=== This is the CP2K Performance-Test === Updating 7d91dcfa1..e6ae6671d Fast-forward src/response_solver.F | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: e6ae6671d4bc88fbd380deaa153891801fbee65d ################# 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 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/01 job id: 42056600 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/02 job id: 42056601 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/03 job id: 42056603 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/04 job id: 42056604 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/05 job id: 42056606 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/06 job id: 42056607 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/07 job id: 42056609 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/08 job id: 42056611 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/09 job id: 42056612 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/10 job id: 42056613 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/11 job id: 42056614 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/12 job id: 42056616 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/13 job id: 42056618 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/14 job id: 42056621 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/15 job id: 42056626 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/16 job id: 42056627 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/17 job id: 42056629 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/18 job id: 42056631 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/19 job id: 42056632 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/20 job id: 42056633 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/21 job id: 42056635 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/22 job id: 42056636 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/23 job id: 42056637 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/24 job id: 42056640 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/25 job id: 42056641 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/26 job id: 42056642 --- 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/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 0.022 0.036 134.516 134.516 farming_run 1 2.0 133.614 133.618 134.482 134.486 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.454285E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.044 116.505 116.505 qs_energies 1 2.0 0.000 0.000 116.035 116.037 mp2_main 1 3.0 0.000 0.000 113.972 113.974 mp2_gpw_main 1 4.0 0.019 0.022 113.061 113.063 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 94.177 94.576 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.006 55.408 55.807 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.170 41.713 47.042 get_2c_integrals 1 6.0 0.000 0.000 37.612 38.598 integrate_v_rspace 273 8.0 0.436 0.454 25.108 30.275 pw_transfer 6555 10.6 0.376 0.397 27.379 28.248 fft_wrap_pw1pw2 5465 11.4 0.046 0.049 26.039 26.773 grid_integrate_task_list 273 9.0 20.924 26.575 20.924 26.575 fft_wrap_pw1pw2_100 2178 12.4 1.193 1.408 23.602 24.334 compute_2c_integrals 1 7.0 0.003 0.003 19.773 19.774 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.917 19.465 mp2_eri_2c_integrate_gpw 1 9.0 2.383 2.428 18.914 19.462 rpa_ri_compute_en 1 5.0 0.000 0.001 18.775 18.909 cp_fm_cholesky_decompose 12 8.2 17.852 18.830 17.852 18.830 cholesky_decomp 1 7.0 0.000 0.000 16.697 17.676 fft3d_s 5443 13.4 16.147 16.674 16.169 16.695 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.566 10.855 15.566 calculate_wavefunction 272 8.0 5.407 5.550 12.526 13.133 rpa_num_int 1 6.0 0.000 0.001 10.558 10.558 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.540 10.543 calc_mat_Q 8 8.0 0.000 0.000 9.343 9.428 contract_S_to_Q 8 9.0 0.000 0.000 8.764 8.849 calc_potential_gpw 544 9.5 0.005 0.006 8.227 8.639 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.225 8.601 potential_pw2rs 545 10.0 0.107 0.109 7.722 8.448 parallel_gemm_fm 14 9.1 0.000 0.000 8.356 8.422 parallel_gemm_fm_cosma 14 10.1 8.356 8.422 8.356 8.422 collocate_single_gaussian 272 10.0 0.040 0.043 7.490 7.782 create_integ_mat 1 6.0 0.022 0.026 7.723 7.723 array2fm 1 7.0 0.000 0.000 6.625 7.104 pw_scatter_s 2720 13.7 4.422 4.616 4.422 4.616 pw_gather_s 2722 13.2 3.908 4.266 3.908 4.266 array2fm_buffer_send 1 8.0 2.925 3.158 2.925 3.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.060250, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2727.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.037 397.572 397.574 farming_run 1 2.0 396.784 396.809 397.525 397.529 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.222734E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 209.761 209.762 qs_energies 1 2.0 0.004 0.027 209.523 209.529 scf_env_do_scf 1 3.0 0.000 0.000 106.169 106.169 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.154 105.162 rebuild_ks_matrix 4 6.0 0.000 0.000 105.152 105.160 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.067 105.152 105.160 hfx_ks_matrix 4 8.0 0.001 0.001 104.755 104.758 integrate_four_center 4 9.0 0.144 0.459 104.755 104.758 mp2_main 1 3.0 0.009 0.044 103.068 103.077 mp2_gpw_main 1 4.0 0.055 0.137 101.788 101.806 integrate_four_center_main 4 10.0 0.089 0.469 96.673 99.350 integrate_four_center_bin 266 11.0 96.584 99.336 96.584 99.336 init_scf_loop 1 4.0 0.000 0.000 91.887 91.887 mp2_ri_gpw_compute_in 1 5.0 0.068 0.081 74.884 75.929 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.443 55.474 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.163 42.130 47.273 integrate_v_rspace 95 8.0 0.398 0.572 28.520 33.457 pw_transfer 2240 10.6 0.146 0.175 29.891 30.344 fft_wrap_pw1pw2 1868 11.4 0.019 0.023 28.907 29.380 grid_integrate_task_list 95 9.0 23.793 28.938 23.793 28.938 mp2_ri_gpw_compute_en 1 5.0 0.056 0.066 26.673 28.381 ao_to_mo_and_store_B_mult_1 91 7.0 10.622 27.533 10.622 27.533 fft_wrap_pw1pw2_100 730 12.4 1.262 1.516 26.638 27.134 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.895 24.995 25.006 get_2c_integrals 1 6.0 0.003 0.018 20.350 20.389 compute_2c_integrals 1 7.0 0.006 0.018 19.321 19.343 compute_2c_integrals_loop_lm 1 8.0 0.003 0.030 18.829 19.181 mp2_eri_2c_integrate_gpw 1 9.0 1.743 1.892 18.825 19.180 fft3d_s 1823 13.4 18.431 18.880 18.444 18.894 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.279 14.279 calculate_wavefunction 91 8.0 2.018 2.045 9.741 9.993 potential_pw2rs 186 10.0 0.033 0.035 8.633 9.298 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.590 8.789 9.265 local_gemm 172 8.0 8.232 8.675 8.232 8.675 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.210 8.517 mp2_ri_gpw_compute_en_comm 22 7.0 0.494 0.513 7.970 8.483 calc_potential_gpw 182 9.5 0.002 0.002 7.900 8.193 collocate_single_gaussian 91 10.0 0.017 0.024 7.869 8.120 mp_sendrecv_dm3 2068 8.0 6.017 6.532 6.017 6.532 mp2_ri_gpw_compute_en_ener 172 7.0 6.354 6.441 6.354 6.441 pw_gather_s 912 13.2 4.921 5.462 4.921 5.462 mp_sync 38 10.4 2.726 4.826 2.726 4.826 pw_scatter_s 910 13.7 3.915 4.295 3.915 4.295 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.789589, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1507.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 449.433600E+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 884531. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 49.703 49.705 qs_mol_dyn_low 1 2.0 0.003 0.003 49.417 49.426 qs_forces 11 3.9 0.002 0.002 49.357 49.357 qs_energies 11 4.9 0.001 0.002 47.911 47.920 scf_env_do_scf 11 5.9 0.000 0.001 41.677 41.677 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 39.600 39.601 dbcsr_multiply_generic 2286 12.5 0.094 0.099 31.620 32.088 qs_scf_new_mos 108 7.5 0.000 0.001 29.713 30.001 qs_scf_loop_do_ot 108 8.5 0.000 0.001 29.713 30.001 ot_scf_mini 108 9.5 0.002 0.002 28.156 28.332 multiply_cannon 2286 13.5 0.179 0.188 25.234 26.929 multiply_cannon_loop 2286 14.5 1.486 1.573 24.502 26.240 velocity_verlet 10 3.0 0.001 0.001 23.866 23.867 ot_mini 108 10.5 0.001 0.001 17.794 18.041 qs_ot_get_derivative 108 11.5 0.001 0.001 15.009 15.195 mp_waitall_1 267858 16.1 7.729 13.863 7.729 13.863 multiply_cannon_metrocomm3 54864 15.5 0.067 0.072 5.394 12.176 multiply_cannon_multrec 54864 15.5 4.319 6.751 7.755 11.310 rebuild_ks_matrix 119 8.3 0.000 0.000 7.851 7.998 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.850 7.998 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.931 7.071 multiply_cannon_sync_h2d 54864 15.5 5.990 6.896 5.990 6.896 qs_ot_get_p 119 10.4 0.001 0.001 5.962 6.226 mp_sum_l 7207 12.9 4.357 5.796 4.357 5.796 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.875 5.330 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.186 5.291 init_scf_run 11 5.9 0.000 0.001 4.993 4.993 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.993 4.993 dbcsr_mm_accdrv_process 76910 16.1 1.137 1.847 3.359 4.774 sum_up_and_integrate 119 10.3 0.012 0.015 4.516 4.524 integrate_v_rspace 119 11.3 0.002 0.002 4.503 4.513 qs_rho_update_rho_low 119 7.7 0.000 0.001 3.996 4.083 calculate_rho_elec 119 8.7 0.011 0.016 3.995 4.082 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.368 3.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.871 3.052 apply_single 119 13.6 0.000 0.000 2.871 3.052 calculate_dm_sparse 119 9.5 0.000 0.000 2.818 2.966 rs_pw_transfer 974 11.9 0.012 0.013 2.795 2.896 calculate_first_density_matrix 1 7.0 0.000 0.000 2.845 2.848 jit_kernel_multiply 13 15.8 2.162 2.837 2.162 2.837 multiply_cannon_metrocomm1 54864 15.5 0.054 0.060 1.406 2.635 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.626 2.626 ot_diis_step 108 11.5 0.006 0.006 2.586 2.587 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.431 2.434 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.396 2.397 cp_fm_redistribute_end 50 14.0 2.175 2.373 2.180 2.374 cp_fm_diag_elpa_base 50 14.0 0.193 2.330 0.194 2.341 density_rs2pw 119 9.7 0.004 0.004 2.191 2.286 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.215 2.276 acc_transpose_blocks 54864 15.5 0.220 0.244 1.754 2.248 grid_integrate_task_list 119 12.3 2.024 2.131 2.024 2.131 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.062 2.119 wfi_extrapolate 11 7.9 0.001 0.001 2.085 2.085 init_scf_loop 11 6.9 0.000 0.000 2.058 2.058 potential_pw2rs 119 12.3 0.004 0.004 1.881 1.893 make_m2s 4572 13.5 0.053 0.056 1.715 1.762 pw_transfer 1439 11.6 0.051 0.055 1.659 1.728 make_images 4572 14.5 0.133 0.138 1.633 1.679 mp_sum_d 4129 12.0 1.074 1.667 1.074 1.667 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.584 1.655 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.423 1.442 fft3d_ps 1201 14.6 0.356 0.462 1.363 1.424 mp_waitany 12084 13.8 1.244 1.374 1.244 1.374 mp_alltoall_d11v 2130 13.8 1.244 1.367 1.244 1.367 grid_collocate_task_list 119 9.7 1.283 1.366 1.283 1.366 fft_wrap_pw1pw2_140 487 13.2 0.080 0.093 1.211 1.283 parallel_gemm_fm 81 9.0 0.000 0.000 1.061 1.074 parallel_gemm_fm_cosma 81 10.0 1.061 1.074 1.061 1.074 acc_transpose_blocks_kernels 54864 16.5 0.244 0.384 0.783 1.071 prepare_preconditioner 11 7.9 0.000 0.000 0.978 1.010 make_preconditioner 11 8.9 0.000 0.000 0.978 1.010 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=49.705000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=428.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 486.551552E+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 325404. