11 5.9 0.000 0.001 48.911 48.911 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.041 40.041 velocity_verlet 10 3.0 0.004 0.006 32.741 32.743 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.667 28.934 qs_scf_new_mos 99 7.5 0.001 0.001 24.825 24.941 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.824 24.940 ot_scf_mini 99 9.5 0.002 0.003 23.607 23.740 multiply_cannon 2055 13.4 0.214 0.242 22.343 23.733 multiply_cannon_loop 2055 14.4 0.616 0.632 21.142 22.086 ot_mini 99 10.5 0.001 0.001 13.529 13.664 rebuild_ks_matrix 110 8.3 0.000 0.001 12.816 12.951 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.815 12.951 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.327 11.446 multiply_cannon_multrec 16440 15.4 3.830 5.084 9.828 11.014 mp_waitall_1 139946 16.5 7.100 9.861 7.100 9.861 qs_ot_get_derivative 99 11.5 0.001 0.001 9.031 9.167 init_scf_loop 11 6.9 0.000 0.000 8.833 8.834 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.416 6.966 init_scf_run 11 5.9 0.000 0.001 6.862 6.862 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.862 6.862 prepare_preconditioner 11 7.9 0.000 0.000 6.731 6.752 make_preconditioner 11 8.9 0.000 0.000 6.731 6.752 sum_up_and_integrate 110 10.3 0.060 0.061 6.605 6.621 integrate_v_rspace 110 11.3 0.002 0.003 6.545 6.561 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.074 6.460 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.055 6.063 calculate_rho_elec 110 8.6 0.058 0.058 6.054 6.062 dbcsr_mm_accdrv_process 34862 16.1 4.743 5.410 5.852 6.027 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.040 5.487 apply_single 110 13.6 0.000 0.000 5.040 5.487 qs_ot_get_p 110 10.4 0.001 0.001 5.244 5.425 make_m2s 4110 13.4 0.050 0.051 4.218 4.592 make_images 4110 14.4 0.397 0.520 4.103 4.475 ot_diis_step 99 11.5 0.011 0.011 4.471 4.471 density_rs2pw 110 9.6 0.004 0.005 3.150 4.393 multiply_cannon_sync_h2d 16440 15.4 3.636 4.206 3.636 4.206 rs_pw_transfer 902 11.9 0.010 0.011 2.687 3.923 calculate_first_density_matrix 1 7.0 0.000 0.000 3.798 3.799 pw_transfer 1331 11.6 0.066 0.074 3.752 3.763 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.052 3.727 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.651 3.661 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.644 3.658 grid_integrate_task_list 110 12.3 3.162 3.418 3.162 3.418 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.303 3.303 fft_wrap_pw1pw2_140 451 13.1 0.214 0.219 3.147 3.161 wfi_extrapolate 11 7.9 0.001 0.001 2.978 2.978 fft3d_ps 1111 14.6 1.086 1.094 2.943 2.961 make_images_data 4110 15.4 0.044 0.047 2.474 2.943 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.902 2.903 hybrid_alltoall_any 4261 16.3 0.105 0.376 2.261 2.847 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.780 2.782 cp_fm_redistribute_end 48 14.0 1.734 2.754 1.736 2.754 cp_fm_diag_elpa_base 48 14.0 0.956 2.605 1.013 2.709 calculate_dm_sparse 110 9.5 0.001 0.001 2.642 2.676 cp_fm_cholesky_invert 11 10.9 2.633 2.639 2.633 2.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.565 2.623 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.911 2.530 mp_waitany 17072 13.8 1.184 2.490 1.184 2.490 grid_collocate_task_list 110 9.6 2.082 2.469 2.082 2.469 mp_irecv_dv 48980 15.7 0.841 2.403 0.841 2.403 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.311 2.396 potential_pw2rs 110 12.3 0.010 0.010 2.327 2.333 mp_alltoall_d11v 2046 13.8 1.816 2.156 1.816 2.156 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.180 0.901 2.137 mp_sum_l 6514 12.8 1.508 2.063 1.508 2.063 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.970 1.971 dbcsr_complete_redistribute 325 12.2 0.323 0.365 1.424 1.889 make_basis_sm 11 9.8 0.008 0.014 1.885 1.888 parallel_gemm_fm 81 9.0 0.000 0.000 1.862 1.868 parallel_gemm_fm_cosma 81 10.0 1.862 1.868 1.862 1.868 cp_fm_upper_to_full 70 13.6 1.399 1.854 1.399 1.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.707 mp_allgather_i34 2055 14.4 0.526 1.663 0.526 1.663 jit_kernel_multiply 8 16.5 0.717 1.624 0.717 1.624 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.606 1.618 cp_fm_cholesky_decompose 22 10.9 1.580 1.613 1.580 1.613 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.356 1.469 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.984 1.443 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.266 1.274 acc_transpose_blocks 16440 15.4 0.073 0.075 1.259 1.270 rs_gather_matrices 110 12.3 0.233 0.259 0.978 1.269 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.836000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=8.144359 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 726.585344E+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_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.040 0.104 65.430 65.432 qs_mol_dyn_low 1 2.0 0.003 0.004 64.978 64.987 qs_forces 11 3.9 0.003 0.005 64.906 64.907 qs_energies 11 4.9 0.004 0.005 61.470 61.475 scf_env_do_scf 11 5.9 0.001 0.001 52.877 52.880 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.369 41.369 velocity_verlet 10 3.0 0.011 0.013 36.834 36.836 dbcsr_multiply_generic 2055 12.4 0.113 0.116 29.615 29.815 qs_scf_new_mos 99 7.5 0.001 0.001 26.251 26.373 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.251 26.372 ot_scf_mini 99 9.5 0.003 0.004 24.634 24.750 multiply_cannon 2055 13.4 0.243 0.265 22.791 23.680 multiply_cannon_loop 2055 14.4 0.889 0.908 21.360 21.960 ot_mini 99 10.5 0.001 0.001 14.099 14.244 multiply_cannon_multrec 24660 15.4 4.229 6.964 12.987 14.144 rebuild_ks_matrix 110 8.3 0.000 0.000 12.264 12.367 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.029 12.264 12.366 init_scf_loop 11 6.9 0.001 0.003 11.466 11.466 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.823 10.916 qs_ot_get_derivative 99 11.5 0.001 0.001 9.956 10.079 prepare_preconditioner 11 7.9 0.000 0.000 9.715 9.732 make_preconditioner 11 8.9 0.000 0.001 9.715 9.732 dbcsr_mm_accdrv_process 52304 16.0 7.013 8.279 8.608 9.583 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.990 9.409 sum_up_and_integrate 110 10.3 0.067 0.069 6.542 6.555 integrate_v_rspace 110 11.3 0.003 0.003 6.475 6.489 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.143 6.153 calculate_rho_elec 110 8.6 0.077 0.081 6.142 6.152 mp_waitall_1 121746 16.5 4.107 6.011 4.107 6.011 init_scf_run 11 5.9 0.000 0.001 5.765 5.765 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.764 5.765 qs_ot_get_p 110 10.4 0.001 0.001 5.562 5.714 make_m2s 4110 13.4 0.060 0.063 5.249 5.631 make_images 4110 14.4 0.581 0.699 5.107 5.486 cp_fm_upper_to_full 70 13.8 3.230 4.575 3.230 4.575 ot_diis_step 99 11.5 0.011 0.011 4.106 4.107 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.019 4.098 apply_single 110 13.6 0.000 0.000 4.019 4.098 pw_transfer 1331 11.6 0.066 0.076 3.873 3.901 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.813 3.827 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.767 3.799 dbcsr_complete_redistribute 325 12.2 0.415 0.465 2.591 3.707 density_rs2pw 110 9.6 0.004 0.004 3.088 3.606 grid_integrate_task_list 110 12.3 3.269 3.491 3.269 3.491 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.324 3.380 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.337 3.337 multiply_cannon_sync_h2d 24660 15.4 3.168 3.331 3.168 3.331 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 3.224 3.261 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.441 3.226 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.078 3.172 fft3d_ps 1111 14.6 1.092 1.129 3.059 3.084 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.981 3.051 calculate_dm_sparse 110 9.5 0.001 0.001 3.014 3.049 wfi_extrapolate 11 7.9 0.001 0.001 2.995 2.995 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.233 2.981 make_images_data 4110 15.4 