=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: afe0d05cf1032bb96baeefe089b5bc565e4f53ca ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, 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 (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/01 job id: 45097478 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/02 job id: 45097479 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/03 job id: 45097480 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/04 job id: 45097481 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/05 job id: 45097482 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/06 job id: 45097483 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/07 job id: 45097484 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/08 job id: 45097485 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/09 job id: 45097486 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/10 job id: 45097487 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/11 job id: 45097489 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/12 job id: 45097490 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/13 job id: 45097491 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/14 job id: 45097492 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/15 job id: 45097493 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/16 job id: 45097494 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/17 job id: 45097495 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/18 job id: 45097496 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/19 job id: 45097497 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/20 job id: 45097498 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/21 job id: 45097499 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/22 job id: 45097500 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/23 job id: 45097501 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/24 job id: 45097502 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/25 job id: 45097503 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/26 job id: 45097504 --- 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/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 0.024 0.034 138.353 138.355 farming_run 1 2.0 136.438 136.440 138.322 138.326 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.473413E+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 231 1098811. MP_Allreduce 491 2254389. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 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 0.014 0.035 115.940 115.941 qs_energies 1 2.0 0.000 0.000 115.651 115.652 mp2_main 1 3.0 0.000 0.000 113.404 113.405 mp2_gpw_main 1 4.0 0.019 0.026 112.400 112.402 mp2_ri_gpw_compute_in 1 5.0 0.173 0.183 93.440 93.584 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.478 55.620 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.659 46.668 get_2c_integrals 1 6.0 0.001 0.007 37.339 37.789 integrate_v_rspace 273 8.0 0.437 0.449 25.065 29.730 pw_transfer 6555 10.6 0.373 0.383 27.358 27.722 fft_wrap_pw1pw2 5465 11.4 0.045 0.046 26.012 26.391 grid_integrate_task_list 273 9.0 20.885 25.974 20.885 25.974 fft_wrap_pw1pw2_100 2178 12.4 1.186 1.290 23.520 23.940 compute_2c_integrals 1 7.0 0.003 0.003 19.376 19.378 compute_2c_integrals_loop_lm 1 8.0 0.004 0.005 18.876 19.042 mp2_eri_2c_integrate_gpw 1 9.0 2.389 2.436 18.873 19.037 rpa_ri_compute_en 1 5.0 0.001 0.001 18.853 18.980 cp_fm_cholesky_decompose 12 8.2 17.982 18.539 17.982 18.539 cholesky_decomp 1 7.0 0.000 0.000 16.799 17.258 fft3d_s 5443 13.4 16.123 16.458 16.145 16.478 ao_to_mo_and_store_B_mult_1 272 7.0 10.843 15.538 10.843 15.538 calculate_wavefunction 272 8.0 5.411 5.486 12.514 