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.030 39.969 39.971 qs_mol_dyn_low 1 2.0 0.003 0.003 39.694 39.703 qs_forces 11 3.9 0.002 0.002 39.636 39.636 qs_energies 11 4.9 0.001 0.002 37.926 37.929 scf_env_do_scf 11 5.9 0.000 0.001 32.568 32.568 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.838 29.838 dbcsr_multiply_generic 2286 12.5 0.101 0.104 21.878 22.266 qs_scf_new_mos 108 7.5 0.001 0.001 20.530 20.768 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.530 20.767 ot_scf_mini 108 9.5 0.002 0.003 19.608 19.778 velocity_verlet 10 3.0 0.001 0.001 19.103 19.104 multiply_cannon 2286 13.5 0.207 0.216 16.500 18.277 multiply_cannon_loop 2286 14.5 0.909 0.977 15.307 16.952 ot_mini 108 10.5 0.001 0.001 12.184 12.420 mp_waitall_1 217478 16.2 6.499 11.567 6.499 11.567 qs_ot_get_derivative 108 11.5 0.001 0.001 9.670 9.844 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.250 9.595 multiply_cannon_multrec 27432 15.5 1.944 4.256 5.880 8.602 rebuild_ks_matrix 119 8.3 0.000 0.000 7.363 7.511 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.363 7.511 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.514 6.649 dbcsr_mm_accdrv_process 47894 16.0 2.963 5.206 3.868 5.685 qs_ot_get_p 119 10.4 0.001 0.001 4.457 4.696 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.734 4.581 sum_up_and_integrate 119 10.3 0.024 0.026 4.361 4.367 integrate_v_rspace 119 11.3 0.002 0.002 4.336 4.344 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.086 4.163 apply_single 119 13.6 0.000 0.000 3.085 4.163 init_scf_run 11 5.9 0.000 0.001 4.141 4.141 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.141 4.141 mp_sum_l 7207 12.9 2.193 4.108 2.193 4.108 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.793 3.829 calculate_rho_elec 119 8.7 0.021 0.024 3.792 3.829 make_m2s 4572 13.5 0.052 0.054 2.739 3.002 make_images 4572 14.5 0.200 0.235 2.650 2.912 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.890 2.912 rs_pw_transfer 974 11.9 0.010 0.011 2.811 2.889 multiply_cannon_sync_h2d 27432 15.5 2.163 2.813 2.163 2.813 init_scf_loop 11 6.9 0.000 0.000 2.705 2.705 calculate_first_density_matrix 1 7.0 0.000 0.000 2.537 2.539 ot_diis_step 108 11.5 0.011 0.011 2.454 2.455 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.441 2.441 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.253 2.344 density_rs2pw 119 9.7 0.004 0.004 2.147 2.208 calculate_dm_sparse 119 9.5 0.000 0.000 2.122 2.196 jit_kernel_multiply 11 16.0 0.853 2.166 0.853 2.166 potential_pw2rs 119 12.3 0.006 0.007 2.086 2.098 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.054 2.054 cp_fm_redistribute_end 50 14.0 1.689 2.021 1.693 2.023 cp_fm_diag_elpa_base 50 14.0 0.320 1.951 0.327 1.981 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.968 1.970 grid_integrate_task_list 119 12.3 1.838 1.927 1.838 1.927 pw_transfer 1439 11.6 0.063 0.066 1.878 1.912 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.788 1.824 prepare_preconditioner 11 7.9 0.000 0.000 1.755 1.783 make_preconditioner 11 8.9 0.000 0.000 1.755 1.783 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.709 1.750 make_images_data 4572 15.5 0.045 0.051 1.249 1.727 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.652 1.711 hybrid_alltoall_any 4725 16.4 0.050 0.112 1.097 1.623 mp_alltoall_d11v 2130 13.8 1.302 1.564 1.302 1.564 wfi_extrapolate 11 7.9 0.001 0.001 1.549 1.549 fft3d_ps 1201 14.6 0.499 0.551 1.495 1.529 acc_transpose_blocks 27432 15.5 0.108 0.112 1.219 1.516 mp_allgather_i34 2286 14.5 0.637 1.442 0.637 1.442 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.328 1.362 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.315 1.322 grid_collocate_task_list 119 9.7 1.241 1.317 1.241 1.317 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.220 1.262 mp_sum_d 4129 12.0 0.619 1.069 0.619 1.069 make_images_sizes 4572 15.5 0.005 0.005 0.671 0.950 mp_alltoall_i44 4572 16.5 0.666 0.945 0.666 0.945 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.936 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.907 0.920 acc_transpose_blocks_kernels 27432 16.5 0.182 0.268 0.673 0.885 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.608 0.816 0.867 mp_alltoall_z22v 1201 16.6 0.744 0.820 0.744 0.820 cp_fm_cholesky_invert 11 10.9 0.805 0.810 0.805 0.810 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.971000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.272727, yerr=2.048987 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 522.088448E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.035 33.652 33.653 qs_mol_dyn_low 1 2.0 0.003 0.003 33.318 33.326 qs_forces 11 3.9 0.002 0.002 33.259 33.259 qs_energies 11 4.9 0.001 0.002 31.681 31.684 scf_env_do_scf 11 5.9 0.000 0.001 26.773 26.773 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.014 24.014 dbcsr_multiply_generic 2286 12.5 0.095 0.096 16.980 17.111 velocity_verlet 10 3.0 0.001 0.001 15.890 15.892 qs_scf_new_mos 108 7.5 0.001 0.001 15.705 15.725 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.705 15.725 ot_scf_mini 108 9.5 0.002 0.002 14.948 14.959 multiply_cannon 2286 13.5 0.194 0.199 13.264 14.053 multiply_cannon_loop 2286 14.5 0.641 0.670 12.389 13.247 ot_mini 108 10.5 0.001 0.001 9.181 9.192 qs_ot_get_derivative 108 11.5 0.001 0.001 7.624 7.635 multiply_cannon_multrec 18288 15.5 1.920 2.896 6.823 7.105 rebuild_ks_matrix 119 8.3 0.000 0.000 6.519 6.535 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.519 6.534 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.774 5.789 dbcsr_mm_accdrv_process 38222 16.0 4.106 5.276 4.820 5.749 mp_waitall_1 169478 16.3 3.427 4.343 3.427 4.343 sum_up_and_integrate 119 10.3 0.030 0.031 4.142 4.147 integrate_v_rspace 119 11.3 0.002 0.003 4.111 4.121 init_scf_run 11 5.9 0.000 0.001 3.727 3.727 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.727 3.727 qs_ot_get_p 119 10.4 0.001 0.001 3.523 3.542 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.430 3.443 calculate_rho_elec 119 8.7 0.030 0.031 3.429 3.442 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.803 3.425 init_scf_loop 11 6.9 0.000 0.000 2.738 2.738 rs_pw_transfer 974 11.9 0.009 0.010 2.590 2.685 calculate_first_density_matrix 1 7.0 0.000 0.000 2.444 2.445 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.071 2.395 apply_single 119 13.6 0.000 0.000 2.071 2.395 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.541 2.284 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.264 2.273 make_m2s 4572 13.5 0.045 0.046 2.031 2.193 density_rs2pw 119 9.7 0.004 0.004 2.017 2.127 make_images 4572 14.5 0.190 0.202 1.945 2.107 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.976 1.977 calculate_dm_sparse 119 9.5 0.000 0.000 1.949 1.961 potential_pw2rs 119 12.3 0.007 0.008 1.926 1.936 jit_kernel_multiply 10 16.3 0.662 1.931 0.662 1.931 prepare_preconditioner 11 7.9 0.000 0.000 1.908 1.909 make_preconditioner 11 8.9 0.000 0.000 1.907 1.909 grid_integrate_task_list 119 12.3 1.792 1.869 1.792 1.869 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.759 1.842 pw_transfer 1439 11.6 0.063 0.066 1.784 1.796 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.742 1.747 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.715 1.716 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.693 1.708 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.674 1.676 mp_sum_l 7207 12.9 1.299 1.657 1.299 1.657 cp_fm_redistribute_end 50 14.0 1.239 1.645 1.241 1.645 cp_fm_diag_elpa_base 50 14.0 0.386 1.569 0.401 1.608 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.527 1.534 ot_diis_step 108 11.5 0.011 0.011 1.524 1.524 multiply_cannon_sync_h2d 18288 15.5 1.324 1.481 1.324 1.481 fft3d_ps 1201 14.6 0.502 0.518 1.381 1.398 grid_collocate_task_list 119 9.7 1.205 1.310 1.205 1.310 acc_transpose_blocks 18288 15.5 0.075 0.076 1.281 1.299 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.292 1.297 fft_wrap_pw1pw2_140 487 13.2 0.085 0.091 1.253 1.266 wfi_extrapolate 11 7.9 0.001 0.001 1.232 1.232 make_images_data 4572 15.5 0.045 0.049 0.905 1.088 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.377 1.010 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.953 0.978 hybrid_alltoall_any 4725 16.4 0.054 0.115 0.781 0.967 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.938 0.938 mp_alltoall_d11v 2130 13.8 0.777 0.927 0.777 0.927 acc_transpose_blocks_kernels 18288 16.5 0.212 0.219 0.829 0.840 cp_fm_cholesky_invert 11 10.9 0.822 0.827 0.822 0.827 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.817 0.819 mp_alltoall_z22v 1201 16.6 0.698 0.771 0.698 0.771 mp_allgather_i34 2286 14.5 0.337 0.742 0.337 0.742 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.720 dbcsr_complete_redistribute 329 12.2 0.095 0.116 0.579 0.685 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=33.653000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.090909, yerr=1.781447 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.877504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.045 37.495 37.497 qs_mol_dyn_low 1 2.0 0.003 0.003 37.214 37.222 qs_forces 11 3.9 0.002 0.003 37.096 37.097 qs_energies 11 4.9 0.001 0.004 35.322 35.327 scf_env_do_scf 11 5.9 0.001 0.001 29.954 29.956 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.920 25.921 velocity_verlet 10 3.0 0.001 0.001 19.372 19.376 dbcsr_multiply_generic 2286 12.5 0.101 0.105 18.713 18.831 qs_scf_new_mos 108 7.5 0.001 0.001 16.993 17.050 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.993 17.049 ot_scf_mini 108 9.5 0.002 0.003 16.016 16.076 multiply_cannon 2286 13.5 0.224 0.265 14.614 15.081 multiply_cannon_loop 2286 14.5 0.941 0.971 13.496 13.880 ot_mini 108 10.5 0.001 0.001 9.715 9.782 multiply_cannon_multrec 27432 15.5 2.329 3.003 8.583 8.975 qs_ot_get_derivative 108 11.5 0.001 0.001 7.813 7.872 dbcsr_mm_accdrv_process 47916 15.9 5.385 6.949 6.160 7.342 rebuild_ks_matrix 119 8.3 0.000 0.000 6.983 7.038 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.983 7.038 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.224 6.273 sum_up_and_integrate 119 10.3 0.035 0.038 4.195 4.202 integrate_v_rspace 119 11.3 0.002 0.003 4.160 4.168 init_scf_loop 11 6.9 0.001 0.003 4.006 4.007 init_scf_run 11 5.9 0.000 0.001 3.941 3.942 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.941 3.941 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.739 3.773 calculate_rho_elec 119 8.7 0.040 0.046 3.739 3.772 mp_waitall_1 145218 16.4 2.993 3.701 2.993 3.701 qs_ot_get_p 119 10.4 0.001 0.001 3.590 3.661 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.943 3.344 prepare_preconditioner 11 7.9 0.000 0.000 3.062 3.071 make_preconditioner 11 8.9 0.000 0.001 3.062 3.071 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.637 2.974 make_m2s 4572 13.5 0.054 0.056 2.584 2.798 rs_pw_transfer 974 11.9 0.009 0.009 2.683 2.794 make_images 4572 14.5 0.273 0.336 2.476 2.689 calculate_first_density_matrix 1 7.0 0.000 0.001 2.393 2.394 density_rs2pw 119 9.7 0.003 0.004 2.207 2.334 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.110 2.239 apply_single 119 13.6 0.000 0.000 2.110 2.238 calculate_dm_sparse 119 9.5 0.000 0.000 2.081 2.127 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.108 2.118 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.061 2.090 pw_transfer 1439 11.6 0.063 0.066 1.976 2.024 potential_pw2rs 119 12.3 0.009 0.009 1.951 1.971 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.886 1.938 grid_integrate_task_list 119 12.3 1.838 1.911 1.838 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.888 1.889 ot_diis_step 108 11.5 0.012 0.012 1.855 1.855 jit_kernel_multiply 10 16.0 0.714 1.826 0.714 1.826 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.759 1.760 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.693 fft3d_ps 1201 14.6 0.534 0.584 1.569 1.614 fft_wrap_pw1pw2_140 487 13.2 0.083 0.091 1.465 1.517 acc_transpose_blocks 27432 15.5 0.112 0.114 1.482 1.514 wfi_extrapolate 11 7.9 0.001 0.001 1.486 1.486 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.789 1.484 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.404 1.407 cp_fm_redistribute_end 50 14.0 0.917 1.362 0.920 1.365 mp_sum_l 7207 12.9 1.096 1.359 1.096 1.359 cp_fm_diag_elpa_base 50 14.0 0.420 1.283 0.440 1.322 make_images_data 4572 15.5 0.045 0.049 1.061 1.313 grid_collocate_task_list 119 9.7 1.221 1.288 1.221 1.288 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.250 dbcsr_complete_redistribute 329 12.2 0.116 0.143 0.963 1.247 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.902 1.200 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.169 1.184 cp_fm_upper_to_full 72 13.5 0.829 1.154 0.829 1.154 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.111 1.111 mp_alltoall_d11v 2130 13.8 0.919 1.077 0.919 1.077 cp_fm_cholesky_invert 11 10.9 1.060 1.064 1.060 1.064 multiply_cannon_sync_h2d 27432 15.5 0.972 1.022 0.972 1.022 mp_alltoall_z22v 1201 16.6 0.917 0.958 0.917 0.958 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.671 0.949 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.796 0.873 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.839 0.