0.047 0.051 2.603 2.975 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.952 2.954 rs_pw_transfer 902 11.9 0.010 0.011 2.417 2.922 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.796 2.798 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.716 2.792 cp_fm_redistribute_end 48 14.0 1.398 2.767 1.400 2.768 mp_alltoall_i22 605 13.7 1.602 2.756 1.602 2.756 cp_fm_diag_elpa_base 48 14.0 1.288 2.631 1.366 2.741 cp_fm_cholesky_invert 11 10.9 2.701 2.710 2.701 2.710 calculate_first_density_matrix 1 7.0 0.001 0.002 2.678 2.680 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.589 2.619 grid_collocate_task_list 110 9.6 2.182 2.477 2.182 2.477 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.279 2.280 potential_pw2rs 110 12.3 0.013 0.013 2.214 2.219 jit_kernel_multiply 10 15.8 1.261 2.100 1.261 2.100 mp_alltoall_d11v 2046 13.8 1.680 1.911 1.680 1.911 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.689 1.725 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.603 1.701 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.687 1.699 cp_fm_cholesky_decompose 22 10.9 1.618 1.664 1.618 1.664 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.636 1.652 acc_transpose_blocks 24660 15.4 0.109 0.111 1.592 1.622 mp_waitany 13376 13.8 1.075 1.556 1.075 1.556 mp_sum_l 6514 12.8 0.963 1.511 0.963 1.511 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.854 1.502 mp_irecv_dv 62702 16.1 0.755 1.427 0.755 1.427 qs_env_update_s_mstruct 11 6.9 0.000 0.005 1.227 1.326 mp_allgather_i34 2055 14.4 0.530 1.324 0.530 1.324 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.174 0.814 1.320 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.432000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.090909, yerr=6.444935 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 831.897600E+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_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.046 0.061 56.095 56.099 qs_mol_dyn_low 1 2.0 0.003 0.003 55.628 55.638 qs_forces 11 3.9 0.002 0.003 55.565 55.566 qs_energies 11 4.9 0.001 0.002 51.843 51.848 scf_env_do_scf 11 5.9 0.000 0.001 43.353 43.353 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.626 35.626 velocity_verlet 10 3.0 0.002 0.002 31.298 31.301 dbcsr_multiply_generic 2055 12.4 0.105 0.106 23.002 23.142 qs_scf_new_mos 99 7.5 0.001 0.001 20.377 20.446 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.377 20.446 ot_scf_mini 99 9.5 0.002 0.002 19.145 19.204 multiply_cannon 2055 13.4 0.246 0.256 17.616 18.906 multiply_cannon_loop 2055 14.4 0.321 0.334 16.293 16.534 rebuild_ks_matrix 110 8.3 0.000 0.000 12.056 12.097 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.056 12.097 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.712 10.749 ot_mini 99 10.5 0.001 0.001 10.347 10.405 multiply_cannon_multrec 8220 15.4 3.215 4.535 7.675 8.869 init_scf_loop 11 6.9 0.000 0.000 7.678 7.679 mp_waitall_1 103326 16.6 6.027 7.639 6.027 7.639 sum_up_and_integrate 110 10.3 0.079 0.080 6.575 6.589 qs_ot_get_derivative 99 11.5 0.001 0.001 6.525 6.583 integrate_v_rspace 110 11.3 0.003 0.003 6.495 6.509 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.390 6.404 calculate_rho_elec 110 8.6 0.115 0.116 6.390 6.404 prepare_preconditioner 11 7.9 0.000 0.000 6.002 6.008 make_preconditioner 11 8.9 0.000 0.000 6.002 6.008 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.592 5.660 init_scf_run 11 5.9 0.000 0.001 5.348 5.348 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.348 5.348 dbcsr_mm_accdrv_process 17442 15.9 2.912 4.095 4.331 5.270 qs_ot_get_p 110 10.4 0.001 0.001 4.871 4.932 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.179 4.529 make_m2s 4110 13.4 0.038 0.040 4.152 4.417 make_images 4110 14.4 0.643 0.696 4.022 4.291 pw_transfer 1331 11.6 0.066 0.071 4.165 4.176 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.057 4.071 ot_diis_step 99 11.5 0.012 0.014 3.803 3.803 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.713 3.755 apply_single 110 13.6 0.000 0.000 3.712 3.755 grid_integrate_task_list 110 12.3 3.334 3.582 3.334 3.582 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.541 3.545 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.514 3.528 density_rs2pw 110 9.6 0.004 0.004 3.132 3.505 fft3d_ps 1111 14.6 1.144 1.170 3.300 3.314 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.238 3.239 multiply_cannon_sync_h2d 8220 15.4 2.890 3.053 2.890 3.053 cp_fm_cholesky_invert 11 10.9 2.903 2.907 2.903 2.907 make_images_data 4110 15.4 0.038 0.043 2.355 2.774 wfi_extrapolate 11 7.9 0.001 0.001 2.753 2.753 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.716 2.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.692 2.693 cp_fm_redistribute_end 48 14.0 0.691 2.688 0.695 2.689 hybrid_alltoall_any 4261 16.3 0.201 0.868 2.297 2.667 cp_fm_diag_elpa_base 48 14.0 1.808 2.495 1.987 2.658 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.637 2.640 calculate_dm_sparse 110 9.5 0.001 0.001 2.579 2.620 rs_pw_transfer 902 11.9 0.010 0.010 2.161 2.560 grid_collocate_task_list 110 9.6 2.277 2.551 2.277 2.551 calculate_first_density_matrix 1 7.0 0.000 0.000 2.498 2.499 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.300 2.313 potential_pw2rs 110 12.3 0.015 0.015 2.121 2.123 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.776 2.006 mp_alltoall_d11v 2046 13.8 1.657 1.969 1.657 1.969 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.936 1.963 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.768 1.797 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.680 1.685 cp_fm_cholesky_decompose 22 10.9 1.659 1.680 1.659 1.680 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.527 1.656 jit_kernel_multiply 8 15.6 1.106 1.626 1.106 1.626 mp_allgather_i34 2055 14.4 0.512 1.579 0.512 1.579 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.548 1.559 dbcsr_complete_redistribute 325 12.2 0.554 0.577 1.415 1.500 mp_waitany 9240 13.8 1.046 1.472 1.046 1.472 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.482 1.386 qs_create_task_list 11 7.9 0.008 0.015 1.236 1.342 generate_qs_task_list 11 8.9 0.378 0.445 1.228 1.342 mp_irecv_dv 24056 15.7 0.458 1.339 0.458 1.339 rs_gather_matrices 110 12.3 0.324 0.377 0.952 1.259 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.166 0.792 1.199 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.139 1.179 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.099000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=785.000000, yerr=11.622861 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 1.391366E+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_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.053 87.669 87.671 qs_mol_dyn_low 1 2.0 0.003 0.003 87.286 87.294 qs_forces 11 3.9 0.002 0.003 87.212 87.217 qs_energies 11 4.9 0.002 0.002 83.003 83.009 scf_env_do_scf 11 5.9 0.001 0.001 72.708 72.708 velocity_verlet 10 3.0 0.007 0.009 55.941 55.947 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.489 44.491 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.814 28.850 init_scf_loop 11 6.9 0.000 0.000 28.146 28.150 prepare_preconditioner 11 7.9 0.000 0.000 26.088 26.095 make_preconditioner 11 8.9 0.000 0.000 26.088 26.095 qs_scf_new_mos 99 7.5 0.001 0.001 26.011 26.046 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.010 26.046 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.460 25.562 ot_scf_mini 99 9.5 0.002 0.002 24.233 24.255 multiply_cannon 2055 13.4 0.348 0.381 21.724 22.508 multiply_cannon_loop 2055 14.4 0.341 0.346 19.922 20.344 cp_fm_upper_to_full 70 14.2 12.573 17.911 12.573 17.911 rebuild_ks_matrix 110 8.3 0.001 0.001 14.278 14.315 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.278 14.315 ot_mini 99 10.5 0.001 0.001 13.550 