13.140 rpa_num_int 1 6.0 0.000 0.001 10.716 10.725 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.587 10.711 calc_mat_Q 8 8.0 0.000 0.000 9.360 9.462 contract_S_to_Q 8 9.0 0.000 0.000 8.783 8.886 calc_potential_gpw 544 9.5 0.005 0.005 8.258 8.664 parallel_gemm_fm 14 9.1 0.000 0.000 8.390 8.463 parallel_gemm_fm_cosma 14 10.1 8.390 8.463 8.390 8.463 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.210 8.460 potential_pw2rs 545 10.0 0.107 0.109 7.673 8.390 create_integ_mat 1 6.0 0.014 0.026 7.897 7.907 collocate_single_gaussian 272 10.0 0.040 0.043 7.454 7.665 array2fm 1 7.0 0.000 0.000 6.743 7.268 pw_scatter_s 2720 13.7 4.433 4.624 4.433 4.624 pw_gather_s 2722 13.2 3.848 4.164 3.848 4.164 array2fm_buffer_send 1 8.0 2.910 3.176 2.910 3.176 ao_to_mo_and_store_B_E_Ex_1 272 7.0 1.249 1.393 2.200 2.386 pw_poisson_solve 545 10.5 1.119 1.170 2.173 2.342 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.400067, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.038 404.068 404.070 farming_run 1 2.0 401.735 401.742 404.025 404.029 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.227489E+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 744 385880. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.051 211.629 211.629 qs_energies 1 2.0 0.091 0.363 211.200 211.212 scf_env_do_scf 1 3.0 0.000 0.000 107.260 107.260 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.152 106.160 rebuild_ks_matrix 4 6.0 0.000 0.000 106.150 106.158 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.066 106.150 106.158 hfx_ks_matrix 4 8.0 0.001 0.001 105.730 105.733 integrate_four_center 4 9.0 0.144 0.454 105.729 105.733 mp2_main 1 3.0 0.009 0.047 103.506 103.602 mp2_gpw_main 1 4.0 0.033 0.051 101.999 102.010 integrate_four_center_main 4 10.0 0.120 0.569 96.640 100.119 integrate_four_center_bin 268 11.0 96.520 100.026 96.520 100.026 init_scf_loop 1 4.0 0.000 0.000 92.921 92.921 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.878 76.092 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.580 55.795 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.136 47.212 integrate_v_rspace 95 8.0 0.401 0.573 28.552 33.409 pw_transfer 2240 10.6 0.145 0.165 29.935 30.406 ao_to_mo_and_store_B_mult_1 91 7.0 10.757 29.739 10.757 29.739 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.959 29.459 grid_integrate_task_list 95 9.0 23.803 28.892 23.803 28.892 mp2_ri_gpw_compute_en 1 5.0 0.055 0.064 26.942 28.847 fft_wrap_pw1pw2_100 730 12.4 1.256 1.414 26.634 27.173 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.829 1.887 25.047 25.058 get_2c_integrals 1 6.0 0.000 0.000 20.208 20.233 compute_2c_integrals 1 7.0 0.005 0.015 19.188 19.194 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.793 18.998 mp2_eri_2c_integrate_gpw 1 9.0 1.734 1.810 18.792 18.998 fft3d_s 1823 13.4 18.427 18.830 18.440 18.843 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.338 14.338 calculate_wavefunction 91 8.0 2.009 2.039 9.732 9.946 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.579 8.839 9.423 potential_pw2rs 186 10.0 0.033 0.035 8.613 9.359 local_gemm 172 8.0 8.285 8.852 8.285 8.852 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.208 8.628 mp2_ri_gpw_compute_en_comm 22 7.0 0.511 0.531 7.998 8.347 calc_potential_gpw 182 9.5 0.002 0.002 7.937 8.192 collocate_single_gaussian 91 10.0 0.017 0.024 7.847 8.106 mp_sync 37 10.5 3.719 7.117 3.719 7.117 mp2_ri_gpw_compute_en_ener 172 7.0 6.337 6.447 6.337 6.447 mp_sendrecv_dm3 2068 8.0 6.007 6.349 6.007 6.349 pw_gather_s 912 13.2 4.892 5.503 4.892 5.503 pw_scatter_s 910 13.7 3.950 4.251 3.950 4.251 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.006232, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 454.451200E+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 272. MP_Sync 530 MP_Alltoall 2083 589622. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.108 60.451 60.453 qs_mol_dyn_low 1 2.0 0.004 0.008 59.703 59.715 qs_forces 11 3.9 0.008 0.049 59.259 59.267 qs_energies 11 4.9 0.003 0.011 57.588 57.601 scf_env_do_scf 11 5.9 0.001 0.002 50.515 50.515 scf_env_do_scf_inner_loop 108 6.5 0.046 0.181 47.942 47.943 qs_scf_new_mos 108 7.5 0.000 0.001 35.634 35.956 qs_scf_loop_do_ot 108 8.5 0.001 0.001 35.634 35.956 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.911 35.444 ot_scf_mini 108 9.5 0.002 0.004 33.805 33.982 velocity_verlet 10 3.0 0.001 0.003 28.441 28.448 multiply_cannon 2286 13.5 0.185 0.192 26.214 27.861 multiply_cannon_loop 2286 14.5 1.438 1.538 25.259 26.961 ot_mini 108 10.5 0.001 0.004 20.089 20.340 qs_ot_get_derivative 108 11.5 0.001 0.002 17.113 17.299 mp_waitall_1 245248 16.5 9.204 15.500 9.204 15.500 multiply_cannon_metrocomm3 54864 15.5 0.070 0.074 5.956 12.887 multiply_cannon_multrec 54864 15.5 4.301 6.748 7.478 10.907 rebuild_ks_matrix 119 8.3 0.000 0.000 9.057 9.211 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.018 9.057 9.211 qs_ot_get_p 119 10.4 0.001 0.001 8.780 9.088 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.987 8.121 mp_sum_l 7207 12.9 5.882 7.780 5.882 7.780 multiply_cannon_sync_h2d 54864 15.5 5.943 7.180 5.943 7.180 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.144 6.249 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.679 6.115 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.597 5.767 calculate_rho_elec 119 8.7 0.012 0.017 5.596 5.766 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.668 5.751 init_scf_run 11 5.9 0.000 0.001 5.709 5.709 scf_env_initial_rho_setup 11 6.9 0.001 0.004 5.709 5.709 sum_up_and_integrate 119 10.3 0.012 0.016 5.437 5.443 integrate_v_rspace 119 11.3 0.002 0.003 5.425 5.433 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.686 4.687 dbcsr_mm_accdrv_process 76910 16.1 1.078 1.832 3.098 4.462 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.324 4.326 cp_fm_redistribute_end 50 14.0 2.218 4.291 2.224 4.295 cp_fm_diag_elpa_base 50 14.0 2.063 4.152 2.067 4.160 rs_pw_transfer 974 11.9 0.012 0.013 3.739 3.874 density_rs2pw 119 9.7 0.004 0.005 3.263 3.395 calculate_dm_sparse 119 9.5 0.000 0.001 3.174 3.360 calculate_first_density_matrix 1 7.0 0.001 0.004 3.165 3.170 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.938 3.139 apply_single 119 13.6 0.000 0.000 2.938 3.139 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.956 3.138 pw_transfer 1439 11.6 0.053 0.059 2.790 2.891 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.719 2.823 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.714 2.819 ot_diis_step 108 11.5 0.006 0.010 2.711 2.713 potential_pw2rs 119 12.3 0.004 0.004 2.664 2.704 make_m2s 4572 13.5 0.054 0.056 2.537 2.628 fft3d_ps 1201 14.6 0.373 0.480 2.485 2.586 init_scf_loop 11 6.9 0.001 0.005 2.543 2.547 make_images 4572 14.5 0.134 0.139 2.454 2.543 jit_kernel_multiply 13 15.8 1.962 2.540 1.962 2.540 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.485 2.487 wfi_extrapolate 11 7.9 0.001 0.003 2.436 2.436 mp_alltoall_d11v 2130 13.8 2.033 2.305 2.033 2.305 acc_transpose_blocks 54864 15.5 0.218 0.252 1.762 2.283 fft_wrap_pw1pw2_140 487 13.2 0.080 0.095 2.057 2.162 grid_integrate_task_list 119 12.3 2.012 2.100 2.012 2.100 mp_sum_d 4133 12.0 1.432 2.099 1.432 2.099 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 2.036 2.098 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.521 1.547 mp_waitany 12084 13.8 1.372 1.542 1.372 1.542 mp_alltoall_z22v 1201 16.6 1.387 1.472 1.387 1.472 parallel_gemm_fm 81 9.0 0.000 0.000 1.415 1.426 parallel_gemm_fm_cosma 81 