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.851 0.858 mp_alltoall_i22 627 13.8 0.465 0.770 0.465 0.770 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.497000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.454545, yerr=3.677045 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 598.241280E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.036 31.515 31.516 qs_mol_dyn_low 1 2.0 0.003 0.003 31.274 31.282 qs_forces 11 3.9 0.002 0.002 31.214 31.215 qs_energies 11 4.9 0.001 0.001 29.404 29.407 scf_env_do_scf 11 5.9 0.000 0.001 23.917 23.918 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.905 20.906 velocity_verlet 10 3.0 0.001 0.001 15.923 15.926 dbcsr_multiply_generic 2286 12.5 0.093 0.095 13.608 13.724 qs_scf_new_mos 108 7.5 0.001 0.001 12.512 12.551 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.511 12.550 ot_scf_mini 108 9.5 0.002 0.002 11.776 11.810 multiply_cannon 2286 13.5 0.233 0.241 10.544 11.139 multiply_cannon_loop 2286 14.5 0.333 0.344 9.498 9.789 multiply_cannon_multrec 9144 15.5 1.784 2.016 6.217 6.565 ot_mini 108 10.5 0.001 0.001 6.475 6.521 rebuild_ks_matrix 119 8.3 0.000 0.000 6.322 6.361 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.322 6.361 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.618 5.655 qs_ot_get_derivative 108 11.5 0.001 0.001 5.083 5.119 dbcsr_mm_accdrv_process 12550 15.8 2.959 3.482 4.332 4.447 init_scf_run 11 5.9 0.000 0.001 4.001 4.001 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.001 4.001 sum_up_and_integrate 119 10.3 0.037 0.041 3.980 3.985 integrate_v_rspace 119 11.3 0.003 0.003 3.943 3.948 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.691 3.736 calculate_rho_elec 119 8.7 0.059 0.061 3.690 3.735 mp_waitall_1 121218 16.5 2.751 3.501 2.751 3.501 qs_ot_get_p 119 10.4 0.001 0.001 3.378 3.429 init_scf_loop 11 6.9 0.000 0.000 2.983 2.985 calculate_first_density_matrix 1 7.0 0.000 0.000 2.631 2.638 make_m2s 4572 13.5 0.035 0.035 2.117 2.382 make_images 4572 14.5 0.268 0.303 2.027 2.290 rs_pw_transfer 974 11.9 0.008 0.008 2.169 2.237 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.210 2.214 prepare_preconditioner 11 7.9 0.000 0.000 2.193 2.200 make_preconditioner 11 8.9 0.000 0.000 2.193 2.200 jit_kernel_multiply 10 15.6 1.335 2.170 1.335 2.170 density_rs2pw 119 9.7 0.004 0.004 2.061 2.116 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.062 2.098 pw_transfer 1439 11.6 0.063 0.066 2.020 2.038 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.987 1.989 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.985 1.986 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.929 1.946 grid_integrate_task_list 119 12.3 1.859 1.935 1.859 1.935 calculate_dm_sparse 119 9.5 0.000 0.000 1.903 1.931 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.760 1.776 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.737 1.754 potential_pw2rs 119 12.3 0.010 0.011 1.661 1.677 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.654 1.656 cp_fm_redistribute_end 50 14.0 0.827 1.616 0.829 1.618 fft3d_ps 1201 14.6 0.543 0.556 1.597 1.610 cp_fm_diag_elpa_base 50 14.0 0.740 1.524 0.785 1.581 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.512 1.534 grid_collocate_task_list 119 9.7 1.284 1.388 1.284 1.388 make_images_data 4572 15.5 0.038 0.042 1.036 1.380 ot_diis_step 108 11.5 0.012 0.013 1.373 1.374 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.345 1.365 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.957 1.317 wfi_extrapolate 11 7.9 0.001 0.001 1.302 1.302 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.228 1.260 apply_single 119 13.6 0.000 0.000 1.227 1.260 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.237 1.238 cp_fm_cholesky_invert 11 10.9 1.168 1.172 1.168 1.172 mp_alltoall_d11v 2130 13.8 1.017 1.153 1.017 1.153 mp_alltoall_z22v 1201 16.6 0.927 0.969 0.927 0.969 acc_transpose_blocks 9144 15.5 0.039 0.040 0.960 0.968 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.438 0.947 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.923 mp_allgather_i34 2286 14.5 0.377 0.895 0.377 0.895 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.833 0.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.830 0.832 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.580 0.800 multiply_cannon_sync_h2d 9144 15.5 0.704 0.774 0.704 0.774 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.689 0.738 dbcsr_complete_redistribute 329 12.2 0.160 0.168 0.707 0.737 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.712 0.717 mp_sum_l 7207 12.9 0.519 0.654 0.519 0.654 make_images_sizes 4572 15.5 0.005 0.005 0.420 0.645 mp_alltoall_i44 4572 16.5 0.415 0.640 0.415 0.640 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.516000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.363636, yerr=5.547690 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 751.730688E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.036 44.016 44.017 qs_mol_dyn_low 1 2.0 0.003 0.003 43.765 43.796 qs_forces 11 3.9 0.002 0.002 43.699 43.700 qs_energies 11 4.9 0.001 0.001 41.717 41.721 scf_env_do_scf 11 5.9 0.000 0.001 35.699 35.699 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.915 26.916 velocity_verlet 10 3.0 0.001 0.001 24.746 24.752 dbcsr_multiply_generic 2286 12.5 0.101 0.102 18.362 18.654 qs_scf_new_mos 108 7.5 0.001 0.001 17.460 17.564 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.459 17.564 ot_scf_mini 108 9.5 0.002 0.002 16.326 16.424 multiply_cannon 2286 13.5 0.304 0.313 13.838 14.964 multiply_cannon_loop 2286 14.5 0.344 0.350 12.494 13.471 ot_mini 108 10.5 0.001 0.001 9.470 9.597 init_scf_loop 11 6.9 0.000 0.000 8.752 8.755 multiply_cannon_multrec 9144 15.5 3.346 4.666 8.584 8.709 prepare_preconditioner 11 7.9 0.000 0.000 7.719 7.733 make_preconditioner 11 8.9 0.000 0.000 7.719 7.733 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.266 7.591 qs_ot_get_derivative 108 11.5 0.001 0.001 7.342 7.439 rebuild_ks_matrix 119 8.3 0.000 0.000 7.210 7.364 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.210 7.364 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.537 6.675 dbcsr_mm_accdrv_process 12550 15.8 4.299 5.711 5.115 6.451 cp_fm_upper_to_full 72 14.2 3.278 4.653 3.278 4.653 mp_waitall_1 97218 16.6 3.240 4.141 3.240 4.141 qs_ot_get_p 119 10.4 0.001 0.001 3.983 4.134 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.989 4.009 calculate_rho_elec 119 8.7 0.118 0.121 3.988 4.009 sum_up_and_integrate 119 10.3 0.064 0.065 3.959 3.965 integrate_v_rspace 119 11.3 0.003 0.003 3.894 3.902 init_scf_run 11 5.9 0.000 0.001 3.830 3.830 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.829 3.829 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.640 3.105 dbcsr_complete_redistribute 329 12.2 0.284 0.288 2.103 2.930 make_m2s 4572 13.5 0.038 0.038 2.680 2.900 make_images 4572 14.5 0.352 0.380 2.560 2.780 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.721 2.540 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.295 2.531 apply_single 119 13.6 0.000 0.000 2.295 2.530 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.478 2.480 calculate_dm_sparse 119 9.5 0.000 0.000 2.342 2.358 pw_transfer 1439 11.6 0.066 0.068 2.300 2.309 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.444 2.280 mp_alltoall_i22 627 13.8 1.426 2.280 1.426 2.280 calculate_first_density_matrix 1 7.0 0.000 0.000 2.272 2.278 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.448 2.257 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.204 2.213 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.177 2.178 mp_sum_l 7207 12.9 1.316 2.143 1.316 2.143 ot_diis_step 108 11.5 0.014 0.014 2.086 2.087 cp_fm_cholesky_invert 11 10.9 2.064 2.068 2.064 2.068 density_rs2pw 119 9.7 0.003 0.003 2.038 2.060 grid_integrate_task_list 119 12.3 2.024 2.049 2.024 2.049 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.857 1.909 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.905 1.906 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.855 1.856 fft3d_ps 1201 14.6 0.568 0.580 1.839 1.847 fft_wrap_pw1pw2_140 487 13.2 0.087 0.088 1.804 1.813 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.806 1.807 cp_fm_diag_elpa_base 50 14.0 1.651 1.705 1.803 1.805 rs_pw_transfer 974 11.9 0.009 0.009 1.669 1.711 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.368 1.690 make_images_data 4572 15.5 0.042 0.045 1.363 1.681 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.561 1.603 wfi_extrapolate 11 7.9 0.001 0.001 1.488 1.488 grid_collocate_task_list 119 9.7 1.463 1.477 1.463 1.477 potential_pw2rs 119 12.3 0.014 0.014 1.443 1.451 mp_alltoall_d11v 2130 13.8 1.339 1.389 1.339 1.389 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.362 1.374 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.230 1.268 qs_env_update_s_mstruct 11 6.9 0.059 0.129 1.173 1.252 mp_alltoall_z22v 1201 16.6 1.135 1.161 1.135 1.161 multiply_cannon_sync_h2d 9144 15.5 1.036 1.039 1.036 1.039 jit_kernel_multiply 6 15.5 0.789 1.033 0.789 1.033 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.989 1.020 qs_create_task_list 11 7.9 0.001 0.001 0.956 0.966 generate_qs_task_list 11 8.9 0.371 0.390 0.956 0.965 acc_transpose_blocks 9144 15.5 0.039 0.039 0.906 0.908 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.017000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=701.454545, yerr=14.002361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 499.232768E+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 1174467. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.101 84.542 84.545 qs_mol_dyn_low 1 2.0 0.003 0.003 84.103 84.115 qs_forces 11 3.9 0.002 0.002 84.023 84.024 qs_energies 11 4.9 0.001 0.001 81.161 81.170 scf_env_do_scf 11 5.9 0.000 0.001 72.023 72.026 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 66.165 66.165 dbcsr_multiply_generic 2055 12.4 0.105 0.108 52.697 52.881 qs_scf_new_mos 99 7.5 0.000 0.001 48.722 48.832 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.721 48.831 ot_scf_mini 99 9.5 0.002 0.002 46.310 46.391 multiply_cannon 2055 13.4 0.174 0.179 42.917 43.826 velocity_verlet 10 3.0 0.001 0.001 42.961 42.962 multiply_cannon_loop 2055 14.4 1.570 1.606 41.863 42.716 ot_mini 99 10.5 0.001 0.001 28.283 28.369 qs_ot_get_derivative 99 11.5 0.001 0.001 21.457 21.555 multiply_cannon_multrec 49320 15.4 12.062 12.786 17.227 17.984 rebuild_ks_matrix 110 8.3 0.000 0.001 14.745 14.890 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.744 14.890 mp_waitall_1 241148 16.1 12.331 13.219 12.331 13.219 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.952 13.071 multiply_cannon_sync_h2d 49320 15.4 9.884 10.502 9.884 10.502 qs_ot_get_p 110 10.4 0.001 0.001 9.510 9.635 multiply_cannon_metrocomm3 49320 15.4 0.076 0.080 7.063 8.221 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.306 7.849 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.185 7.657 apply_single 110 13.6 0.000 0.000 7.184 7.657 sum_up_and_integrate 110 10.3 0.036 0.043 7.262 7.275 integrate_v_rspace 110 11.3 0.003 0.003 7.226 7.247 init_scf_run 11 5.9 0.000 0.001 7.108 7.108 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.108 7.108 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.878 6.938 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.519 6.665 calculate_rho_elec 110 8.6 0.020 0.024 6.519 6.664 ot_diis_step 99 11.5 0.006 0.006 6.562 6.562 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.130 6.158 mp_sum_l 6514 12.8 5.232 5.989 5.232 5.989 init_scf_loop 11 6.9 0.000 0.000 5.826 5.827 dbcsr_mm_accdrv_process 87628 16.1 2.121 2.212 5.044 5.339 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.117 5.117 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.592 4.593 cp_fm_redistribute_end 48 14.0 3.993 4.557 3.997 4.559 cp_fm_diag_elpa_base 48 14.0 0.555 4.451 0.559 4.481 make_m2s 4110 13.4 0.061 0.066 4.100 4.220 multiply_cannon_metrocomm1 49320 15.4 0.062 0.066 2.810 4.207 rs_pw_transfer 902 11.9 0.012 0.013 4.037 4.200 wfi_extrapolate 11 7.9 0.001 0.001 4.162 4.162 make_images 4110 14.4 0.178 0.192 4.005 4.126 calculate_dm_sparse 110 9.5 0.000 0.001 3.776 3.858 density_rs2pw 110 9.6 0.004 0.004 3.531 3.678 prepare_preconditioner 11 7.9 0.000 0.000 3.628 3.647 make_preconditioner 11 8.9 0.000 0.000 3.628 3.646 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.619 3.624 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.406 3.451 grid_integrate_task_list 110 12.3 3.232 3.374 3.232 3.374 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.246 3.295 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.231 3.272 pw_transfer 1331 11.6 0.053 0.064 3.121 3.207 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.034 3.124 potential_pw2rs 110 12.3 0.006 0.007 2.825 2.859 calculate_first_density_matrix 1 7.0 0.000 0.000 2.847 2.858 mp_alltoall_d11v 2046 13.8 2.287 2.733 2.287 2.733 jit_kernel_multiply 13 15.9 2.639 2.669 2.639 2.669 fft3d_ps 1111 14.6 0.748 0.835 2.573 2.652 fft_wrap_pw1pw2_140 451 13.1 0.167 0.187 2.506 2.598 acc_transpose_blocks 49320 15.4 0.217 0.227 2.225 2.310 grid_collocate_task_list 110 9.6 2.084 2.202 2.084 2.202 mp_waitany 14300 13.8 1.833 2.162 1.833 2.162 make_images_data 4110 15.4 0.043 0.046 1.828 1.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.957 cp_fm_cholesky_invert 11 10.9 1.946 1.950 1.946 1.950 mp_sum_d 3883 11.9 1.450 1.890 1.450 1.890 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.583 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.545000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=474.818182, yerr=2.124240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 586.792960E+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 1377358. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 70.492 70.494 qs_mol_dyn_low 1 2.0 0.003 0.003 70.281 70.290 qs_forces 11 3.9 0.002 0.002 68.987 68.988 qs_energies 11 4.9 0.001 0.001 65.649 65.654 scf_env_do_scf 11 5.9 0.000 0.001 57.146 57.149 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.131 49.132 dbcsr_multiply_generic 2055 12.4 0.119 0.124 38.070 38.321 velocity_verlet 10 3.0 0.001 0.001 37.842 37.868 qs_scf_new_mos 99 7.5 0.001 0.001 33.086 33.225 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.086 33.224 multiply_cannon 2055 13.4 0.221 0.246 31.106 32.221 ot_scf_mini 99 9.5 0.003 0.003 31.415 31.542 multiply_cannon_loop 2055 14.4 0.934 0.954 29.749 30.662 ot_mini 99 10.5 0.001 0.001 18.630 18.764 multiply_cannon_multrec 24660 15.4 7.578 9.618 13.800 15.661 rebuild_ks_matrix 110 8.3 0.000 0.000 13.759 13.876 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.759 13.876 qs_ot_get_derivative 99 11.5 0.001 0.001 12.774 12.904 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.127 12.223 mp_waitall_1 186928 16.3 8.104 10.576 8.104 10.576 multiply_cannon_sync_h2d 24660 15.4 7.009 8.297 7.009 8.297 init_scf_loop 11 6.9 0.000 0.000 7.978 7.978 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.259 7.686 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.710 7.328 apply_single 110 13.6 0.000 0.001 6.709 7.328 sum_up_and_integrate 110 10.3 0.052 0.060 6.660 6.672 integrate_v_rspace 110 11.3 0.002 0.003 6.608 6.623 dbcsr_mm_accdrv_process 52282 16.1 4.751 5.531 6.064 6.502 qs_ot_get_p 110 10.4 0.001 0.001 6.114 6.302 init_scf_run 11 5.9 0.000 0.001 6.122 6.122 scf_env_initial_rho_setup 11 6.9 0.007 0.008 6.121 6.122 prepare_preconditioner 11 7.9 0.000 0.000 5.947 5.970 make_preconditioner 11 8.9 0.000 0.000 5.947 5.970 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.862 5.873 calculate_rho_elec 110 8.6 0.039 0.047 5.862 5.873 ot_diis_step 99 11.5 0.010 0.010 5.803 5.803 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.515 5.679 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.742 5.607 make_m2s 4110 13.4 0.057 0.060 4.318 4.788 make_images 4110 14.4 0.396 0.441 4.210 4.677 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.159 4.183 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.714 3.715 wfi_extrapolate 11 7.9 0.001 0.001 3.550 3.551 pw_transfer 1331 11.6 0.065 0.069 3.398 3.541 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.292 3.430 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.288 3.360 density_rs2pw 110 9.6 0.004 0.004 3.142 3.320 grid_integrate_task_list 110 12.3 3.145 3.319 3.145 3.319 rs_pw_transfer 902 11.9 0.012 0.014 3.062 3.275 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.262 3.264 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.134 3.136 cp_fm_redistribute_end 48 14.0 2.335 3.104 2.337 3.104 cp_fm_diag_elpa_base 48 14.0 0.733 2.979 0.764 3.060 calculate_dm_sparse 110 9.5 0.001 0.001 2.969 3.002 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.913 2.960 mp_sum_l 6514 12.8 2.015 2.889 2.015 2.889 make_images_data 4110 15.4 0.047 0.052 2.411 2.875 cp_fm_cholesky_invert 11 10.9 2.851 2.860 2.851 2.860 fft_wrap_pw1pw2_140 451 13.1 0.200 0.217 2.720 2.852 hybrid_alltoall_any 4261 16.3 0.101 0.441 2.135 2.836 fft3d_ps 1111 14.6 1.072 1.272 2.649 2.777 potential_pw2rs 110 12.3 0.008 0.009 2.505 2.524 calculate_first_density_matrix 1 7.0 0.000 0.001 2.478 2.481 jit_kernel_multiply 12 16.4 0.960 2.286 0.960 2.286 grid_collocate_task_list 110 9.6 2.119 2.266 2.119 2.266 mp_alltoall_d11v 2046 13.8 1.858 2.116 1.858 2.116 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.992 2.012 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.862 1.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.796 1.814 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.787 1.778 mp_allgather_i34 2055 14.4 0.685 1.766 0.685 1.766 mp_irecv_dv 57340 16.2 0.661 1.562 0.661 1.562 acc_transpose_blocks 24660 15.4 0.107 0.109 1.543 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.532 1.545 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.375 1.485 dbcsr_complete_redistribute 325 12.2 0.242 0.299 1.208 1.481 cp_fm_cholesky_decompose 22 10.9 1.442 1.448 1.442 1.448 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=70.494000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.000000, yerr=7.236272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 657.960960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.123 61.070 61.071 qs_mol_dyn_low 1 2.0 0.003 0.003 60.620 60.630 qs_forces 11 3.9 0.002 0.004 60.554 60.554 qs_energies 11 4.9 0.002 0.006 57.355 57.360 scf_env_do_scf 11 5.9 0.001 0.001 49.313 49.313 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.249 40.250 velocity_verlet 10 3.0 0.001 0.001 33.186 33.188 dbcsr_multiply_generic 2055 12.4 0.109 0.112 29.011 29.305 qs_scf_new_mos 99 7.5 0.001 0.001 25.547 25.653 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.546 25.652 ot_scf_mini 99 9.5 0.002 0.003 24.316 24.431 multiply_cannon 2055 13.4 0.209 0.221 22.307 23.474 multiply_cannon_loop 2055 14.4 0.623 0.634 21.058 22.375 ot_mini 99 10.5 0.001 0.001 14.017 14.137 rebuild_ks_matrix 110 8.3 0.000 0.000 12.311 12.466 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.311 12.465 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.861 11.003 mp_waitall_1 146766 16.3 7.695 10.819 7.695 10.819 multiply_cannon_multrec 16440 15.4 3.810 4.953 9.570 10.599 qs_ot_get_derivative 99 11.5 0.001 0.001 9.447 9.565 init_scf_loop 11 6.9 0.001 0.004 9.026 9.027 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.576 7.608 prepare_preconditioner 11 7.9 0.000 0.000 7.223 7.244 make_preconditioner 11 8.9 0.000 0.002 7.223 7.244 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.546 6.902 sum_up_and_integrate 110 10.3 0.060 0.061 6.509 6.523 integrate_v_rspace 110 11.3 0.003 0.003 6.449 6.462 dbcsr_mm_accdrv_process 34862 16.1 4.724 5.328 5.616 5.745 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.644 5.655 calculate_rho_elec 110 8.6 0.058 0.059 5.643 5.655 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.145 5.628 apply_single 110 13.6 0.000 0.000 5.145 5.628 init_scf_run 11 5.9 0.000 0.001 5.623 5.624 scf_env_initial_rho_setup 11 6.9 0.002 0.012 5.623 5.623 qs_ot_get_p 110 10.4 0.001 0.001 5.400 5.566 make_m2s 4110 13.4 0.050 0.051 4.398 4.752 make_images 4110 14.4 0.391 0.515 4.283 4.636 ot_diis_step 99 11.5 0.011 0.011 4.535 4.535 multiply_cannon_sync_h2d 16440 15.4 3.633 4.183 3.633 4.183 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.175 3.887 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.695 3.702 grid_integrate_task_list 110 12.3 3.175 3.394 3.175 3.394 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.335 3.335 pw_transfer 1331 11.6 0.065 0.071 3.249 3.256 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.143 3.152 density_rs2pw 110 9.6 0.004 0.004 2.910 3.147 rs_pw_transfer 902 11.9 0.011 0.011 2.807 3.049 make_images_data 4110 15.4 0.044 0.047 2.567 3.033 wfi_extrapolate 11 7.9 0.001 0.001 3.003 3.003 cp_fm_cholesky_invert 11 10.9 2.994 3.001 2.994 3.001 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.265 2.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.828 2.830 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.764 2.766 cp_fm_redistribute_end 48 14.0 1.720 2.737 1.722 2.737 cp_fm_diag_elpa_base 48 14.0 0.952 2.590 1.010 2.697 fft_wrap_pw1pw2_140 451 13.1 0.210 0.214 2.653 2.663 calculate_dm_sparse 110 9.5 0.001 0.001 2.524 2.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.485 2.538 calculate_first_density_matrix 1 7.0 0.000 0.003 2.534 2.537 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.898 2.528 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.446 2.504 fft3d_ps 1111 14.6 1.061 1.069 2.458 2.467 mp_sum_l 6514 12.8 1.712 2.403 1.712 2.403 mp_irecv_dv 48980 15.7 0.827 2.401 0.827 2.401 grid_collocate_task_list 110 9.6 2.173 2.330 2.173 2.330 potential_pw2rs 110 12.3 0.011 0.011 2.315 2.324 mp_alltoall_d11v 2046 13.8 1.861 2.194 1.861 2.194 dbcsr_complete_redistribute 325 12.2 0.323 0.357 1.499 1.959 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.956 1.957 cp_fm_upper_to_full 70 13.6 1.377 1.845 1.377 1.845 cp_fm_cholesky_decompose 22 10.9 1.733 1.752 1.733 1.752 mp_allgather_i34 2055 14.4 0.575 1.743 0.575 1.743 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.666 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.480 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.360 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.008 1.461 mp_waitany 17072 13.8 1.166 1.399 1.166 1.399 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.297 1.303 acc_transpose_blocks 16440 15.4 0.072 0.075 1.243 1.256 rs_gather_matrices 110 12.3 0.138 0.151 0.906 1.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.071000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.727273, yerr=8.995867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 737.529856E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.076 67.136 67.137 qs_mol_dyn_low 1 2.0 0.003 0.003 66.782 66.795 qs_forces 11 3.9 0.002 0.002 66.709 66.711 qs_energies 11 4.9 0.002 0.006 63.262 63.267 scf_env_do_scf 11 5.9 0.001 0.001 54.886 54.889 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.393 42.393 velocity_verlet 10 3.0 0.001 0.001 38.155 38.157 dbcsr_multiply_generic 2055 12.4 0.117 0.119 29.986 30.259 qs_scf_new_mos 99 7.5 0.001 0.001 27.576 27.691 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.575 27.691 ot_scf_mini 99 9.5 0.003 0.003 25.930 26.026 multiply_cannon 2055 13.4 0.244 0.261 22.760 23.684 multiply_cannon_loop 2055 14.4 0.883 0.906 21.161 21.858 ot_mini 99 10.5 0.001 0.001 14.603 14.718 multiply_cannon_multrec 24660 15.4 4.202 6.966 12.638 13.890 init_scf_loop 11 6.9 0.001 0.004 12.450 12.451 rebuild_ks_matrix 110 8.3 0.000 0.000 12.201 12.311 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.200 12.311 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.799 10.894 prepare_preconditioner 11 7.9 0.000 0.000 10.675 10.689 make_preconditioner 11 8.9 0.000 0.002 10.675 10.689 qs_ot_get_derivative 99 11.5 0.001 0.001 10.424 10.523 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.856 10.374 dbcsr_mm_accdrv_process 52304 16.0 6.899 8.480 8.289 9.189 mp_waitall_1 126806 16.4 4.963 6.872 4.963 6.872 sum_up_and_integrate 110 10.3 0.068 0.071 6.575 6.587 integrate_v_rspace 110 11.3 0.003 0.003 6.507 6.522 qs_ot_get_p 110 10.4 0.001 0.001 6.258 6.391 make_m2s 4110 13.4 0.059 0.061 5.552 5.985 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.844 5.851 calculate_rho_elec 110 8.6 0.077 0.081 5.844 5.851 make_images 4110 14.4 0.575 0.690 5.411 5.840 init_scf_run 11 5.9 0.000 0.001 5.619 5.620 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.619 5.620 cp_fm_upper_to_full 70 13.8 3.385 4.819 3.385 4.819 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.349 4.363 ot_diis_step 99 11.5 0.011 0.011 4.138 4.138 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.038 4.115 apply_single 110 13.6 0.000 0.000 4.038 4.114 dbcsr_complete_redistribute 325 12.2 0.410 0.449 2.813 3.994 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.858 3.859 grid_integrate_task_list 110 12.3 3.254 3.449 3.254 3.449 pw_transfer 1331 11.6 0.065 0.072 3.407 3.442 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.378 3.422 multiply_cannon_sync_h2d 24660 15.4 3.158 3.382 3.158 3.382 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.214 3.378 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.302 3.339 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.539 3.287 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.243 3.245 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.410 3.233 make_images_data 4110 15.4 0.046 0.050 2.747 3.230 density_rs2pw 110 9.6 0.004 0.004 3.001 3.206 cp_fm_redistribute_end 48 14.0 1.612 3.201 1.614 3.202 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.127 3.180 cp_fm_diag_elpa_base 48 14.0 1.502 3.058 1.584 3.171 cp_fm_cholesky_invert 11 10.9 3.115 3.124 3.115 3.124 wfi_extrapolate 11 7.9 0.001 0.001 3.076 3.076 calculate_dm_sparse 110 9.5 0.001 0.001 2.959 2.981 rs_pw_transfer 902 11.9 0.010 0.011 2.747 2.972 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.810 2.958 mp_alltoall_i22 605 13.7 1.710 2.888 1.710 2.888 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.869 2.871 fft_wrap_pw1pw2_140 451 13.1 0.201 0.211 2.740 2.779 fft3d_ps 1111 14.6 1.065 1.092 2.605 2.