13.574 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.934 12.968 dbcsr_complete_redistribute 325 12.2 1.025 1.045 7.209 10.295 multiply_cannon_multrec 8220 15.4 4.405 4.589 9.816 9.932 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.174 9.268 qs_ot_get_derivative 99 11.5 0.001 0.001 8.973 8.999 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.612 8.679 mp_alltoall_i22 605 13.7 5.234 8.342 5.234 8.342 mp_waitall_1 84994 16.7 7.534 8.319 7.534 8.319 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.823 7.858 calculate_rho_elec 110 8.6 0.227 0.227 7.822 7.858 sum_up_and_integrate 110 10.3 0.150 0.151 7.405 7.419 integrate_v_rspace 110 11.3 0.004 0.004 7.254 7.268 init_scf_run 11 5.9 0.000 0.001 6.009 6.009 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.009 6.009 make_m2s 4110 13.4 0.043 0.043 5.360 5.889 make_images 4110 14.4 0.883 0.919 5.173 5.701 qs_ot_get_p 110 10.4 0.001 0.001 5.492 5.522 dbcsr_mm_accdrv_process 11614 15.7 3.265 3.661 5.268 5.514 cp_fm_cholesky_invert 11 10.9 5.374 5.379 5.374 5.379 pw_transfer 1331 11.6 0.076 0.077 5.363 5.371 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.245 5.253 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.694 5.149 apply_single 110 13.6 0.000 0.000 4.694 5.148 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.664 5.130 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.568 4.577 ot_diis_step 99 11.5 0.015 0.015 4.552 4.552 fft3d_ps 1111 14.6 1.294 1.305 4.441 4.452 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.990 3.997 multiply_cannon_sync_h2d 8220 15.4 3.942 3.949 3.942 3.949 density_rs2pw 110 9.6 0.004 0.004 3.842 3.876 hybrid_alltoall_any 4261 16.3 0.257 0.557 2.972 3.716 grid_integrate_task_list 110 12.3 3.658 3.716 3.658 3.716 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.689 3.694 make_images_data 4110 15.4 0.042 0.045 2.993 3.632 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.550 3.550 wfi_extrapolate 11 7.9 0.001 0.001 3.326 3.326 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.855 3.307 calculate_dm_sparse 110 9.5 0.001 0.001 3.268 3.291 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.041 3.045 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.955 2.955 cp_fm_diag_elpa_base 48 14.0 2.432 2.619 2.953 2.953 grid_collocate_task_list 110 9.6 2.630 2.659 2.630 2.659 calculate_first_density_matrix 1 7.0 0.000 0.000 2.560 2.561 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.547 2.560 potential_pw2rs 110 12.3 0.021 0.022 2.518 2.524 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.382 2.393 rs_pw_transfer 902 11.9 0.011 0.011 2.261 2.291 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.217 2.275 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.085 2.196 mp_alltoall_d11v 2046 13.8 1.959 2.055 1.959 2.055 cp_fm_cholesky_decompose 22 10.9 1.998 2.015 1.998 2.015 jit_kernel_multiply 10 15.3 1.800 1.965 1.800 1.965 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.947 1.951 qs_create_task_list 11 7.9 0.001 0.001 1.897 1.950 generate_qs_task_list 11 8.9 0.734 0.788 1.897 1.950 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.787 1.800 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.748 1.781 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.671000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1249.545455, yerr=66.499984 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.678080E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520949. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 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 0.015 0.042 203.740 203.744 qs_mol_dyn_low 1 2.0 0.003 0.003 203.310 203.340 qs_forces 11 3.9 0.003 0.004 203.127 203.128 qs_energies 11 4.9 0.001 0.002 197.545 197.555 scf_env_do_scf 11 5.9 0.001 0.002 181.125 181.129 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 160.643 160.645 dbcsr_multiply_generic 2507 12.6 0.177 0.182 123.830 124.613 velocity_verlet 10 3.0 0.003 0.006 122.649 122.650 qs_scf_new_mos 117 7.6 0.001 0.001 120.390 120.650 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.389 120.649 ot_scf_mini 117 9.6 0.003 0.003 113.857 114.080 multiply_cannon 2507 13.6 0.235 0.243 100.520 102.579 multiply_cannon_loop 2507 14.6 2.130 2.199 98.324 100.214 ot_mini 117 10.6 0.001 0.001 65.585 65.831 multiply_cannon_multrec 60168 15.6 33.089 35.432 41.516 44.074 qs_ot_get_derivative 117 11.6 0.001 0.001 40.820 41.063 rebuild_ks_matrix 128 8.3 0.001 0.001 33.827 34.200 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.022 33.826 34.199 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.400 30.774 mp_waitall_1 267128 16.5 28.011 30.502 28.011 30.502 multiply_cannon_sync_h2d 60168 15.6 27.190 29.194 27.190 29.194 qs_ot_get_p 128 10.4 0.001 0.001 26.468 26.853 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.200 25.334 apply_single 128 13.6 0.001 0.001 24.200 25.334 ot_diis_step 117 11.6 0.008 0.008 24.434 24.435 init_scf_loop 11 6.9 0.000 0.000 20.404 20.405 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.817 19.905 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.107 19.274 multiply_cannon_metrocomm3 60168 15.6 0.115 0.120 15.506 18.275 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.257 17.259 prepare_preconditioner 11 7.9 0.000 0.000 15.812 15.868 make_preconditioner 11 8.9 0.000 0.000 15.812 15.868 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.041 15.218 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.483 14.487 cp_fm_redistribute_end 83 14.4 11.457 14.418 11.468 14.421 sum_up_and_integrate 128 10.3 0.089 0.106 14.303 14.316 integrate_v_rspace 128 11.3 0.003 0.004 14.214 14.231 cp_fm_diag_elpa_base 83 14.4 2.912 14.040 2.942 14.170 make_m2s 5014 13.6 0.104 0.114 13.791 14.136 make_images 5014 14.6 0.401 0.420 13.610 13.962 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.850 13.943 calculate_rho_elec 128 8.7 0.045 0.065 13.850 13.942 init_scf_run 11 5.9 0.000 0.001 12.288 12.288 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.287 12.288 density_rs2pw 128 9.7 0.006 0.007 7.226 10.356 mp_sum_l 7870 13.0 8.251 9.630 8.251 9.630 wfi_extrapolate 11 7.9 0.001 0.001 9.070 9.070 cp_fm_cholesky_invert 11 10.9 8.972 8.979 8.972 8.979 rs_pw_transfer 1046 11.9 0.016 0.018 5.701 8.827 dbcsr_mm_accdrv_process 124484 16.2 3.329 3.462 7.990 8.525 calculate_dm_sparse 128 9.5 0.001 0.001 8.334 8.398 pw_transfer 1547 11.6 0.077 0.104 8.088 8.324 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.882 8.126 multiply_cannon_metrocomm1 60168 15.6 0.089 0.095 5.978 8.070 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.856 7.979 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.755 7.897 make_images_data 5014 15.6 0.067 0.073 6.808 7.724 grid_integrate_task_list 128 12.3 7.062 7.488 7.062 7.488 hybrid_alltoall_any 5200 16.5 0.290 2.247 5.976 7.257 fft_wrap_pw1pw2_140 523 13.2 0.445 0.512 6.868 7.155 fft3d_ps 1291 14.7 2.106 2.895 6.672 6.854 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.649 6.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.797 5.902 grid_collocate_task_list 128 9.7 4.551 5.814 4.551 5.814 mp_waitany 16020 13.9 2.627 5.776 2.627 5.776 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.296 2.103 5.269 mp_alltoall_d11v 2415 14.1 4.194 5.171 4.194 5.171 potential_pw2rs 128 12.3 0.009 0.011 4.701 4.726 cp_fm_cholesky_decompose 22 10.9 4.646 4.659 4.646 4.659 mp_sum_d 4455 12.2 3.536 4.349 3.536 4.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.744000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.799368 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 828.952576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5619064. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.062 189.618 189.626 qs_mol_dyn_low 1 2.0 0.003 0.004 189.150 189.165 qs_forces 11 3.9 0.003 0.005 189.068 189.070 qs_energies 11 4.9 0.003 0.014 182.243 182.254 scf_env_do_scf 11 5.9 0.001 0.001 165.945 165.955 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 133.177 133.179 velocity_verlet 10 3.0 0.002 0.002 119.484 119.485 dbcsr_multiply_generic 2507 12.6 0.184 0.190 97.206 98.304 qs_scf_new_mos 117 7.6 0.001 0.001 92.868 93.439 qs_scf_loop_do_ot 117 8.6 0.001 0.002 92.867 93.439 ot_scf_mini 117 9.6 0.004 0.005 88.066 88.707 multiply_cannon 2507 13.6 0.480 0.533 77.318 81.717 multiply_cannon_loop 2507 14.6 1.254 1.291 74.110 76.364 ot_mini 117 10.6 0.001 0.001 49.418 49.995 mp_waitall_1 214728 16.6 24.241 37.373 24.241 37.373 multiply_cannon_multrec 30084 15.6 22.173 26.100 31.790 36.299 rebuild_ks_matrix 128 8.3 0.001 0.001 33.563 34.071 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.563 34.070 init_scf_loop 11 6.9 0.001 0.006 32.681 32.687 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.202 30.662 prepare_preconditioner 11 7.9 0.000 0.001 28.180 28.242 make_preconditioner 11 8.9 0.000 0.001 28.180 28.242 qs_ot_get_derivative 117 11.6 0.001 0.002 27.532 28.164 multiply_cannon_metrocomm3 30084 15.6 0.095 0.099 15.457 27.696 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.902 27.441 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.992 23.089 apply_single 128 13.6 0.001 0.001 21.992 23.088 multiply_cannon_sync_h2d 30084 15.6 19.423 22.214 19.423 22.214 ot_diis_step 117 11.6 0.014 0.015 21.715 21.717 qs_ot_get_p 128 10.4 0.001 0.004 20.445 21.121 cp_fm_cholesky_invert 11 10.9 16.529 16.541 16.529 16.541 make_m2s 5014 13.6 0.090 0.095 14.292 15.853 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.658 15.690 make_images 5014 14.6 1.175 1.364 14.083 15.645 sum_up_and_integrate 128 10.3 0.113 0.131 14.807 14.842 integrate_v_rspace 128 11.3 0.003 0.003 14.693 14.734 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.533 14.534 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.136 14.164 calculate_rho_elec 128 8.7 0.088 0.106 14.135 14.163 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.611 11.617 cp_fm_redistribute_end 83 14.4 6.782 11.552 6.793 11.553 init_scf_run 11 5.9 0.000 0.001 11.488 11.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.487 11.489 cp_fm_diag_elpa_base 83 14.4 4.528 11.043 4.745 11.410 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.814 11.236 multiply_cannon_metrocomm4 27577 15.6 0.094 0.108 3.706 10.598 make_images_data 5014 15.6 0.065 0.073 8.727 10.576 mp_irecv_dv 69486 16.3 3.514 10.213 3.514 10.213 density_rs2pw 128 9.7 0.006 0.007 7.489 9.999 hybrid_alltoall_any 5200 16.5 0.343 1.523 7.404 9.700 dbcsr_mm_accdrv_process 62242 16.2 4.547 5.311 9.073 9.674 pw_transfer 1547 11.6 0.087 0.113 9.008 9.081 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.781 8.859 wfi_extrapolate 11 7.9 0.001 0.001 8.415 8.415 rs_pw_transfer 1046 11.9 0.014 0.017 5.681 8.179 fft_wrap_pw1pw2_140 523 13.2 0.471 0.525 7.779 7.886 grid_integrate_task_list 128 12.3 7.192 7.585 7.192 7.585 fft3d_ps 1291 14.7 2.753 2.935 7.151 7.188 cp_fm_cholesky_decompose 22 10.9 6.852 6.921 6.852 6.921 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.144 6.867 calculate_dm_sparse 128 9.5 0.001 0.001 6.450 6.572 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.142 6.151 grid_collocate_task_list 128 9.7 4.710 5.878 4.710 5.878 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.294 5.469 mp_sum_l 7870 13.0 3.823 5.456 3.823 5.456 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.285 5.344 potential_pw2rs 128 12.3 0.015 0.018 5.067 5.086 mp_waitany 11748 13.9 2.513 5.046 2.513 5.046 mp_allgather_i34 2507 14.6 1.787 5.011 1.787 5.011 rs_pw_transfer_RS2PW_140 139 11.5 0.350 0.377 2.099 4.584 mp_alltoall_d11v 2415 14.1 3.902 4.456 3.902 4.456 mp_sum_d 4455 12.2 2.619 3.940 2.619 3.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.767 3.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.626000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.090909, yerr=2.108621 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/17/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 2.951022E+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 4017216 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.3 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 941.756416E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1052064 MPI messages size (bytes): total size 2.737208E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.601750E+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 284544 37295751168 131072 < size <= 4194304 665856 1004519030784 4194304 < size <= 16777216 66080 937891119088 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56602. MP_Allreduce 11361 981. MP_Sync 172 MP_Alltoall 1724 9394234. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_comm_split 84 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 177.028 177.029 qs_mol_dyn_low 1 2.0 0.003 0.005 176.542 176.555 qs_forces 11 3.9 0.003 0.005 176.278 176.289 qs_energies 11 4.9 0.003 0.007 169.688 169.697 scf_env_do_scf 11 5.9 0.001 0.003 154.059 154.060 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 119.210 119.211 velocity_verlet 10 3.0 0.001 0.002 113.909 113.916 dbcsr_multiply_generic 2529 12.6 0.182 0.185 81.263 82.354 qs_scf_new_mos 118 7.6 0.001 0.001 80.765 81.116 qs_scf_loop_do_ot 118 8.6 0.001 0.002 80.764 81.115 ot_scf_mini 118 9.6 0.004 0.005 76.621 77.031 multiply_cannon 2529 13.6 0.506 0.530 61.862 65.736 multiply_cannon_loop 2529 14.6 0.865 0.891 58.639 61.258 ot_mini 118 10.6 0.001 0.001 42.482 42.883 init_scf_loop 11 6.9 0.001 0.005 34.749 34.750 mp_waitall_1 172006 16.6 23.885 34.092 23.885 34.092 rebuild_ks_matrix 129 8.3 0.001 0.001 30.852 31.283 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 30.852 31.282 prepare_preconditioner 11 7.9 0.000 0.000 30.720 30.770 make_preconditioner 11 8.9 0.000 0.002 30.719 30.770 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.365 29.803 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.821 28.222 multiply_cannon_multrec 20232 15.6 13.830 16.843 22.756 25.785 multiply_cannon_metrocomm3 20232 15.6 0.061 0.064 14.487 24.591 qs_ot_get_derivative 118 11.6 0.002 0.002 22.570 22.974 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.999 20.937 apply_single 129 13.6 0.001 0.001 19.998 20.937 qs_ot_get_p 129 10.4 0.001 0.001 19.616 20.053 ot_diis_step 118 11.6 0.018 0.019 19.814 19.815 multiply_cannon_sync_h2d 20232 15.6 14.490 16.083 14.490 16.083 qs_ot_p2m_diag 84 11.4 0.269 0.276 15.459 15.467 make_m2s 5058 13.6 0.082 0.086 14.422 15.303 make_images 5058 14.6 1.198 1.278 14.187 15.066 sum_up_and_integrate 129 10.3 0.133 0.144 14.800 14.827 integrate_v_rspace 129 11.3 0.004 0.004 14.667 14.694 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.447 14.494 calculate_rho_elec 129 8.7 0.133 0.148 14.446 14.493 cp_dbcsr_syevd 84 12.4 0.005 0.006 14.438 14.439 cp_fm_cholesky_invert 11 10.9 14.342 14.350 14.342 14.350 cp_fm_diag_elpa 84 13.4 0.000 0.001 11.259 11.260 cp_fm_redistribute_end 84 14.4 4.257 11.200 4.271 11.202 cp_fm_diag_elpa_base 84 14.4 6.501 10.609 6.912 11.082 init_scf_run 11 5.9 0.000 0.001 10.606 10.606 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.605 10.606 make_images_data 5058 15.6 0.062 0.069 8.800 10.028 hybrid_alltoall_any 5245 16.5 0.439 2.006 7.704 9.392 density_rs2pw 129 9.7 0.006 0.006 7.296 9.373 multiply_cannon_metrocomm4 17703 15.6 0.062 0.071 3.366 9.319 pw_transfer 1559 11.6 0.088 0.108 9.180 9.297 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.855 9.146 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.952 9.077 mp_irecv_dv 50659 16.2 3.243 9.073 3.243 9.073 dbcsr_mm_accdrv_process 41846 16.2 4.539 5.429 8.369 8.513 fft_wrap_pw1pw2_140 527 13.2 0.481 0.529 7.906 8.035 grid_integrate_task_list 129 12.3 7.362 7.861 7.362 7.861 cp_fm_upper_to_full 106 14.5 5.738 7.411 5.738 7.411 wfi_extrapolate 11 7.9 0.001 0.001 7.377 7.378 cp_fm_cholesky_decompose 22 10.9 7.279 7.312 7.279 7.312 fft3d_ps 1301 14.7 2.690 2.926 7.200 7.285 rs_pw_transfer 1054 12.0 0.014 0.015 5.139 7.194 dbcsr_complete_redistribute 397 12.7 1.183 1.233 4.590 6.346 calculate_dm_sparse 129 9.5 0.001 0.001 5.767 5.857 grid_collocate_task_list 129 9.7 4.948 5.789 4.948 5.789 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.569 5.575 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.551 5.231 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.433 5.187 mp_alltoall_d11v 2429 14.1 4.192 5.174 4.192 5.174 potential_pw2rs 129 12.3 0.020 0.023 4.863 4.