10.0 1.415 1.425 1.415 1.425 make_images_sizes 4572 15.5 0.004 0.005 1.114 1.421 mp_alltoall_i44 4572 16.5 1.109 1.417 1.109 1.417 grid_collocate_task_list 119 9.7 1.292 1.374 1.292 1.374 prepare_preconditioner 11 7.9 0.000 0.001 1.226 1.261 make_preconditioner 11 8.9 0.001 0.003 1.226 1.261 rs_scatter_matrices 130 9.7 0.141 0.170 1.047 1.217 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.453000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=433.727273, yerr=0.616575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.173568E+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 54 MP_Alltoall 2060 1523321. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.029 42.726 42.726 qs_mol_dyn_low 1 2.0 0.003 0.003 42.543 42.551 qs_forces 11 3.9 0.002 0.002 42.478 42.480 qs_energies 11 4.9 0.001 0.001 40.710 40.715 scf_env_do_scf 11 5.9 0.000 0.001 34.460 34.461 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 31.652 31.652 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.972 23.334 qs_scf_new_mos 108 7.5 0.001 0.001 21.377 21.624 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.376 21.623 ot_scf_mini 108 9.5 0.002 0.003 20.455 20.627 velocity_verlet 10 3.0 0.001 0.001 19.894 19.896 multiply_cannon 2286 13.5 0.209 0.218 17.486 19.024 multiply_cannon_loop 2286 14.5 0.897 0.971 16.254 17.703 ot_mini 108 10.5 0.001 0.001 12.316 12.550 mp_waitall_1 200699 16.5 6.811 11.940 6.811 11.940 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.618 10.113 qs_ot_get_derivative 108 11.5 0.001 0.001 9.822 9.999 multiply_cannon_multrec 27432 15.5 1.980 4.287 5.870 8.803 rebuild_ks_matrix 119 8.3 0.000 0.000 7.922 8.060 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.921 8.059 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.027 7.152 dbcsr_mm_accdrv_process 47894 16.0 2.947 5.030 3.821 5.878 qs_ot_get_p 119 10.4 0.001 0.001 5.142 5.372 init_scf_run 11 5.9 0.000 0.001 4.985 4.985 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.985 4.985 sum_up_and_integrate 119 10.3 0.024 0.028 4.810 4.828 integrate_v_rspace 119 11.3 0.002 0.003 4.786 4.801 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.861 4.640 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.346 4.379 calculate_rho_elec 119 8.7 0.021 0.024 4.346 4.378 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.000 4.150 apply_single 119 13.6 0.000 0.000 3.000 4.150 mp_sum_l 7207 12.9 2.078 4.081 2.078 4.081 rs_pw_transfer 974 11.9 0.010 0.011 3.359 3.822 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.487 3.505 calculate_first_density_matrix 1 7.0 0.000 0.000 3.288 3.290 make_m2s 4572 13.5 0.052 0.054 2.956 3.192 make_images 4572 14.5 0.201 0.237 2.868 3.102 density_rs2pw 119 9.7 0.004 0.004 2.641 3.093 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.030 3.031 init_scf_loop 11 6.9 0.000 0.000 2.780 2.780 multiply_cannon_sync_h2d 27432 15.5 2.203 2.776 2.203 2.776 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.601 2.604 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.567 2.568 cp_fm_redistribute_end 50 14.0 1.297 2.521 1.307 2.526 potential_pw2rs 119 12.3 0.006 0.006 2.487 2.513 calculate_dm_sparse 119 9.5 0.000 0.001 2.405 2.484 ot_diis_step 108 11.5 0.010 0.011 2.443 2.445 cp_fm_diag_elpa_base 50 14.0 1.181 2.387 1.212 2.441 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.338 2.429 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.340 2.378 pw_transfer 1439 11.6 0.065 0.073 2.310 2.356 acc_transpose_blocks 27432 15.5 0.110 0.115 1.629 2.315 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.218 2.267 fft3d_ps 1201 14.6 0.517 0.568 1.920 1.966 grid_integrate_task_list 119 12.3 1.827 1.923 1.827 1.923 jit_kernel_multiply 10 16.1 0.822 1.910 0.822 