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.476 2.518 calculate_first_density_matrix 1 7.0 0.000 0.002 2.452 2.454 grid_collocate_task_list 110 9.6 2.218 2.374 2.218 2.374 potential_pw2rs 110 12.3 0.012 0.013 2.331 2.355 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.246 2.248 mp_alltoall_d11v 2046 13.8 1.938 2.159 1.938 2.159 jit_kernel_multiply 11 15.6 1.060 2.011 1.060 2.011 cp_fm_cholesky_decompose 22 10.9 1.871 1.920 1.871 1.920 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.748 1.776 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.606 1.712 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.650 1.665 mp_sum_l 6514 12.8 1.016 1.594 1.016 1.594 acc_transpose_blocks 24660 15.4 0.104 0.106 1.538 1.570 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.556 1.569 multiply_cannon_metrocomm4 20550 15.4 0.057 0.059 0.851 1.500 mp_irecv_dv 62702 16.1 0.752 1.421 0.752 1.421 mp_waitany 13376 13.8 1.076 1.395 1.076 1.395 mp_alltoall_z22v 1111 16.6 1.324 1.389 1.324 1.389 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.137000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.909091, yerr=15.029998 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 819.482624E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.035 56.198 56.199 qs_mol_dyn_low 1 2.0 0.003 0.003 55.884 55.904 qs_forces 11 3.9 0.002 0.002 55.817 55.818 qs_energies 11 4.9 0.001 0.001 52.170 52.175 scf_env_do_scf 11 5.9 0.000 0.001 43.700 43.700 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.644 35.645 velocity_verlet 10 3.0 0.001 0.001 31.302 31.305 dbcsr_multiply_generic 2055 12.4 0.105 0.106 23.005 23.200 qs_scf_new_mos 99 7.5 0.001 0.001 21.360 21.400 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.359 21.399 ot_scf_mini 99 9.5 0.002 0.002 20.118 20.140 multiply_cannon 2055 13.4 0.248 0.258 17.336 18.856 multiply_cannon_loop 2055 14.4 0.324 0.336 15.970 16.352 rebuild_ks_matrix 110 8.3 0.000 0.000 11.581 11.602 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.580 11.602 ot_mini 99 10.5 0.001 0.001 10.646 10.660 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.297 10.314 multiply_cannon_multrec 8220 15.4 3.201 4.551 7.436 8.385 mp_waitall_1 106626 16.5 6.423 8.040 6.423 8.040 init_scf_loop 11 6.9 0.000 0.000 8.006 8.007 qs_ot_get_derivative 99 11.5 0.001 0.001 6.879 6.900 prepare_preconditioner 11 7.9 0.000 0.000 6.359 6.363 make_preconditioner 11 8.9 0.000 0.000 6.359 6.363 sum_up_and_integrate 110 10.3 0.079 0.081 6.283 6.295 integrate_v_rspace 110 11.3 0.003 0.003 6.203 6.215 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.940 6.013 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.722 5.737 calculate_rho_elec 110 8.6 0.115 0.115 5.721 5.736 qs_ot_get_p 110 10.4 0.001 0.001 5.508 5.543 init_scf_run 11 5.9 0.000 0.001 5.321 5.321 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.321 5.321 dbcsr_mm_accdrv_process 17442 15.9 2.812 3.975 4.105 5.026 make_m2s 4110 13.4 0.039 0.040 4.343 4.616 make_images 4110 14.4 0.634 0.690 4.213 4.486 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.992 4.208 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.043 4.048 ot_diis_step 99 11.5 0.012 0.012 3.745 3.746 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.735 3.736 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.662 3.688 apply_single 110 13.6 0.000 0.000 3.662 3.688 grid_integrate_task_list 110 12.3 3.358 3.526 3.358 3.526 pw_transfer 1331 11.6 0.065 0.070 3.358 3.374 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.251 3.272 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.163 3.164 cp_fm_cholesky_invert 11 10.9 3.134 3.137 3.134 3.137 cp_fm_redistribute_end 48 14.0 0.802 3.123 0.807 3.123 cp_fm_diag_elpa_base 48 14.0 2.134 2.926 2.307 3.083 multiply_cannon_sync_h2d 8220 15.4 2.902 3.023 2.902 3.023 make_images_data 4110 15.4 0.038 0.043 2.489 2.949 density_rs2pw 110 9.6 0.004 0.004 2.770 2.938 hybrid_alltoall_any 4261 16.3 0.199 0.859 2.394 2.828 fft_wrap_pw1pw2_140 451 13.1 0.211 0.219 2.763 2.787 wfi_extrapolate 11 7.9 0.001 0.001 2.748 2.748 qs_energies_init_hamiltonians 11 5.9 0.005 0.013 2.684 2.685 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.631 2.632 fft3d_ps 1111 14.6 1.123 1.162 2.510 2.531 calculate_dm_sparse 110 9.5 0.001 0.001 2.464 2.504 grid_collocate_task_list 110 9.6 2.324 2.490 2.324 2.490 calculate_first_density_matrix 1 7.0 0.000 0.000 2.483 2.485 rs_pw_transfer 902 11.9 0.010 0.010 2.304 2.471 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.159 2.167 potential_pw2rs 110 12.3 0.015 0.016 2.024 2.035 mp_alltoall_d11v 2046 13.8 1.708 2.007 1.708 2.007 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.989 2.004 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.778 1.988 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.873 1.888 cp_fm_cholesky_decompose 22 10.9 1.771 1.788 1.771 1.788 dbcsr_complete_redistribute 325 12.2 0.628 0.693 1.600 1.699 mp_allgather_i34 2055 14.4 0.548 1.691 0.548 1.691 qs_env_update_s_mstruct 11 6.9 0.007 0.015 1.514 1.642 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.617 1.620 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.427 1.436 qs_create_task_list 11 7.9 0.008 0.009 1.232 1.334 generate_qs_task_list 11 8.9 0.378 0.445 1.223 1.327 mp_waitany 9240 13.8 1.066 1.233 1.066 1.233 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.860 1.203 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.180 1.197 jit_kernel_multiply 7 15.8 0.984 1.195 0.984 1.195 mp_alltoall_z22v 1111 16.6 1.153 1.176 1.153 1.176 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.199000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=774.272727, yerr=10.154183 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 1.338393E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.037 88.607 88.608 qs_mol_dyn_low 1 2.0 0.003 0.003 88.298 88.311 qs_forces 11 3.9 0.008 0.009 88.232 88.233 qs_energies 11 4.9 0.001 0.001 84.155 84.157 scf_env_do_scf 11 5.9 0.000 0.001 74.123 74.123 velocity_verlet 10 3.0 0.001 0.001 57.108 57.114 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.121 44.123 init_scf_loop 11 6.9 0.000 0.000 29.926 29.929 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.030 29.139 prepare_preconditioner 11 7.9 0.000 0.000 27.935 27.943 make_preconditioner 11 8.9 0.000 0.000 27.935 27.943 qs_scf_new_mos 99 7.5 0.001 0.001 27.399 27.458 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.398 27.458 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.891 27.393 ot_scf_mini 99 9.5 0.002 0.002 25.594 25.638 multiply_cannon 2055 13.4 0.350 0.369 21.736 22.510 multiply_cannon_loop 2055 14.4 0.343 0.347 19.887 20.189 cp_fm_upper_to_full 70 14.2 13.138 18.894 13.138 18.894 ot_mini 99 10.5 0.001 0.001 13.966 14.014 rebuild_ks_matrix 110 8.3 0.000 0.001 13.382 13.430 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.381 13.429 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.148 12.191 dbcsr_complete_redistribute 325 12.2 1.030 1.053 7.741 11.088 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.638 9.974 multiply_cannon_multrec 8220 15.4 4.350 4.547 9.550 9.677 qs_ot_get_derivative 99 11.5 0.001 0.001 9.332 9.379 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.028 9.327 mp_waitall_1 87304 16.6 8.251 9.218 8.251 9.218 mp_alltoall_i22 605 13.7 5.653 9.014 5.653 9.014 sum_up_and_integrate 110 10.3 0.150 0.151 6.803 6.817 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.690 6.729 calculate_rho_elec 110 8.6 0.227 0.227 6.689 6.729 integrate_v_rspace 110 11.3 0.004 0.004 6.652 6.666 qs_ot_get_p 110 10.4 0.001 0.001 6.336 6.389 make_m2s 4110 13.4 0.043 0.044 5.463 6.002 cp_fm_cholesky_invert 11 10.9 5.927 5.932 5.927 5.932 make_images 4110 14.4 0.879 0.932 5.275 5.814 init_scf_run 11 5.9 0.000 0.001 5.724 5.724 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.724 5.724 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.991 5.312 dbcsr_mm_accdrv_process 11614 15.7 3.234 3.660 5.057 5.310 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.726 5.204 apply_single 110 13.6 0.000 0.000 4.726 5.204 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.680 4.687 ot_diis_step 99 11.5 0.015 0.016 4.605 4.605 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.226 4.226 pw_transfer 1331 11.6 0.074 0.074 3.964 3.975 multiply_cannon_sync_h2d 8220 15.4 3.943 3.947 3.943 3.947 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.848 3.859 hybrid_alltoall_any 4261 16.3 0.256 0.552 3.017 3.731 make_images_data 4110 15.4 0.041 0.044 2.984 3.726 grid_integrate_task_list 110 12.3 3.660 3.715 3.660 3.715 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.672 3.673 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.566 3.568 cp_fm_diag_elpa_base 48 14.0 3.010 3.217 3.564 3.566 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.975 3.417 wfi_extrapolate 11 7.9 0.001 0.001 3.365 3.365 fft_wrap_pw1pw2_140 451 13.1 0.215 0.216 3.311 3.321 calculate_dm_sparse 110 9.5 0.001 0.001 3.170 3.190 density_rs2pw 110 9.6 0.004 0.004 3.148 3.164 fft3d_ps 1111 14.6 1.270 1.277 3.056 3.066 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.900 2.903 grid_collocate_task_list 110 9.6 2.628 2.653 2.628 2.653 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.488 2.514 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.372 2.386 cp_fm_cholesky_decompose 22 10.9 2.300 2.328 2.300 2.328 rs_pw_transfer 902 11.9 0.011 0.011 2.288 2.324 mp_alltoall_d11v 2046 13.8 2.217 2.267 2.217 2.267 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.193 2.256 calculate_first_density_matrix 1 7.0 0.000 0.000 2.251 2.251 potential_pw2rs 110 12.3 0.021 0.021 2.197 2.202 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.086 2.155 qs_create_task_list 11 7.9 0.001 0.001 1.904 1.948 generate_qs_task_list 11 8.9 0.737 0.792 1.904 1.948 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.862 1.901 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.839 1.844 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.608000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1195.454545, yerr=57.752393 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 625.754112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 56756. MP_Allreduce 11253 786. MP_Sync 168 MP_Alltoall 2210 1425383. MP_ISendRecv 48260 18752. MP_Wait 66280 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.027 204.065 204.066 qs_mol_dyn_low 1 2.0 0.003 0.003 203.681 203.694 qs_forces 11 3.9 0.003 0.003 203.579 203.580 qs_energies 11 4.9 0.001 0.003 198.016 198.035 scf_env_do_scf 11 5.9 0.001 0.001 181.387 181.390 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 160.232 160.234 dbcsr_multiply_generic 2485 12.5 0.174 0.177 123.797 124.360 velocity_verlet 10 3.0 0.001 0.001 121.663 121.664 qs_scf_new_mos 116 7.6 0.001 0.001 120.886 121.120 qs_scf_loop_do_ot 116 8.6 0.001 0.001 120.885 121.120 ot_scf_mini 116 9.6 0.003 0.003 114.312 114.469 multiply_cannon 2485 13.5 0.236 0.244 100.054 102.177 multiply_cannon_loop 2485 14.5 2.105 2.175 97.703 99.388 ot_mini 116 10.6 0.001 0.001 65.537 65.734 multiply_cannon_multrec 59640 15.5 32.762 35.176 41.113 43.307 qs_ot_get_derivative 116 11.6 0.001 0.001 40.907 41.127 rebuild_ks_matrix 127 8.3 0.001 0.001 33.469 33.674 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.468 33.673 mp_waitall_1 288940 16.2 29.313 32.376 29.313 32.376 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.134 30.326 multiply_cannon_sync_h2d 59640 15.5 26.985 29.171 26.985 29.171 qs_ot_get_p 127 10.4 0.001 0.001 27.044 27.248 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.113 24.876 apply_single 127 13.6 0.001 0.001 24.113 24.876 ot_diis_step 116 11.6 0.007 0.008 24.230 24.232 init_scf_loop 11 6.9 0.000 0.000 21.070 21.071 qs_ot_p2m_diag 82 11.4 0.076 0.090 20.230 20.326 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.140 19.319 multiply_cannon_metrocomm3 59640 15.5 0.108 0.112 15.535 17.749 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.536 17.537 prepare_preconditioner 11 7.9 0.000 0.000 16.444 16.486 make_preconditioner 11 8.9 0.000 0.000 16.444 16.486 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.643 15.803 cp_fm_diag_elpa 82 13.4 0.000 0.001 14.517 14.523 make_m2s 4970 13.5 0.102 0.110 14.156 14.483 cp_fm_redistribute_end 82 14.4 11.458 14.428 11.471 14.432 sum_up_and_integrate 127 10.3 0.089 0.107 14.326 14.343 make_images 4970 14.5 0.400 0.419 13.977 14.313 integrate_v_rspace 127 11.3 0.003 0.004 14.237 14.254 cp_fm_diag_elpa_base 82 14.4 2.917 14.076 2.947 14.205 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.138 13.208 calculate_rho_elec 127 8.7 0.044 0.063 13.137 13.208 init_scf_run 11 5.9 0.000 0.001 12.438 12.439 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.438 12.439 cp_fm_cholesky_invert 11 10.9 9.444 9.451 9.444 9.451 mp_sum_l 7804 13.0 8.332 9.205 8.332 9.205 wfi_extrapolate 11 7.9 0.001 0.001 9.083 9.083 calculate_dm_sparse 127 9.5 0.001 0.001 8.372 8.478 dbcsr_mm_accdrv_process 123452 16.2 3.330 3.504 7.918 8.477 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.982 8.154 make_images_data 4970 15.5 0.069 0.075 7.039 8.040 multiply_cannon_metrocomm1 59640 15.5 0.090 0.094 6.010 8.022 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.824 7.920 density_rs2pw 127 9.7 0.006 0.007 6.951 7.842 pw_transfer 1535 11.6 0.074 0.106 7.216 7.519 grid_integrate_task_list 127 12.3 7.057 7.492 7.057 7.492 hybrid_alltoall_any 5155 16.4 0.288 2.250 6.119 7.460 fft_wrap_pw1pw2 1281 12.7 0.010 0.014 7.014 7.305 rs_pw_transfer 1038 11.9 0.017 0.019 6.177 7.193 