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.651 4.775 mp_allgather_i34 2529 14.6 1.740 4.732 1.740 4.732 mp_sum_l 7936 13.1 3.143 4.587 3.143 4.587 mp_waitany 11836 13.9 2.319 4.410 2.319 4.410 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.042 4.076 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.335 4.014 rs_pw_transfer_RS2PW_140 140 11.5 0.333 0.353 1.914 3.943 qs_energies_init_hamiltonians 11 5.9 0.002 0.011 3.781 3.786 mp_alltoall_i22 720 14.1 1.946 3.778 1.946 3.778 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.547 3.582 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.029000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.272727, yerr=9.845081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 4.353792E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 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 1.126715E+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 665379572320 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 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 0.036 0.057 187.893 187.894 qs_mol_dyn_low 1 2.0 0.003 0.004 187.356 187.370 qs_forces 11 3.9 0.003 0.005 187.248 187.253 qs_energies 11 4.9 0.011 0.016 180.054 180.066 scf_env_do_scf 11 5.9 0.001 0.002 163.395 163.403 velocity_verlet 10 3.0 0.002 0.002 124.867 124.870 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 116.856 116.857 qs_scf_new_mos 117 7.6 0.001 0.001 79.347 79.630 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.346 79.629 dbcsr_multiply_generic 2507 12.6 0.189 0.194 78.789 79.476 ot_scf_mini 117 9.6 0.003 0.004 74.875 75.168 multiply_cannon 2507 13.6 0.556 0.589 54.440 57.853 multiply_cannon_loop 2507 14.6 1.187 1.212 50.644 52.264 init_scf_loop 11 6.9 0.001 0.006 46.417 46.418 ot_mini 117 10.6 0.001 0.001 42.057 42.341 prepare_preconditioner 11 7.9 0.000 0.000 42.279 42.315 make_preconditioner 11 8.9 0.000 0.001 42.279 42.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.978 40.905 multiply_cannon_multrec 30084 15.6 14.285 19.887 26.239 31.300 rebuild_ks_matrix 128 8.3 0.001 0.001 29.937 30.241 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 29.936 30.241 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.975 27.250 mp_waitall_1 147882 16.7 16.566 25.788 16.566 25.788 qs_ot_get_derivative 117 11.6 0.001 0.002 22.381 22.679 make_m2s 5014 13.6 0.095 0.099 20.064 21.084 make_images 5014 14.6 1.990 2.278 19.757 20.778 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.121 19.633 apply_single 128 13.6 0.001 0.001 19.120 19.633 ot_diis_step 117 11.6 0.017 0.018 19.544 19.546 qs_ot_get_p 128 10.4 0.001 0.001 19.111 19.414 cp_fm_cholesky_invert 11 10.9 16.339 16.349 16.339 16.349 cp_fm_upper_to_full 105 14.7 10.899 16.119 10.899 16.119 qs_ot_p2m_diag 83 11.4 0.343 0.390 15.132 15.184 sum_up_and_integrate 128 10.3 0.141 0.153 14.787 14.814 integrate_v_rspace 128 11.3 0.003 0.004 14.645 14.680 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.591 14.623 calculate_rho_elec 128 8.7 0.175 0.190 14.590 14.622 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.007 14.538 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.846 13.848 multiply_cannon_sync_h2d 30084 15.6 11.745 12.702 11.745 12.702 dbcsr_complete_redistribute 395 12.7 1.514 1.624 8.956 12.652 make_images_data 5014 15.6 0.065 0.069 10.691 12.599 dbcsr_mm_accdrv_process 62264 16.2 7.263 8.332 11.524 12.022 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.609 11.295 hybrid_alltoall_any 5200 16.5 0.528 2.217 9.648 11.206 init_scf_run 11 5.9 0.000 0.001 10.787 10.789 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.787 10.789 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.620 10.622 cp_fm_redistribute_end 83 14.4 1.834 10.542 1.847 10.547 cp_fm_diag_elpa_base 83 14.4 8.091 9.932 8.664 10.418 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.277 9.872 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.225 9.431 pw_transfer 1547 11.6 0.089 0.109 9.219 9.306 mp_alltoall_i22 716 14.1 5.496 9.171 5.496 9.171 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.989 9.080 density_rs2pw 128 9.7 0.006 0.006 7.063 8.282 fft_wrap_pw1pw2_140 523 13.2 0.483 0.496 7.972 8.067 grid_integrate_task_list 128 12.3 7.492 7.983 7.492 7.983 cp_fm_cholesky_decompose 22 10.9 7.649 7.763 7.649 7.763 wfi_extrapolate 11 7.9 0.001 0.001 7.640 7.640 fft3d_ps 1291 14.7 2.775 2.865 7.214 7.290 multiply_cannon_metrocomm4 25070 15.6 0.075 0.084 2.749 7.124 mp_irecv_dv 76098 16.2 2.604 6.862 2.604 6.862 calculate_dm_sparse 128 9.5 0.001 0.001 6.123 6.187 rs_pw_transfer 1046 11.9 0.013 0.014 4.803 6.070 grid_collocate_task_list 128 9.7 5.084 5.728 5.084 5.728 mp_alltoall_d11v 2415 14.1 4.800 5.531 4.800 5.531 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.433 5.470 potential_pw2rs 128 12.3 0.022 0.023 4.671 4.690 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.389 4.448 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 4.439 4.443 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.325 4.418 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.187 4.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.894000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1059.090909, yerr=13.331129 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.498120E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58165. MP_Allreduce 11010 1175. MP_Sync 86 MP_Alltoall 1700 18828189. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 173.315 173.316 qs_mol_dyn_low 1 2.0 0.003 0.003 172.863 172.876 qs_forces 11 3.9 0.003 0.003 172.750 172.756 qs_energies 11 4.9 0.001 0.002 165.209 165.219 scf_env_do_scf 11 5.9 0.001 0.001 147.714 147.724 velocity_verlet 10 3.0 0.007 0.010 113.328 113.332 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 111.706 111.708 qs_scf_new_mos 116 7.6 0.001 0.001 73.597 73.673 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.597 73.672 dbcsr_multiply_generic 2485 12.5 0.181 0.185 72.693 73.034 ot_scf_mini 116 9.6 0.004 0.004 69.324 69.348 multiply_cannon 2485 13.5 0.581 0.618 53.585 57.668 multiply_cannon_loop 2485 14.5 0.443 0.453 48.920 49.513 ot_mini 116 10.6 0.001 0.001 38.746 38.766 init_scf_loop 11 6.9 0.000 0.000 35.860 35.862 prepare_preconditioner 11 7.9 0.000 0.000 31.937 31.962 make_preconditioner 11 8.9 0.000 0.000 31.937 31.962 mp_waitall_1 124680 16.7 24.325 31.287 24.325 31.287 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.823 30.087 rebuild_ks_matrix 127 8.3 0.001 0.001 29.472 29.496 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 29.472 29.496 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.754 26.775 multiply_cannon_multrec 9940 15.5 10.402 14.048 17.983 20.948 ot_diis_step 116 11.6 0.019 0.020 19.515 19.516 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.202 19.407 apply_single 127 13.6 0.001 0.001 19.202 19.407 qs_ot_get_derivative 