1.910 prepare_preconditioner 11 7.9 0.000 0.000 1.782 1.809 make_preconditioner 11 8.9 0.000 0.000 1.782 1.809 make_images_data 4572 15.5 0.045 0.051 1.279 1.784 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.659 1.721 acc_transpose_blocks_kernels 27432 16.5 0.182 0.274 1.096 1.686 fft_wrap_pw1pw2_140 487 13.2 0.078 0.085 1.615 1.657 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.114 1.648 wfi_extrapolate 11 7.9 0.001 0.001 1.644 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.601 1.609 mp_alltoall_d11v 2130 13.8 1.452 1.558 1.452 1.558 mp_allgather_i34 2286 14.5 0.673 1.462 0.673 1.462 jit_kernel_transpose 5 15.5 0.915 1.411 0.915 1.411 grid_collocate_task_list 119 9.7 1.231 1.367 1.231 1.367 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.233 1.283 make_images_sizes 4572 15.5 0.005 0.005 0.862 1.188 mp_alltoall_i44 4572 16.5 0.857 1.183 0.857 1.183 multiply_cannon_metrocomm1 27432 15.5 0.032 0.036 0.462 1.076 mp_alltoall_z22v 1201 16.6 0.918 1.061 0.918 1.061 mp_sum_d 4133 12.0 0.604 1.054 0.604 1.054 mp_waitany 5720 13.7 0.534 0.988 0.534 0.988 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.147 0.546 0.984 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.955 0.968 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.956 0.959 rs_pw_transfer_PW2RS_50 119 14.3 0.588 0.609 0.880 0.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=42.726000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.388032E+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 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 1172916. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 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 0.076 0.136 36.942 36.943 qs_mol_dyn_low 1 2.0 0.003 0.003 35.926 35.932 qs_forces 11 3.9 0.009 0.031 35.863 35.864 qs_energies 11 4.9 0.004 0.026 34.128 34.140 scf_env_do_scf 11 5.9 0.001 0.001 28.983 28.983 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.946 25.948 dbcsr_multiply_generic 2286 12.5 0.096 0.100 17.627 17.852 velocity_verlet 10 3.0 0.001 0.001 17.140 17.141 qs_scf_new_mos 108 7.5 0.001 0.001 16.740 16.756 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.740 16.756 ot_scf_mini 108 9.5 0.002 0.003 15.949 15.963 multiply_cannon 2286 13.5 0.197 0.203 13.480 14.452 multiply_cannon_loop 2286 14.5 0.633 0.667 12.565 13.641 ot_mini 108 10.5 0.001 0.001 9.690 9.720 qs_ot_get_derivative 108 11.5 0.001 0.002 8.099 8.114 multiply_cannon_multrec 18288 15.5 1.912 2.864 6.754 7.120 rebuild_ks_matrix 119 8.3 0.000 0.000 7.067 7.113 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.067 7.112 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.213 6.253 dbcsr_mm_accdrv_process 38222 16.0 4.079 5.323 4.760 5.545 mp_waitall_1 158411 16.6 3.485 4.764 3.485 4.764 sum_up_and_integrate 119 10.3 0.031 0.033 4.497 4.505 integrate_v_rspace 119 11.3 0.003 0.003 4.465 4.478 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.027 4.046 calculate_rho_elec 119 8.7 0.031 0.031 4.026 4.045 qs_ot_get_p 119 10.4 0.001 0.001 3.960 3.989 init_scf_run 11 5.9 0.000 0.001 3.764 3.764 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.764 3.764 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.001 3.721 rs_pw_transfer 974 11.9 0.010 0.010 2.922 3.162 init_scf_loop 11 6.9 0.001 0.004 3.006 3.012 multiply_cannon_metrocomm3 18288 15.5 0.046 0.046 1.688 2.771 density_rs2pw 119 9.7 0.004 0.005 2.504 2.743 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.629 2.640 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.170 2.563 apply_single 119 13.6 0.000 0.000 2.170 2.563 calculate_first_density_matrix 1 7.0 0.000 0.003 2.380 2.381 pw_transfer 1439 11.6 0.066 0.071 2.348 2.375 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.339 