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.812 6.819 fft_wrap_pw1pw2_140 519 13.2 0.436 0.507 5.860 6.075 fft3d_ps 1281 14.7 2.075 2.549 5.822 6.054 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.870 5.968 mp_alltoall_d11v 2401 14.1 4.389 5.766 4.389 5.766 grid_collocate_task_list 127 9.7 4.691 5.148 4.691 5.148 potential_pw2rs 127 12.3 0.009 0.011 4.815 4.861 cp_fm_cholesky_decompose 22 10.9 4.752 4.765 4.752 4.765 mp_sum_d 4446 12.1 3.577 4.363 3.577 4.363 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.066000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.545455, yerr=6.155022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 2.199914E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6022464 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.2 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 832.987136E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2425920 MPI messages size (bytes): total size 4.132350E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703416E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71436 2336489472 32768 < size <= 131072 728832 55956209664 131072 < size <= 4194304 1386864 1409906900992 4194304 < size <= 16777216 155760 1473826772352 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4113 56823. MP_Allreduce 11322 944. MP_Sync 170 MP_Alltoall 1983 4696046. MP_ISendRecv 24252 47072. MP_Wait 38240 MP_comm_split 83 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 195.299 195.301 qs_mol_dyn_low 1 2.0 0.003 0.003 194.960 194.974 qs_forces 11 3.9 0.003 0.003 193.547 193.548 qs_energies 11 4.9 0.001 0.002 186.816 186.828 scf_env_do_scf 11 5.9 0.001 0.001 170.495 170.505 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 136.955 136.957 velocity_verlet 10 3.0 0.001 0.001 123.399 123.428 dbcsr_multiply_generic 2527 12.6 0.187 0.194 98.178 99.422 qs_scf_new_mos 118 7.6 0.001 0.001 97.254 97.922 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.253 97.921 ot_scf_mini 118 9.6 0.004 0.004 92.395 93.133 multiply_cannon 2527 13.6 0.480 0.536 77.678 81.886 multiply_cannon_loop 2527 14.6 1.270 1.318 74.495 77.386 ot_mini 118 10.6 0.001 0.001 50.574 51.226 mp_waitall_1 228564 16.4 25.663 38.978 25.663 38.978 multiply_cannon_multrec 30324 15.6 22.014 25.781 31.691 35.935 rebuild_ks_matrix 129 8.3 0.001 0.001 33.096 33.656 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 33.096 33.656 init_scf_loop 11 6.9 0.000 0.000 33.448 33.449 qs_ks_update_qs_env 129 7.6 0.001 0.001 29.853 30.368 prepare_preconditioner 11 7.9 0.000 0.000 29.082 29.131 make_preconditioner 11 8.9 0.000 0.000 29.082 29.131 qs_ot_get_derivative 118 11.6 0.001 0.002 28.370 29.083 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.789 28.315 multiply_cannon_metrocomm3 30324 15.6 0.095 0.100 15.996 28.147 qs_ot_get_p 129 10.4 0.001 0.001 23.508 24.104 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.231 23.313 apply_single 129 13.6 0.001 0.001 22.231 23.312 ot_diis_step 118 11.6 0.014 0.015 22.037 22.039 multiply_cannon_sync_h2d 30324 15.6 19.250 21.428 19.250 21.428 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.512 18.559 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.308 17.309 cp_fm_cholesky_invert 11 10.9 17.127 17.140 17.127 17.140 make_m2s 5054 13.6 0.089 0.094 14.600 15.954 make_images 5054 14.6 1.155 1.351 14.392 15.748 sum_up_and_integrate 129 10.3 0.117 0.133 14.688 14.718 integrate_v_rspace 129 11.3 0.003 0.004 14.571 14.606 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.084 14.093 cp_fm_redistribute_end 83 14.4 8.221 14.005 8.235 14.008 cp_fm_diag_elpa_base 83 14.4 5.531 13.525 5.753 13.882 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.272 13.306 calculate_rho_elec 129 8.7 0.088 0.106 13.272 13.305 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.189 11.704 init_scf_run 11 5.9 0.000 0.001 11.545 11.547 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.545 11.547 multiply_cannon_metrocomm4 27797 15.6 0.098 0.111 3.823 10.881 make_images_data 5054 15.6 0.067 0.075 8.892 10.521 mp_irecv_dv 70031 16.3 3.625 10.486 3.625 10.486 dbcsr_mm_accdrv_process 62734 16.2 4.595 5.359 9.140 9.748 hybrid_alltoall_any 5240 16.5 0.345 1.522 7.508 9.666 wfi_extrapolate 11 7.9 0.001 0.001 8.356 8.356 pw_transfer 1559 11.6 0.085 0.109 7.976 8.042 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 7.750 7.808 grid_integrate_task_list 129 12.3 7.229 7.635 7.229 7.635 density_rs2pw 129 9.7 0.006 0.006 6.977 7.359 cp_fm_cholesky_decompose 22 10.9 7.056 7.125 7.056 7.125 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 6.382 7.098 fft_wrap_pw1pw2_140 527 13.2 0.471 0.520 6.719 6.786 calculate_dm_sparse 129 9.5 0.001 0.001 6.523 6.661 rs_pw_transfer 1054 12.0 0.015 0.017 5.870 6.300 fft3d_ps 1301 14.7 2.795 2.946 6.112 6.162 mp_sum_l 7930 13.1 4.101 6.154 4.101 6.154 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.079 6.088 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.421 5.482 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.223 5.375 grid_collocate_task_list 129 9.7 4.944 5.372 4.944 5.372 mp_alltoall_d11v 2423 14.1 4.274 5.076 4.274 5.076 potential_pw2rs 129 12.3 0.015 0.018 5.025 5.045 mp_allgather_i34 2527 14.6 1.749 4.990 1.749 4.990 dbcsr_complete_redistribute 395 12.7 0.765 0.841 3.218 4.080 mp_sum_d 4496 12.2 2.674 4.034 2.674 4.034 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=195.301000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.363636, yerr=4.333227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 936.120320E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.046 179.448 179.449 qs_mol_dyn_low 1 2.0 0.003 0.004 178.941 178.954 qs_forces 11 3.9 0.003 0.004 178.834 178.841 qs_energies 11 4.9 0.015 0.048 172.300 172.311 scf_env_do_scf 11 5.9 0.001 0.002 156.551 156.553 scf_env_do_scf_inner_loop 117 6.6 0.003 0.010 120.994 120.994 velocity_verlet 10 3.0 0.001 0.001 114.213 114.215 qs_scf_new_mos 117 7.6 0.001 0.001 83.907 84.226 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.906 84.225 dbcsr_multiply_generic 2507 12.6 0.184 0.194 82.177 83.209 ot_scf_mini 117 9.6 0.004 0.004 79.742 80.094 multiply_cannon 2507 13.6 0.493 0.511 61.915 66.656 multiply_cannon_loop 2507 14.6 0.861 0.886 58.890 61.449 ot_mini 117 10.6 0.001 0.001 42.978 43.334 mp_waitall_1 178456 16.5 26.558 36.909 26.558 36.909 init_scf_loop 11 6.9 0.001 0.002 35.451 35.452 prepare_preconditioner 11 7.9 0.000 0.000 31.407 31.457 make_preconditioner 11 8.9 0.000 0.000 31.407 31.457 rebuild_ks_matrix 128 8.3 0.001 0.001 30.452 30.896 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.452 30.896 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.062 30.413 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.530 27.932 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.919 25.977 multiply_cannon_multrec 20056 15.6 13.275 15.999 21.916 24.607 qs_ot_get_derivative 117 11.6 0.001 0.002 22.958 23.305 qs_ot_get_p 128 10.4 0.001 0.001 22.041 22.461 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.140 21.147 apply_single 128 13.6 0.001 0.001 20.140 21.146 ot_diis_step 117 11.6 0.018 0.018 19.914 19.915 qs_ot_p2m_diag 83 11.4 0.266 0.272 17.711 17.724 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.666 16.667 make_m2s 5014 13.6 0.081 0.085 15.258 15.973 make_images 5014 14.6 1.175 1.281 15.026 15.735 multiply_cannon_sync_h2d 20056 15.6 14.090 15.685 14.090 15.685 cp_fm_cholesky_invert 11 10.9 14.819 14.829 14.819 14.829 sum_up_and_integrate 128 10.3 0.134 0.145 14.444 14.473 integrate_v_rspace 128 11.3 0.003 0.004 14.309 14.336 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.334 13.337 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.252 13.294 calculate_rho_elec 128 8.7 0.132 0.147 13.251 13.293 cp_fm_redistribute_end 83 14.4 5.043 13.257 5.061 13.262 cp_fm_diag_elpa_base 83 14.4 7.763 12.608 8.175 13.120 init_scf_run 11 5.9 0.000 0.001 10.621 10.621 scf_env_initial_rho_setup 11 6.9 0.011 0.065 10.621 10.621 make_images_data 5014 15.6 0.060 0.067 9.498 10.612 hybrid_alltoall_any 5200 16.5 0.430 1.964 8.115 9.939 multiply_cannon_metrocomm4 17549 15.6 0.063 0.073 3.532 9.541 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.083 9.343 mp_irecv_dv 50230 16.2 3.407 9.289 3.407 9.289 dbcsr_mm_accdrv_process 41502 16.2 4.465 5.265 8.103 8.210 pw_transfer 1547 11.6 0.084 0.103 8.020 8.132 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.797 7.913 grid_integrate_task_list 128 12.3 7.271 7.733 7.271 7.733 cp_fm_cholesky_decompose 22 10.9 7.500 7.542 7.500 7.542 wfi_extrapolate 11 7.9 0.001 0.001 7.471 7.471 cp_fm_upper_to_full 105 14.5 5.683 7.430 5.683 7.430 density_rs2pw 128 9.7 0.006 0.006 6.696 7.287 fft_wrap_pw1pw2_140 523 13.2 0.476 0.529 6.806 6.936 dbcsr_complete_redistribute 395 12.7 1.180 1.210 4.711 6.467 fft3d_ps 1291 14.7 2.692 2.901 6.068 6.137 rs_pw_transfer 1046 11.9 0.014 0.014 5.330 5.982 calculate_dm_sparse 128 9.5 0.001 0.001 5.771 5.858 mp_alltoall_d11v 2415 14.1 4.501 5.675 4.501 5.675 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.562 5.565 grid_collocate_task_list 128 9.7 5.079 5.440 5.079 5.440 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.564 5.235 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.465 5.221 potential_pw2rs 128 12.3 0.020 0.022 4.751 4.779 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.600 4.737 mp_allgather_i34 2507 14.6 1.551 4.656 1.551 4.656 mp_sum_l 7870 13.0 3.199 4.572 3.199 4.572 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.052 4.086 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.325 4.053 mp_alltoall_i22 716 14.1 1.917 3.802 1.917 3.802 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.762 3.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=179.449000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=887.909091, yerr=10.004131 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.148359E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4087 57164. MP_Allreduce 11256 1068. MP_Sync 168 MP_Alltoall 1700 12496381. MP_ISendRecv 11684 75008. MP_Wait 28114 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.034 189.376 189.377 qs_mol_dyn_low 1 2.0 0.003 0.003 188.981 188.993 qs_forces 11 3.9 0.003 0.003 188.862 188.869 qs_energies 11 4.9 0.001 0.002 181.815 181.826 scf_env_do_scf 11 5.9 0.001 0.001 165.131 165.142 velocity_verlet 10 3.0 0.001 0.001 124.596 124.599 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.613 117.613 qs_scf_new_mos 116 7.6 0.001 0.001 81.921 82.228 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.920 82.227 dbcsr_multiply_generic 2485 12.5 0.190 0.196 78.974 79.615 ot_scf_mini 116 9.6 0.003 0.003 77.460 77.748 multiply_cannon 2485 13.5 0.550 0.578 54.305 56.773 multiply_cannon_loop 2485 14.5 1.170 1.202 50.542 52.139 init_scf_loop 11 6.9 0.000 0.000 47.390 47.391 prepare_preconditioner 11 7.9 0.000 0.000 43.308 43.331 make_preconditioner 11 8.9 0.000 0.000 43.308 43.331 ot_mini 116 10.6 0.001 0.001 42.082 42.360 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.867 41.949 multiply_cannon_multrec 29820 15.5 14.017 18.900 25.850 30.584 rebuild_ks_matrix 127 8.3 0.001 0.001 29.130 29.430 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 29.129 29.429 mp_waitall_1 152434 16.5 17.862 27.862 17.862 27.862 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.338 26.605 qs_ot_get_derivative 116 11.6 0.001 0.002 22.616 22.898 qs_ot_get_p 127 10.4 0.001 0.001 21.625 21.919 make_m2s 4970 13.5 0.095 0.100 20.304 21.388 make_images 4970 14.5 1.935 2.222 20.001 21.084 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.930 19.455 apply_single 127 13.6 0.001 0.001 18.929 19.455 ot_diis_step 116 11.6 0.017 0.018 19.339 19.340 qs_ot_p2m_diag 82 11.4 0.339 0.385 17.458 17.510 cp_fm_cholesky_invert 11 10.9 16.695 16.703 16.695 16.703 cp_fm_upper_to_full 104 14.7 11.266 16.632 11.266 16.632 cp_dbcsr_syevd 82 12.4 0.005 0.005 16.164 16.166 multiply_cannon_metrocomm3 29820 15.5 0.046 0.048 6.526 15.396 sum_up_and_integrate 127 10.3 0.140 0.153 14.214 14.239 integrate_v_rspace 127 11.3 0.003 0.004 14.073 14.105 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.212 13.253 calculate_rho_elec 127 8.7 0.174 0.189 13.211 13.252 dbcsr_complete_redistribute 393 12.7 1.503 1.629 9.121 13.025 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.007 13.009 cp_fm_redistribute_end 82 14.4 2.236 12.917 2.254 12.921 cp_fm_diag_elpa_base 82 14.4 10.050 12.290 10.630 12.771 make_images_data 4970 15.5 0.063 0.067 10.981 12.734 multiply_cannon_sync_h2d 29820 15.5 11.547 12.295 11.547 12.295 dbcsr_mm_accdrv_process 61748 16.2 7.274 8.359 11.410 11.954 hybrid_alltoall_any 5155 16.4 0.521 2.203 9.863 11.785 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.694 11.565 init_scf_run 11 5.9 0.000 0.001 10.802 10.803 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.801 10.803 