116 11.6 0.002 0.002 19.161 19.183 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 11.885 18.447 cp_fm_cholesky_invert 11 10.9 18.312 18.318 18.312 18.318 qs_ot_get_p 127 10.4 0.001 0.001 17.973 18.011 make_m2s 4970 13.5 0.066 0.069 15.505 17.949 make_images 4970 14.5 2.298 2.624 15.202 17.644 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.242 15.293 calculate_rho_elec 127 8.7 0.256 0.266 15.242 15.293 sum_up_and_integrate 127 10.3 0.179 0.189 14.918 14.970 integrate_v_rspace 127 11.3 0.004 0.004 14.738 14.798 qs_ot_p2m_diag 82 11.4 0.490 0.497 14.368 14.384 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.252 13.254 multiply_cannon_sync_h2d 9940 15.5 11.541 11.982 11.541 11.982 make_images_data 4970 15.5 0.052 0.059 9.259 11.665 hybrid_alltoall_any 5155 16.4 0.815 3.815 9.067 11.370 init_scf_run 11 5.9 0.000 0.001 10.661 10.661 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.661 10.661 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.209 10.219 cp_fm_diag_elpa_base 82 14.4 9.967 10.045 10.201 10.212 pw_transfer 1535 11.6 0.086 0.095 9.590 9.625 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.368 9.410 fft_wrap_pw1pw2_140 519 13.2 0.491 0.511 8.249 8.292 cp_fm_cholesky_decompose 22 10.9 8.009 8.201 8.009 8.201 grid_integrate_task_list 127 12.3 7.671 8.193 7.671 8.193 dbcsr_mm_accdrv_process 20590 16.1 2.638 3.408 7.213 7.897 density_rs2pw 127 9.7 0.005 0.006 6.937 7.806 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.621 7.642 fft3d_ps 1281 14.7 2.681 2.760 7.551 7.567 wfi_extrapolate 11 7.9 0.001 0.001 7.438 7.438 mp_allgather_i34 2485 14.5 2.893 7.090 2.893 7.090 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 3.909 6.409 calculate_dm_sparse 127 9.5 0.001 0.001 6.017 6.109 grid_collocate_task_list 127 9.7 5.348 6.002 5.348 6.002 mp_alltoall_d11v 2401 14.1 4.846 5.814 4.846 5.814 dbcsr_complete_redistribute 393 12.7 2.132 2.200 5.164 5.601 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.363 5.375 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.320 5.322 rs_pw_transfer 1038 11.9 0.012 0.013 4.193 5.098 potential_pw2rs 127 12.3 0.026 0.027 4.537 4.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.262 4.286 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.558 3.871 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.825 3.816 mp_irecv_dv 28618 15.9 1.787 3.745 1.787 3.745 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.399 3.732 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.660 3.680 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.617 3.625 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.501 3.606 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.316000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1393.454545, yerr=40.082559 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 3.005587E+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 36954374. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 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 0.039 0.086 297.971 297.986 qs_mol_dyn_low 1 2.0 0.003 0.003 297.335 297.348 qs_forces 11 3.9 0.003 0.003 297.220 297.222 qs_energies 11 4.9 0.002 0.002 287.770 287.774 scf_env_do_scf 11 5.9 0.001 0.001 265.106 265.119 velocity_verlet 10 3.0 0.006 0.008 214.014 214.024 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 140.975 140.977 init_scf_loop 11 6.9 0.000 0.000 123.872 123.873 prepare_preconditioner 11 7.9 0.000 0.000 118.664 118.710 make_preconditioner 11 8.9 0.000 0.000 118.664 118.710 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.466 115.768 qs_scf_new_mos 116 7.6 0.001 0.001 90.391 90.553 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.390 90.552 ot_scf_mini 116 9.6 0.004 0.004 85.580 85.697 dbcsr_multiply_generic 2485 12.5 0.212 0.219 83.913 84.757 cp_fm_upper_to_full 104 14.8 52.987 75.047 52.987 75.047 multiply_cannon 2485 13.5 0.706 0.760 59.179 60.088 multiply_cannon_loop 2485 14.5 0.464 0.475 55.426 57.483 ot_mini 116 10.6 0.001 0.001 44.897 45.037 dbcsr_complete_redistribute 393 12.7 3.981 4.057 28.885 41.154 rebuild_ks_matrix 127 8.3 0.001 0.001 38.419 38.584 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.418 38.584 copy_fm_to_dbcsr 208 11.6 0.002 0.002 25.464 37.746 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.320 35.475 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.154 35.241 cp_fm_cholesky_invert 11 10.9 33.418 33.426 33.418 33.426 mp_alltoall_i22 712 14.1 20.946 33.389 20.946 33.389 mp_waitall_1 102768 16.8 28.725 33.203 28.725 33.203 qs_ot_get_p 127 10.4 0.001 0.001 25.337 25.479 qs_ot_get_derivative 116 11.6 0.002 0.002 24.155 24.272 qs_ot_p2m_diag 82 11.4 0.867 0.873 21.299 21.329 make_m2s 4970 13.5 0.075 0.078 19.987 21.089 multiply_cannon_metrocomm3 9940 15.5 0.024 0.025 19.158 21.042 ot_diis_step 116 11.6 0.022 0.023 20.698 20.698 make_images 4970 14.5 3.773 3.889 19.514 20.617 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.514 20.538 calculate_rho_elec 127 8.7 0.478 0.479 20.514 20.537 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.796 20.219 apply_single 127 13.6 0.001 0.001 19.796 20.219 sum_up_and_integrate 127 10.3 0.319 0.320 19.651 19.733 cp_dbcsr_syevd 82 12.4 0.005 0.006 19.604 19.605 integrate_v_rspace 127 11.3 0.004 0.004 19.332 19.416 multiply_cannon_multrec 9940 15.5 10.445 12.201 18.113 18.191 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.442 16.443 cp_fm_diag_elpa_base 82 14.4 12.133 13.673 16.439 16.439 multiply_cannon_sync_h2d 9940 15.5 15.528 15.537 15.528 15.537 hybrid_alltoall_any 5155 16.4 1.293 3.014 10.982 12.958 make_images_data 4970 15.5 0.061 0.064 10.771 12.647 init_scf_run 11 5.9 0.000 0.001 12.634 12.634 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.634 12.634 pw_transfer 1535 11.6 0.093 0.093 12.206 12.221 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.972 11.987 fft_wrap_pw1pw2_140 519 13.2 0.538 0.540 10.640 10.654 fft3d_ps 1281 14.7 2.724 2.740 10.022 10.038 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.322 9.413 wfi_extrapolate 11 7.9 0.001 0.001 9.253 9.253 dbcsr_mm_accdrv_process 20590 16.0 3.769 5.621 7.433 9.213 cp_fm_cholesky_decompose 22 10.9 8.802 8.852 8.802 8.852 mp_alltoall_d11v 2401 14.1 8.104 8.834 8.104 8.834 grid_integrate_task_list 127 12.3 8.522 8.715 8.522 8.715 density_rs2pw 127 9.7 0.005 0.005 8.246 8.367 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.064 8.065 calculate_dm_sparse 127 9.5 0.001 0.001 6.776 6.844 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.632 6.679 grid_collocate_task_list 127 9.7 6.311 6.364 6.311 6.364 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.079 6.153 rs_scatter_matrices 138 9.7 3.639 4.515 5.930 6.145 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=297.986000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2683.909091, yerr=147.093325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 1.261961E+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 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.054 0.124 85.597 85.598 qs_energies 1 2.0 0.001 0.011 85.030 85.057 ls_scf 1 3.0 0.001 0.010 84.113 84.122 dbcsr_multiply_generic 111 6.7 0.015 0.015 73.119 73.261 multiply_cannon 111 7.7 0.017 0.020 56.238 57.958 multiply_cannon_loop 111 8.7 0.207 0.220 52.729 54.505 ls_scf_main 1 4.0 0.001 0.019 52.546 52.547 density_matrix_trs4 2 5.0 0.003 0.018 46.914 47.021 ls_scf_init_scf 1 4.0 0.000 0.001 28.498 28.499 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.323 27.369 mp_waitall_1 11031 10.9 22.929 25.587 22.929 25.587 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.208 25.239 multiply_cannon_multrec 2664 9.7 8.131 8.859 15.552 17.300 multiply_cannon_sync_h2d 2664 9.7 13.415 15.630 13.415 15.630 