2.340 make_m2s 4572 13.5 0.045 0.045 2.183 2.314 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 2.254 2.285 make_images 4572 14.5 0.191 0.201 2.097 2.228 potential_pw2rs 119 12.3 0.007 0.008 2.209 2.226 mp_sum_l 7207 12.9 1.598 2.152 1.598 2.152 prepare_preconditioner 11 7.9 0.000 0.001 2.110 2.112 make_preconditioner 11 8.9 0.000 0.003 2.109 2.112 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.027 2.038 cp_fm_diag_elpa_base 50 14.0 1.999 2.017 2.025 2.035 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.937 2.024 fft3d_ps 1201 14.6 0.530 0.553 1.934 1.966 calculate_dm_sparse 119 9.5 0.000 0.000 1.911 1.925 grid_integrate_task_list 119 12.3 1.788 1.904 1.788 1.904 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.868 1.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.776 1.778 fft_wrap_pw1pw2_140 487 13.2 0.090 0.095 1.682 1.708 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.595 1.603 ot_diis_step 108 11.5 0.011 0.011 1.553 1.553 jit_kernel_multiply 8 16.3 0.630 1.542 0.630 1.542 multiply_cannon_sync_h2d 18288 15.5 1.322 1.515 1.322 1.515 acc_transpose_blocks 18288 15.5 0.076 0.078 1.262 1.420 grid_collocate_task_list 119 9.7 1.212 1.371 1.212 1.371 wfi_extrapolate 11 7.9 0.001 0.001 1.317 1.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.227 1.235 make_images_data 4572 15.5 0.045 0.048 1.012 1.176 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.504 1.135 mp_alltoall_z22v 1201 16.6 0.959 1.093 0.959 1.093 hybrid_alltoall_any 4725 16.4 0.056 0.114 0.865 1.064 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.975 1.011 cp_fm_cholesky_invert 11 10.9 0.998 1.005 0.998 1.005 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.995 1.003 mp_alltoall_d11v 2130 13.8 0.878 0.973 0.878 0.973 acc_transpose_blocks_kernels 18288 16.5 0.211 0.221 0.820 0.964 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.931 0.937 mp_waitany 9880 13.7 0.577 0.843 0.577 0.843 mp_allgather_i34 2286 14.5 0.373 0.820 0.373 0.820 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.125 0.559 0.802 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.751 jit_kernel_transpose 5 15.6 0.609 0.743 0.609 0.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=36.943000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.363636, yerr=1.666391 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 556.904448E+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 3622 63496. MP_Allreduce 10074 349. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 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 0.013 0.030 38.740 38.741 qs_mol_dyn_low 1 2.0 0.003 0.003 38.481 38.527 qs_forces 11 3.9 0.002 0.003 38.299 38.300 qs_energies 11 4.9 0.001 0.001 36.493 36.498 scf_env_do_scf 11 5.9 0.000 0.001 30.923 30.924 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.076 27.076 velocity_verlet 10 3.0 0.001 0.001 19.529 19.531 dbcsr_multiply_generic 2286 12.5 0.098 0.101 19.122 19.239 qs_scf_new_mos 108 7.5 0.001 0.001 17.551 17.608 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.551 17.607 ot_scf_mini 108 9.5 0.002 0.003 16.585 16.631 multiply_cannon 2286 13.5 0.228 0.267 14.925 15.473 multiply_cannon_loop 2286 14.5 0.936 0.971 13.833 14.264 ot_mini 108 10.5 0.001 0.001 9.830 9.895 multiply_cannon_multrec 27432 15.5 2.326 3.015 9.020 9.426 qs_ot_get_derivative 108 11.5 0.001 0.002 7.924 7.970 dbcsr_mm_accdrv_process 47916 15.9 5.516 7.439 6.600 7.830 rebuild_ks_matrix 119 8.3 0.000 0.000 7.218 7.275 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.218 7.274 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.422 6.469 sum_up_and_integrate 119 10.3 0.036 0.039 4.316 4.324 integrate_v_rspace 119 11.3 0.003 0.003 4.280 4.289 