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.418 10.133 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.372 9.572 mp_alltoall_i22 712 14.1 5.633 9.372 5.633 9.372 pw_transfer 1535 11.6 0.084 0.099 7.995 8.086 cp_fm_cholesky_decompose 22 10.9 7.822 7.916 7.822 7.916 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.772 7.868 grid_integrate_task_list 127 12.3 7.459 7.815 7.459 7.815 wfi_extrapolate 11 7.9 0.001 0.001 7.624 7.624 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.726 7.456 mp_irecv_dv 75445 16.2 2.581 7.194 2.581 7.194 fft_wrap_pw1pw2_140 519 13.2 0.478 0.488 6.844 6.963 density_rs2pw 127 9.7 0.005 0.006 6.364 6.667 calculate_dm_sparse 127 9.5 0.001 0.001 6.118 6.201 fft3d_ps 1281 14.7 2.788 2.886 6.012 6.075 grid_collocate_task_list 127 9.7 5.169 5.658 5.169 5.658 mp_alltoall_d11v 2401 14.1 4.955 5.505 4.955 5.505 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.432 5.497 rs_pw_transfer 1038 11.9 0.013 0.015 4.775 5.077 potential_pw2rs 127 12.3 0.023 0.023 4.430 4.455 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.369 4.428 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.347 4.424 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.400 4.400 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.189 4.250 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.377000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1085.090909, yerr=17.727506 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.519849E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57953. MP_Allreduce 11047 1173. MP_Sync 87 MP_Alltoall 1712 18838222. MP_ISendRecv 7680 122880. MP_Wait 19962 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.074 174.876 174.899 qs_mol_dyn_low 1 2.0 0.003 0.003 174.467 174.480 qs_forces 11 3.9 0.003 0.003 174.356 174.359 qs_energies 11 4.9 0.002 0.002 167.004 167.011 scf_env_do_scf 11 5.9 0.001 0.001 149.637 149.648 velocity_verlet 10 3.0 0.001 0.001 113.912 113.916 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 113.267 113.268 qs_scf_new_mos 117 7.6 0.001 0.001 77.367 77.462 qs_scf_loop_do_ot 117 8.6 0.001 0.001 77.366 77.461 dbcsr_multiply_generic 2507 12.6 0.186 0.196 75.216 75.851 ot_scf_mini 117 9.6 0.004 0.004 72.876 72.967 multiply_cannon 2507 13.6 0.587 0.630 55.090 59.228 multiply_cannon_loop 2507 14.6 0.450 0.460 50.400 51.558 ot_mini 117 10.6 0.001 0.001 40.236 40.344 init_scf_loop 11 6.9 0.000 0.000 36.212 36.214 mp_waitall_1 129618 16.6 27.180 34.801 27.180 34.801 prepare_preconditioner 11 7.9 0.000 0.000 32.366 32.395 make_preconditioner 11 8.9 0.000 0.000 32.366 32.395 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.187 30.464 rebuild_ks_matrix 128 8.3 0.001 0.001 28.719 28.838 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.017 28.718 28.838 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.232 26.339 multiply_cannon_multrec 10028 15.6 10.365 16.332 17.634 22.054 multiply_cannon_metrocomm3 10028 15.6 0.022 0.024 13.059 20.200 qs_ot_get_derivative 117 11.6 0.002 0.002 20.040 20.136 ot_diis_step 117 11.6 0.020 0.020 20.122 20.123 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.657 20.059 apply_single 128 13.6 0.001 0.001 19.656 20.058 qs_ot_get_p 128 10.4 0.001 0.001 19.577 19.711 make_m2s 5014 13.6 0.066 0.072 16.400 18.752 cp_fm_cholesky_invert 11 10.9 18.622 18.628 18.622 18.628 make_images 5014 14.6 2.322 2.888 16.095 18.438 qs_ot_p2m_diag 83 11.4 0.495 0.502 15.683 15.699 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.527 14.528 sum_up_and_integrate 128 10.3 0.180 0.190 14.170 14.217 integrate_v_rspace 128 11.3 0.004 0.004 13.989 14.046 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.280 13.314 calculate_rho_elec 128 8.7 0.258 0.269 13.279 13.313 make_images_data 5014 15.6 0.052 0.060 9.964 12.797 hybrid_alltoall_any 5200 16.5 0.823 3.711 9.798 12.178 multiply_cannon_sync_h2d 10028 15.6 11.503 12.092 11.503 12.092 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.394 11.408 cp_fm_diag_elpa_base 83 14.4 11.140 11.236 11.386 11.400 init_scf_run 11 5.9 0.000 0.001 10.530 10.530 scf_env_initial_rho_setup 11 6.9 0.003 0.016 10.529 10.530 grid_integrate_task_list 128 12.3 7.713 8.179 7.713 8.179 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.112 8.171 cp_fm_cholesky_decompose 22 10.9 8.023 8.139 8.023 8.139 pw_transfer 1547 11.6 0.083 0.092 7.907 7.933 multiply_cannon_metrocomm1 10028 15.6 0.029 0.030 4.562 7.897 fft_wrap_pw1pw2 1291 12.7 0.010 0.010 7.687 7.711 dbcsr_mm_accdrv_process 20762 16.1 2.669 3.437 6.904 7.560 wfi_extrapolate 11 7.9 0.001 0.001 7.377 7.377 mp_allgather_i34 2507 14.6 2.864 7.118 2.864 7.118 fft_wrap_pw1pw2_140 523 13.2 0.501 0.523 6.737 6.769 density_rs2pw 128 9.7 0.005 0.006 6.086 6.591 calculate_dm_sparse 128 9.5 0.001 0.001 6.151 6.238 mp_alltoall_d11v 2415 14.1 4.922 6.094 4.922 6.094 multiply_cannon_metrocomm4 7521 15.6 0.024 0.026 1.876 5.897 fft3d_ps 1291 14.7 2.725 2.791 5.846 5.885 grid_collocate_task_list 128 9.7 5.483 5.873 5.483 5.873 mp_irecv_dv 28860 15.9 1.839 5.808 1.839 5.808 dbcsr_complete_redistribute 395 12.7 2.131 2.220 5.352 5.721 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.448 5.474 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.297 5.298 rs_pw_transfer 1046 11.9 0.013 0.013 4.212 4.739 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.279 4.320 potential_pw2rs 128 12.3 0.026 0.027 4.156 4.175 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.576 3.897 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.526 3.858 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.772 3.813 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.593 3.667 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.500 3.542 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.899000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1416.181818, yerr=59.802430 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_performance_tests/20/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 11.606413E+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 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 3.129078E+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 11.342275E+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 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.090 294.459 294.488 qs_mol_dyn_low 1 2.0 0.003 0.003 293.908 293.920 qs_forces 11 3.9 0.003 0.003 293.815 293.821 qs_energies 11 4.9 0.001 0.002 285.206 285.212 scf_env_do_scf 11 5.9 0.001 0.001 263.210 263.221 velocity_verlet 10 3.0 0.001 0.001 211.306 211.313 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 135.617 135.620 init_scf_loop 11 6.9 0.000 0.000 127.329 127.333 prepare_preconditioner 11 7.9 0.000 0.000 122.655 122.683 make_preconditioner 11 8.9 0.000 0.000 122.655 122.683 make_full_inverse_cholesky 11 9.9 0.000 0.000 98.069 119.697 qs_scf_new_mos 116 7.6 0.001 0.001 93.655 93.743 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.654 93.742 ot_scf_mini 116 9.6 0.004 0.004 88.854 88.898 dbcsr_multiply_generic 2485 12.5 0.212 0.219 83.730 84.384 cp_fm_upper_to_full 104 14.8 54.602 78.091 54.602 78.091 multiply_cannon 2485 13.5 0.701 0.747 58.687 59.493 multiply_cannon_loop 2485 14.5 0.470 0.478 54.959 56.488 ot_mini 116 10.6 0.001 0.001 44.997 45.036 dbcsr_complete_redistribute 393 12.7 3.991 4.032 30.407 44.316 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.939 40.765 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.540 38.177 mp_alltoall_i22 712 14.1 22.324 36.340 22.324 36.340 cp_fm_cholesky_invert 11 10.9 34.223 34.229 34.223 34.229 rebuild_ks_matrix 127 8.3 0.001 0.001 33.430 33.503 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.429 33.502 mp_waitall_1 104546 16.7 29.187 32.896 29.187 32.896 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.178 31.253 qs_ot_get_p 127 10.4 0.001 0.001 28.273 28.341 qs_ot_get_derivative 116 11.6 0.001 0.002 24.611 24.654 qs_ot_p2m_diag 82 11.4 0.868 0.874 24.029 24.060 cp_dbcsr_syevd 82 12.4 0.005 0.006 22.293 22.295 make_m2s 4970 13.5 0.075 0.080 20.304 21.326 make_images 4970 14.5 3.736 3.881 19.831 20.855 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 19.119 20.526 ot_diis_step 116 11.6 0.022 0.022 20.345 20.345 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.602 19.832 apply_single 127 13.6 0.001 0.001 19.601 19.832 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.015 19.017 cp_fm_diag_elpa_base 82 14.4 14.529 16.214 19.011 19.012 multiply_cannon_multrec 9940 15.5 10.358 12.136 17.739 17.809 sum_up_and_integrate 127 10.3 0.318 0.321 15.654 15.741 multiply_cannon_sync_h2d 9940 15.5 15.521 15.539 15.521 15.539 integrate_v_rspace 127 11.3 0.004 0.004 15.336 15.423 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.041 15.058 calculate_rho_elec 127 8.7 0.479 0.480 15.041 15.058 make_images_data 4970 15.5 0.059 0.064 11.038 12.967 hybrid_alltoall_any 5155 16.4 1.293 3.031 11.129 12.954 init_scf_run 11 5.9 0.000 0.001 11.949 11.949 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.948 11.948 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.587 9.628 cp_fm_cholesky_decompose 22 10.9 9.096 9.118 9.096 9.118 dbcsr_mm_accdrv_process 20590 16.0 3.809 5.745 7.146 8.977 wfi_extrapolate 11 7.9 0.001 0.001 8.796 8.796 grid_integrate_task_list 127 12.3 8.476 8.647 8.476 8.647 pw_transfer 1535 11.6 0.090 0.091 8.507 8.522 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.274 8.289 qs_energies_init_hamiltonians 11 5.9 0.031 0.034 8.030 8.031 fft_wrap_pw1pw2_140 519 13.2 0.536 0.539 7.298 7.316 mp_alltoall_d11v 2401 14.1 7.047 7.177 7.047 7.177 calculate_dm_sparse 127 9.5 0.001 0.001 6.633 6.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.418 6.471 fft3d_ps 1281 14.7 2.742 2.763 6.325 6.337 grid_collocate_task_list 127 9.7 6.289 6.321 6.289 6.321 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.158 6.224 density_rs2pw 127 9.7 0.005 0.005 6.137 6.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=294.488000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2771.818182, yerr=179.149414 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 1.260040E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 84.915 84.916 qs_energies 1 2.0 0.000 0.000 84.505 84.512 ls_scf 1 3.0 0.000 0.000 83.590 83.598 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.537 72.713 multiply_cannon 111 7.7 0.017 0.021 55.740 56.882 multiply_cannon_loop 111 8.7 0.213 0.226 52.332 53.752 ls_scf_main 1 4.0 0.000 0.000 52.314 52.314 density_matrix_trs4 2 5.0 0.002 0.003 46.808 46.879 ls_scf_init_scf 1 4.0 0.000 0.000 28.239 28.240 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.178 27.230 mp_waitall_1 11316 10.9 22.421 26.107 22.421 26.107 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.033 25.056 multiply_cannon_multrec 2664 9.7 8.207 8.984 15.535 17.273 multiply_cannon_sync_h2d 2664 9.7 13.626 15.135 13.626 15.135 make_m2s 222 7.7 0.008 0.011 13.229 13.689 make_images 222 8.7 0.099 0.109 13.207 13.668 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.553 12.409 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.397 8.492 make_images_data 222 9.7 0.004 0.005 7.781 8.357 hybrid_alltoall_any 227 10.6 0.215 1.844 6.702 7.938 dbcsr_mm_accdrv_process 4760 10.4 0.517 0.638 6.946 7.885 dbcsr_mm_accdrv_process_sort 4760 11.4 6.231 7.091 6.231 7.091 calculate_norms 4752 9.8 5.559 6.237 5.559 6.237 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.048 5.167 mp_sum_l 807 5.4 3.029 4.454 3.029 4.454 multiply_cannon_metrocomm4 2442 9.7 0.011 0.015 2.055 3.775 mp_irecv_dv 6231 10.9 2.037 3.747 2.037 3.747 make_images_sizes 222 9.7 0.000 0.000 0.720 3.585 mp_alltoall_i44 222 10.7 0.720 3.584 0.720 3.584 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.254 3.476 arnoldi_extremal 4 6.8 0.000 0.000 3.363 3.390 arnoldi_normal_ev 4 7.8 0.001 0.002 3.363 3.390 build_subspace 16 8.4 0.009 0.012 3.261 3.263 ls_scf_post 1 4.0 0.000 0.000 3.037 3.045 ls_scf_store_result 1 5.0 0.000 0.000 2.859 2.899 dbcsr_special_finalize 555 9.7 0.005 0.006 2.375 2.816 dbcsr_merge_single_wm 555 10.7 0.459 0.582 2.367 2.808 make_images_pack 222 9.7 2.207 2.617 2.208 2.619 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.335 2.579 dbcsr_sort_data 658 11.4 2.165 2.543 2.165 2.543 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.468 2.073 2.470 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.241 2.315 buffer_matrices_ensure_size 222 8.7 1.751 2.079 1.751 2.079 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.747 1.748 rebuild_ks_matrix 3 7.3 0.000 0.000 1.737 1.739 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.737 1.739 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.916000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1141.