make_m2s 222 7.7 0.008 0.011 13.058 13.691 make_images 222 8.7 0.099 0.110 13.036 13.670 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 10.006 12.543 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.615 9.454 make_images_data 222 9.7 0.004 0.005 7.639 8.212 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.629 7.038 8.052 hybrid_alltoall_any 227 10.6 0.215 1.850 6.559 7.721 dbcsr_mm_accdrv_process_sort 4760 11.4 6.329 7.242 6.329 7.242 calculate_norms 4752 9.8 5.505 6.114 5.505 6.114 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.163 5.330 mp_sum_l 807 5.4 3.341 4.702 3.341 4.702 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.051 4.204 mp_irecv_dv 6231 10.9 2.033 4.165 2.033 4.165 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.469 3.656 make_images_sizes 222 9.7 0.000 0.000 0.713 3.270 mp_alltoall_i44 222 10.7 0.713 3.270 0.713 3.270 arnoldi_extremal 4 6.8 0.000 0.000 3.206 3.251 arnoldi_normal_ev 4 7.8 0.001 0.003 3.206 3.251 build_subspace 16 8.4 0.009 0.012 3.110 3.113 ls_scf_post 1 4.0 0.001 0.009 3.067 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.888 2.924 dbcsr_special_finalize 555 9.7 0.005 0.006 2.354 2.836 dbcsr_merge_single_wm 555 10.7 0.463 0.636 2.346 2.828 make_images_pack 222 9.7 2.206 2.618 2.208 2.620 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.304 2.546 dbcsr_sort_data 658 11.4 2.142 2.519 2.142 2.519 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.456 2.068 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.297 2.403 buffer_matrices_ensure_size 222 8.7 1.778 2.253 1.778 2.253 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.810 1.813 rebuild_ks_matrix 3 7.3 0.000 0.000 1.800 1.803 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.017 1.800 1.803 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.598000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 2.078052E+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_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.044 90.846 90.847 qs_energies 1 2.0 0.000 0.000 90.293 90.307 ls_scf 1 3.0 0.000 0.000 88.975 88.990 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.988 75.315 multiply_cannon 111 7.7 0.028 0.036 53.152 56.554 ls_scf_main 1 4.0 0.000 0.000 54.556 54.562 multiply_cannon_loop 111 8.7 0.116 0.122 49.897 52.952 density_matrix_trs4 2 5.0 0.002 0.003 48.789 48.993 ls_scf_init_scf 1 4.0 0.000 0.000 30.861 30.863 mp_waitall_1 9105 10.9 21.096 30.147 21.096 30.147 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.657 29.764 multiply_cannon_multrec 1332 9.7 13.071 16.798 22.406 27.320 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.211 27.226 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.679 20.769 make_m2s 222 7.7 0.006 0.008 15.276 15.919 make_images 222 8.7 1.583 1.965 15.246 15.890 dbcsr_mm_accdrv_process 4041 10.4 0.266 0.449 8.932 10.541 dbcsr_mm_accdrv_process_sort 4041 11.4 8.524 10.092 8.524 10.092 make_images_data 222 9.7 0.004 0.004 8.778 9.713 hybrid_alltoall_any 227 10.6 0.526 2.481 8.222 9.291 mp_sum_l 807 5.4 5.254 7.704 5.254 7.704 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.223 7.479 mp_irecv_dv 3311 11.0 3.204 7.422 3.204 7.422 calculate_norms 2376 9.8 6.004 6.895 6.004 6.895 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.049 6.178 multiply_cannon_sync_h2d 1332 9.7 4.743 5.796 4.743 5.796 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.075 5.316 arnoldi_extremal 4 6.8 0.000 0.000 4.612 4.630 arnoldi_normal_ev 4 7.8 0.001 0.005 4.612 4.630 build_subspace 16 8.4 0.014 0.021 4.359 4.361 ls_scf_post 1 4.0 0.000 0.000 3.557 3.571 ls_scf_store_result 1 5.0 0.000 0.000 3.262 3.366 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.114 3.350 dbcsr_matrix_vector_mult_local 304 10.0 2.728 3.205 2.730 3.207 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.276 2.948 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.643 2.760 make_images_pack 222 9.7 2.013 2.392 2.015 2.394 mp_allgather_i34 111 8.7 0.964 2.300 0.964 2.300 dbcsr_sort_data 436 11.2 1.828 2.087 1.828 2.087 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.886 1.888 rebuild_ks_matrix 3 7.3 0.000 0.000 1.873 1.875 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.873 1.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.847000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1706.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 2.682958E+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_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.085 94.005 94.013 qs_energies 1 2.0 0.000 0.000 93.265 93.317 ls_scf 1 3.0 0.004 0.019 91.820 91.871 dbcsr_multiply_generic 111 6.7 0.020 0.056 76.181 76.482 ls_scf_main 1 4.0 0.000 0.008 57.265 57.269 multiply_cannon 111 7.7 0.041 0.080 52.580 56.672 multiply_cannon_loop 111 8.7 0.101 0.130 49.038 52.214 density_matrix_trs4 2 5.0 0.002 0.003 51.287 51.425 mp_waitall_1 7281 11.0 23.546 32.504 23.546 32.504 ls_scf_init_scf 1 4.0 0.000 0.001 30.960 30.963 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.438 29.494 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.051 27.067 multiply_cannon_multrec 888 9.7 12.603 15.257 21.241 24.497 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.982 21.666 make_m2s 222 7.7 0.006 0.007 17.269 18.478 make_images 222 8.7 1.978 2.284 17.231 18.441 hybrid_alltoall_any 227 10.6 0.622 2.884 9.509 10.884 make_images_data 222 9.7 0.004 0.006 9.840 10.858 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.441 8.172 9.382 dbcsr_mm_accdrv_process_sort 3754 11.4 7.796 8.941 7.796 8.941 mp_sum_l 807 5.4 4.781 8.061 4.781 8.061 multiply_cannon_sync_h2d 888 9.7 6.102 7.209 6.102 7.209 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.470 6.922 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.347 6.879 mp_irecv_dv 2335 11.1 2.453 6.866 2.453 6.866 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.614 6.340 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.928 5.137 arnoldi_extremal 4 6.8 0.000 0.000 5.080 5.098 arnoldi_normal_ev 4 7.8 0.001 0.004 5.080 5.098 build_subspace 16 8.4 0.014 0.020 4.775 4.780 calculate_norms 1584 9.8 4.279 4.652 4.279 4.652 mp_allgather_i34 111 8.7 1.381 3.797 1.381 3.797 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.416 3.758 ls_scf_post 1 4.0 0.012 0.050 3.591 3.634 dbcsr_matrix_vector_mult_local 304 10.0 2.995 3.567 2.997 3.569 ls_scf_store_result 1 5.0 0.000 0.000 3.336 3.403 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.867 2.973 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.176 2.178 rebuild_ks_matrix 3 7.3 0.000 0.000 2.158 2.160 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.158 2.160 dbcsr_sort_data 325 11.1 1.908 2.141 1.908 2.141 make_images_pack 222 9.7 1.809 2.129 1.812 2.132 make_images_sizes 222 9.7 0.000 0.000 1.080 2.095 mp_alltoall_i44 222 10.7 1.080 2.095 1.080 2.095 dbcsr_data_release 9322 10.9 1.304 1.995 1.304 1.995 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.013000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2210.