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 2.102981E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.124 0.234 91.724 91.726 qs_energies 1 2.0 0.001 0.006 90.924 90.969 ls_scf 1 3.0 0.005 0.040 89.562 89.606 dbcsr_multiply_generic 111 6.7 0.016 0.024 75.508 75.832 multiply_cannon 111 7.7 0.029 0.054 53.295 57.813 ls_scf_main 1 4.0 0.000 0.008 55.075 55.084 multiply_cannon_loop 111 8.7 0.117 0.123 49.978 53.746 density_matrix_trs4 2 5.0 0.002 0.008 49.397 49.602 ls_scf_init_scf 1 4.0 0.001 0.003 30.869 30.890 mp_waitall_1 9246 10.9 21.030 29.817 21.030 29.817 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.693 29.772 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.005 27.244 27.262 multiply_cannon_multrec 1332 9.7 13.115 16.260 22.318 26.560 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.864 20.703 make_m2s 222 7.7 0.007 0.009 15.062 15.710 make_images 222 8.7 1.569 1.971 15.032 15.680 dbcsr_mm_accdrv_process 4041 10.4 0.285 0.453 8.803 10.391 dbcsr_mm_accdrv_process_sort 4041 11.4 8.395 9.938 8.395 9.938 make_images_data 222 9.7 0.004 0.005 8.551 9.550 mp_sum_l 807 5.4 5.885 9.411 5.885 9.411 hybrid_alltoall_any 227 10.6 0.523 2.479 8.023 8.861 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.245 7.668 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.596 7.649 mp_irecv_dv 3311 11.0 3.225 7.605 3.225 7.605 calculate_norms 2376 9.8 6.001 6.577 6.001 6.577 multiply_cannon_sync_h2d 1332 9.7 4.732 5.858 4.732 5.858 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.103 5.288 arnoldi_extremal 4 6.8 0.000 0.000 4.775 4.800 arnoldi_normal_ev 4 7.8 0.001 0.004 4.775 4.800 build_subspace 16 8.4 0.014 0.021 4.495 4.499 ls_scf_post 1 4.0 0.014 0.064 3.613 3.663 ls_scf_store_result 1 5.0 0.000 0.000 3.297 3.428 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.191 3.395 dbcsr_matrix_vector_mult_local 304 10.0 2.763 3.251 2.765 3.253 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.234 2.724 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.510 2.568 make_images_pack 222 9.7 2.030 2.436 2.032 2.438 mp_allgather_i34 111 8.7 1.021 2.362 1.021 2.362 dbcsr_sort_data 436 11.2 1.820 2.008 1.820 2.008 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.842 1.865 rebuild_ks_matrix 3 7.3 0.000 0.000 1.829 1.853 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.030 1.829 1.853 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.726000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1708.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 2.685690E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.084 93.168 93.170 qs_energies 1 2.0 0.000 0.000 92.533 92.536 ls_scf 1 3.0 0.000 0.001 91.133 91.135 dbcsr_multiply_generic 111 6.7 0.016 0.017 75.768 76.053 ls_scf_main 1 4.0 0.001 0.014 57.200 57.208 multiply_cannon 111 7.7 0.043 0.107 52.539 56.441 multiply_cannon_loop 111 8.7 0.100 0.105 48.971 52.129 density_matrix_trs4 2 5.0 0.002 0.003 51.204 51.361 mp_waitall_1 7374 11.0 23.308 32.853 23.308 32.853 ls_scf_init_scf 1 4.0 0.001 0.006 30.408 30.410 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.192 29.263 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.804 26.821 multiply_cannon_multrec 888 9.7 12.589 15.299 21.097 24.350 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.905 22.191 make_m2s 222 7.7 0.006 0.007 16.780 18.008 make_images 222 8.7 1.964 2.279 16.741 17.970 hybrid_alltoall_any 227 10.6 0.620 2.885 9.054 10.535 make_images_data 222 9.7 0.004 0.008 9.474 10.533 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.416 8.030 9.235 dbcsr_mm_accdrv_process_sort 3754 11.4 7.660 8.820 7.660 8.820 mp_sum_l 807 5.4 4.892 8.003 4.892 8.003 multiply_cannon_sync_h2d 888 9.7 6.068 7.286 6.068 7.286 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.548 6.966 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.452 6.824 mp_irecv_dv 2335 11.1 2.436 6.782 2.436 6.782 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.706 6.156 arnoldi_extremal 4 6.8 0.000 0.000 5.213 5.231 arnoldi_normal_ev 4 7.8 0.001 0.004 5.213 5.231 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.969 5.142 build_subspace 16 8.4 0.014 0.021 4.906 4.914 calculate_norms 1584 9.8 4.272 4.597 4.272 4.597 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.518 3.844 mp_allgather_i34 111 8.7 1.403 3.737 1.403 3.737 dbcsr_matrix_vector_mult_local 304 10.0 3.042 3.616 3.044 3.618 ls_scf_post 1 4.0 0.000 0.001 3.525 3.529 ls_scf_store_result 1 5.0 0.000 0.000 3.280 3.348 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.835 2.944 dbcsr_sort_data 325 11.1 1.871 2.169 1.871 2.169 make_images_pack 222 9.7 1.814 2.128 1.817 2.131 dbcsr_data_release 9322 10.9 1.317 2.025 1.317 2.025 make_images_sizes 222 9.7 0.000 0.000 1.001 1.993 mp_alltoall_i44 222 10.7 1.000 1.993 1.000 1.993 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.965 1.967 rebuild_ks_matrix 3 7.3 0.000 0.000 1.947 1.949 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.947 1.949 dbcsr_finalize 304 7.8 0.026 0.032 1.608 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.170000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2167.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 3.342742E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.197 97.689 97.691 qs_energies 1 2.0 0.000 0.000 96.812 96.838 ls_scf 1 3.0 0.000 0.001 95.156 95.180 dbcsr_multiply_generic 111 6.7 0.017 0.024 78.823 79.109 ls_scf_main 1 4.0 0.001 0.014 58.940 58.941 multiply_cannon 111 7.7 0.062 0.124 51.688 56.347 density_matrix_trs4 2 5.0 0.002 0.002 52.922 53.029 multiply_cannon_loop 111 8.7 0.115 0.126 46.728 49.648 ls_scf_init_scf 1 4.0 0.000 0.002 32.969 32.972 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.746 31.823 mp_waitall_1 6438 11.0 23.018 29.428 23.018 29.428 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.195 29.210 multiply_cannon_multrec 1332 9.7 14.176 17.009 22.014 24.877 make_m2s 222 7.7 0.007 0.008 21.338 22.693 make_images 222 8.7 3.129 3.598 21.288 22.645 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.346 17.857 make_images_data 222 9.7 0.004 0.004 11.883 13.423 hybrid_alltoall_any 227 10.6 0.799 3.818 11.260 12.980 dbcsr_mm_accdrv_process 3641 10.4 0.214 0.420 7.475 8.996 dbcsr_mm_accdrv_process_sort 3641 11.4 7.096 8.573 7.096 8.573 mp_sum_l 807 5.4 4.285 8.073 4.285 8.073 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.262 6.511 multiply_cannon_sync_h2d 1332 9.7 5.542 6.257 5.542 6.257 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.078 6.063 mp_irecv_dv 3229 10.9 2.054 5.980 2.054 5.980 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.612 5.413 arnoldi_extremal 4 6.8 0.000 0.000 5.317 5.332 arnoldi_normal_ev 4 7.8 0.001 0.005 5.317 5.332 build_subspace 16 8.4 0.014 0.021 4.971 4.982 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.592 4.768 calculate_norms 2376 9.8 4.197 4.536 4.197 4.536 mp_allgather_i34 111 8.7 2.083 4.220 2.083 4.220 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.579 3.898 dbcsr_matrix_vector_mult_local 304 10.0 3.184 3.712 3.186 3.714 dbcsr_sort_data 658 11.4 3.156 3.487 3.156 3.487 ls_scf_post 1 4.0 0.003 0.021 3.247 3.272 dbcsr_special_finalize 555 9.7 0.006 0.007 2.903 3.243 dbcsr_merge_single_wm 555 10.7 0.540 0.659 2.894 3.235 ls_scf_store_result 1 5.0 0.000 0.000 2.976 3.047 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.894 2.939 dbcsr_data_release 10477 10.7 1.583 2.397 1.583 2.397 dbcsr_finalize 304 7.8 0.049 0.061 1.804 1.993 make_images_pack 222 9.7 1.610 1.969 1.613 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.691000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2716.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 4.622766E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.057 98.100 98.101 qs_energies 1 2.0 0.000 0.000 97.286 97.292 ls_scf 1 3.0 0.000 0.000 95.349 95.355 dbcsr_multiply_generic 111 6.7 0.017 0.018 76.959 77.180 ls_scf_main 1 4.0 0.000 0.000 61.356 61.357 multiply_cannon 111 7.7 0.085 0.142 54.787 59.510 density_matrix_trs4 2 5.0 0.002 0.003 54.342 54.415 multiply_cannon_loop 111 8.7 0.070 0.077 50.202 51.942 mp_waitall_1 5481 11.0 25.708 30.655 25.708 30.655 ls_scf_init_scf 1 4.0 0.000 0.000 30.422 30.423 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.216 29.254 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.016 27.025 multiply_cannon_multrec 444 9.7 14.049 16.640 21.066 22.842 make_m2s 222 7.7 0.004 0.005 17.539 20.083 make_images 222 8.7 3.715 4.423 17.477 20.022 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.712 15.773 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.021 14.401 make_images_data 222 9.7 0.003 0.004 9.771 12.290 hybrid_alltoall_any 227 10.6 0.789 3.762 9.513 12.042 multiply_cannon_sync_h2d 444 9.7 6.584 8.061 6.584 8.061 dbcsr_mm_accdrv_process 3003 10.4 0.190 0.346 6.722 7.851 dbcsr_mm_accdrv_process_sort 3003 11.4 6.405 7.500 6.405 7.500 mp_allgather_i34 111 8.7 2.824 7.025 2.824 7.025 arnoldi_extremal 4 6.8 0.000 0.000 5.905 5.913 arnoldi_normal_ev 4 7.8 0.001 0.004 5.905 5.913 build_subspace 16 8.4 0.015 0.020 5.511 5.525 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.611 4.727 dbcsr_matrix_vector_mult 304 9.0 0.007 0.017 4.160 4.337 dbcsr_matrix_vector_mult_local 304 10.0 3.655 4.107 3.657 4.109 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.632 3.857 mp_irecv_dv 1241 11.2 1.613 3.834 1.613 3.834 mp_sum_l 807 5.4 2.679 3.756 2.679 3.756 calculate_norms 792 9.8 3.549 3.681 3.549 3.681 ls_scf_post 1 4.0 0.000 0.000 3.571 3.577 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.410 3.527 ls_scf_store_result 1 5.0 0.000 0.000 3.348 3.396 make_images_sizes 222 9.7 0.000 0.000 1.123 3.304 mp_alltoall_i44 222 10.7 1.123 3.304 1.123 3.304 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.798 2.805 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.262 dbcsr_merge_all 275 8.9 0.473 0.524 2.046 2.101 dbcsr_data_release 10123 10.8 1.326 2.005 1.326 2.005 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.988 1.988 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.101000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3614.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e6ae6671d4bc88fbd380deaa153891801fbee65d_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 8.716227E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.109 0.184 106.946 106.948 qs_energies 1 2.0 0.000 0.000 105.393 105.408 ls_scf 1 3.0 0.000 0.000 102.517 102.531 dbcsr_multiply_generic 111 6.7 0.024 0.025 76.651 76.761 ls_scf_main 1 4.0 0.000 0.000 65.050 65.051 density_matrix_trs4 2 5.0 0.002 0.003 56.252 56.305 multiply_cannon 111 7.7 0.127 0.222 49.507 51.527 multiply_cannon_loop 111 8.7 0.067 0.070 46.117 46.805 ls_scf_init_scf 1 4.0 0.000 0.000 33.830 33.831 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.506 32.522 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.696 29.709 mp_waitall_1 4569 11.1 21.635 25.143 21.635 25.143 make_m2s 222 7.7 0.005 0.005 23.658 24.546 make_images 222 8.7 4.582 4.972 23.552 24.437 multiply_cannon_multrec 444 9.7 17.947 18.657 22.536 23.151 hybrid_alltoall_any 227 10.6 1.658 3.625 12.785 15.388 make_images_data 222 9.7 0.003 0.004 12.914 15.242 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.128 10.628 multiply_cannon_sync_h2d 444 9.7 8.853 8.894 8.853 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.422 7.439 arnoldi_normal_ev 4 7.8 0.002 0.008 7.422 7.439 build_subspace 16 8.4 0.026 0.036 6.856 6.872 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.446 5.612 dbcsr_matrix_vector_mult_local 304 10.0 4.984 5.286 4.987 5.288 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.955 5.201 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.789 4.883 dbcsr_mm_accdrv_process 1814 10.4 0.184 0.310 4.413 4.538 dbcsr_mm_accdrv_process_sort 1814 11.4 4.116 4.246 4.116 4.246 ls_scf_post 1 4.0 0.000 0.000 3.637 3.651 make_images_sizes 222 9.7 0.000 0.000 1.425 3.461 mp_alltoall_i44 222 10.7 1.425 3.461 1.425 3.461 ls_scf_store_result 1 5.0 0.000 0.000 3.389 3.399 mp_allgather_i34 111 8.7 1.043 3.290 1.043 3.290 calculate_norms 792 9.8 3.239 3.268 3.239 3.268 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.145 dbcsr_merge_all 275 8.9 0.891 0.921 2.860 2.902 dbcsr_complete_redistribute 5 7.6 1.437 1.470 2.767 2.901 dbcsr_data_release 12724 10.6 2.343 2.869 2.343 2.869 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.846 2.846 matrix_ls_to_qs 2 6.0 0.000 0.000 2.422 2.565 dbcsr_sort_data 325 11.1 2.436 2.495 2.436 2.495 dbcsr_new_transposed 4 7.5 0.244 0.254 2.312 2.326 dbcsr_add_d 103 6.2 0.000 0.000 2.157 2.240 dbcsr_add_anytype 103 7.2 0.860 0.892 2.156 2.239 dbcsr_frobenius_norm 74 6.6 2.058 2.136 2.195 2.228 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.179 2.180 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.948000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6874.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e6ae6671d4bc88fbd380deaa153891801fbee65d Summary: empty Status: OK