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 3.383226E+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_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.079 96.552 96.565 qs_energies 1 2.0 0.000 0.000 95.935 95.941 ls_scf 1 3.0 0.000 0.000 94.292 94.298 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.014 78.269 ls_scf_main 1 4.0 0.000 0.000 58.535 58.536 multiply_cannon 111 7.7 0.052 0.144 51.668 56.303 density_matrix_trs4 2 5.0 0.002 0.003 52.434 52.569 multiply_cannon_loop 111 8.7 0.115 0.127 46.578 49.084 ls_scf_init_scf 1 4.0 0.000 0.000 32.498 32.500 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.278 31.340 mp_waitall_1 6369 11.0 22.592 29.126 22.592 29.126 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.798 28.809 multiply_cannon_multrec 1332 9.7 14.213 17.236 22.171 24.576 make_m2s 222 7.7 0.006 0.008 21.002 22.444 make_images 222 8.7 3.148 3.610 20.952 22.397 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.198 16.715 make_images_data 222 9.7 0.004 0.004 11.704 13.300 hybrid_alltoall_any 227 10.6 0.799 3.836 11.066 12.708 dbcsr_mm_accdrv_process 3641 10.4 0.223 0.427 7.601 9.128 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.715 7.223 8.715 mp_sum_l 807 5.4 3.774 7.493 3.774 7.493 multiply_cannon_sync_h2d 1332 9.7 5.501 6.264 5.501 6.264 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.913 5.974 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.064 5.950 mp_irecv_dv 3229 10.9 2.041 5.883 2.041 5.883 arnoldi_extremal 4 6.8 0.000 0.000 5.222 5.236 arnoldi_normal_ev 4 7.8 0.001 0.005 5.222 5.236 build_subspace 16 8.4 0.014 0.021 4.890 4.898 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.515 4.707 mp_allgather_i34 111 8.7 2.207 4.692 2.207 4.692 calculate_norms 2376 9.8 4.186 4.529 4.186 4.529 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.521 4.467 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.573 3.906 dbcsr_matrix_vector_mult_local 304 10.0 3.171 3.713 3.173 3.715 dbcsr_sort_data 658 11.4 3.089 3.438 3.089 3.438 ls_scf_post 1 4.0 0.000 0.000 3.258 3.263 dbcsr_special_finalize 555 9.7 0.006 0.007 2.842 3.167 dbcsr_merge_single_wm 555 10.7 0.540 0.660 2.834 3.159 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.033 3.075 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.051 dbcsr_data_release 10477 10.7 1.579 2.451 1.579 2.451 dbcsr_finalize 304 7.8 0.049 0.061 1.801 2.024 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.967 1.968 rebuild_ks_matrix 3 7.3 0.000 0.000 1.944 1.945 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.944 1.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=96.565000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2697.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 4.631097E+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_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.055 99.460 99.461 qs_energies 1 2.0 0.000 0.000 98.698 98.704 ls_scf 1 3.0 0.000 0.000 96.757 96.767 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.109 78.341 ls_scf_main 1 4.0 0.000 0.000 62.260 62.262 multiply_cannon 111 7.7 0.085 0.176 55.260 60.971 density_matrix_trs4 2 5.0 0.002 0.003 54.967 55.086 multiply_cannon_loop 111 8.7 0.070 0.080 50.675 52.677 mp_waitall_1 5436 11.0 26.252 32.659 26.252 32.659 ls_scf_init_scf 1 4.0 0.000 0.000 30.870 30.875 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.480 29.533 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.304 27.312 multiply_cannon_multrec 444 9.7 14.008 16.436 21.159 24.295 make_m2s 222 7.7 0.004 0.005 17.752 20.324 make_images 222 8.7 3.728 4.420 17.690 20.264 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.960 16.198 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.175 14.466 make_images_data 222 9.7 0.003 0.004 9.917 12.524 hybrid_alltoall_any 227 10.6 0.788 3.737 9.687 12.227 multiply_cannon_sync_h2d 444 9.7 6.515 8.793 6.515 8.793 dbcsr_mm_accdrv_process 3003 10.4 0.181 0.367 6.854 7.957 dbcsr_mm_accdrv_process_sort 3003 11.4 6.532 7.614 6.532 7.614 mp_allgather_i34 111 8.7 2.818 6.964 2.818 6.964 mp_sum_l 807 5.4 3.135 6.011 3.135 6.011 arnoldi_extremal 4 6.8 0.000 0.000 5.756 5.770 arnoldi_normal_ev 4 7.8 0.001 0.004 5.756 5.770 build_subspace 16 8.4 0.015 0.020 5.374 5.383 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.612 4.788 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.179 4.545 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.679 4.494 mp_irecv_dv 1241 11.2 1.661 4.475 1.661 4.475 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 4.087 4.296 dbcsr_matrix_vector_mult_local 304 10.0 3.630 4.107 3.632 4.109 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.669 3.779 calculate_norms 792 9.8 3.550 3.681 3.550 3.681 make_images_sizes 222 9.7 0.000 0.000 1.193 3.672 mp_alltoall_i44 222 10.7 1.193 3.672 1.193 3.672 ls_scf_post 1 4.0 0.000 0.000 3.627 3.632 ls_scf_store_result 1 5.0 0.000 0.000 3.410 3.457 dbcsr_finalize 304 7.8 0.062 0.078 2.207 2.265 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.215 2.216 rebuild_ks_matrix 3 7.3 0.000 0.000 2.183 2.184 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.000 2.182 2.184 dbcsr_merge_all 275 8.9 0.472 0.521 2.051 2.098 dbcsr_data_release 10123 10.8 1.330 2.003 1.330 2.003 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.461000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3634.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b5786661874f8455978168cb43dd147fdac8cf4f_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 8.684872E+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_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.082 0.100 108.000 108.001 qs_energies 1 2.0 0.000 0.000 106.555 106.567 ls_scf 1 3.0 0.000 0.000 103.603 103.615 dbcsr_multiply_generic 111 6.7 0.023 0.027 77.152 77.274 ls_scf_main 1 4.0 0.000 0.000 65.767 65.767 density_matrix_trs4 2 5.0 0.002 0.003 56.456 56.521 multiply_cannon 111 7.7 0.121 0.160 49.690 50.910 multiply_cannon_loop 111 8.7 0.067 0.069 46.421 47.624 ls_scf_init_scf 1 4.0 0.000 0.000 34.161 34.162 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.504 32.514 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.684 29.695 mp_waitall_1 4527 11.1 22.095 25.682 22.095 25.682 make_m2s 222 7.7 0.005 0.005 23.774 24.570 make_images 222 8.7 4.578 4.977 23.668 24.462 multiply_cannon_multrec 444 9.7 17.827 18.586 22.501 23.163 hybrid_alltoall_any 227 10.6 1.656 3.614 12.855 15.513 make_images_data 222 9.7 0.003 0.003 13.033 15.252 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.470 11.170 multiply_cannon_sync_h2d 444 9.7 8.845 8.898 8.845 8.898 arnoldi_extremal 4 6.8 0.000 0.000 7.247 7.258 arnoldi_normal_ev 4 7.8 0.002 0.008 7.247 7.258 build_subspace 16 8.4 0.026 0.036 6.699 6.711 dbcsr_matrix_vector_mult 304 9.0 0.016 0.032 5.325 5.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.271 5.367 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.023 5.274 dbcsr_matrix_vector_mult_local 304 10.0 4.890 5.196 4.892 5.199 dbcsr_mm_accdrv_process 1814 10.4 0.229 0.305 4.496 4.629 dbcsr_mm_accdrv_process_sort 1814 11.4 4.197 4.334 4.197 4.334 ls_scf_post 1 4.0 0.000 0.000 3.675 3.687 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.436 make_images_sizes 222 9.7 0.000 0.000 1.427 3.426 mp_alltoall_i44 222 10.7 1.427 3.426 1.427 3.426 calculate_norms 792 9.8 3.242 3.282 3.242 3.282 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.166 dbcsr_merge_all 275 8.9 0.893 0.921 2.871 2.946 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.921 2.921 dbcsr_complete_redistribute 5 7.6 1.440 1.491 2.775 2.912 dbcsr_data_release 12724 10.6 2.330 2.857 2.330 2.857 mp_allgather_i34 111 8.7 0.922 2.835 0.922 2.835 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.584 2.586 matrix_ls_to_qs 2 6.0 0.000 0.000 2.412 2.553 rebuild_ks_matrix 3 7.3 0.000 0.000 2.518 2.519 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.518 2.519 dbcsr_sort_data 325 11.1 2.432 2.494 2.432 2.494 dbcsr_new_transposed 4 7.5 0.243 0.251 2.265 2.278 mp_sum_l 807 5.4 1.369 2.271 1.369 2.271 dbcsr_frobenius_norm 74 6.6 2.057 2.138 2.213 2.255 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.205 dbcsr_add_anytype 103 7.2 0.860 0.891 2.133 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.001000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6965.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b5786661874f8455978168cb43dd147fdac8cf4f Summary: empty Status: OK