init_scf_run 11 5.9 0.000 0.001 4.213 4.213 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.213 4.213 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.149 4.178 calculate_rho_elec 119 8.7 0.040 0.046 4.148 4.177 qs_ot_get_p 119 10.4 0.001 0.001 4.050 4.127 init_scf_loop 11 6.9 0.000 0.000 3.821 3.822 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.882 3.362 prepare_preconditioner 11 7.9 0.000 0.000 2.870 2.880 make_preconditioner 11 8.9 0.000 0.000 2.870 2.880 rs_pw_transfer 974 11.9 0.009 0.010 2.674 2.867 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.476 2.808 mp_waitall_1 137007 16.6 2.221 2.748 2.221 2.748 calculate_first_density_matrix 1 7.0 0.000 0.000 2.686 2.688 density_rs2pw 119 9.7 0.004 0.004 2.499 2.682 make_m2s 4572 13.5 0.055 0.057 2.542 2.658 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.152 2.618 apply_single 119 13.6 0.000 0.000 2.152 2.617 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.584 2.594 make_images 4572 14.5 0.271 0.332 2.433 2.547 pw_transfer 1439 11.6 0.066 0.071 2.416 2.460 calculate_dm_sparse 119 9.5 0.000 0.000 2.318 2.386 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.323 2.372 jit_kernel_multiply 10 15.7 1.023 2.260 1.023 2.260 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.255 2.255 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.093 2.117 fft3d_ps 1201 14.6 0.565 0.619 1.997 2.036 potential_pw2rs 119 12.3 0.008 0.010 2.018 2.029 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.013 2.014 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.918 1.926 cp_fm_diag_elpa_base 50 14.0 1.877 1.895 1.915 1.923 grid_integrate_task_list 119 12.3 1.839 1.903 1.839 1.903 fft_wrap_pw1pw2_140 487 13.2 0.087 0.095 1.834 1.888 mp_sum_l 7207 12.9 1.229 1.836 1.229 1.836 ot_diis_step 108 11.5 0.012 0.012 1.830 1.830 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.798 1.813 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.477 1.489 acc_transpose_blocks 27432 15.5 0.113 0.116 1.456 1.482 wfi_extrapolate 11 7.9 0.001 0.001 1.467 1.467 grid_collocate_task_list 119 9.7 1.221 1.343 1.221 1.343 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.721 1.207 cp_fm_upper_to_full 72 14.2 0.829 1.185 0.829 1.185 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.159 1.178 dbcsr_complete_redistribute 329 12.2 0.118 0.145 0.894 1.170 make_images_data 4572 15.5 0.045 0.049 1.009 1.154 mp_alltoall_d11v 2130 13.8 0.946 1.102 0.946 1.102 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.086 1.087 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.867 1.083 mp_alltoall_z22v 1201 16.6 1.015 1.059 1.015 1.059 multiply_cannon_sync_h2d 27432 15.5 0.978 1.040 0.978 1.040 cp_fm_cholesky_invert 11 10.9 0.924 0.928 0.924 0.928 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.896 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.611 0.880 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.870 acc_transpose_blocks_kernels 27432 16.5 0.268 0.277 0.829 0.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.741000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.545455, yerr=2.934548 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 604.934144E+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 3622 63495. MP_Allreduce 10074 348. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 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 0.028 32.685 32.686 qs_mol_dyn_low 1 2.0 0.003 0.003 32.436 32.443 qs_forces 11 3.9 0.002 0.003 32.369 32.370 qs_energies 11 4.9 0.001 0.001 30.564 30.567 scf_env_do_scf 11 5.9 0.000 0.001 25.367 25.367 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 22.162 22.163 velocity_verlet 10 3.0 0.001 0.001 16.177 16.180 dbcsr_multiply_generic 2286 12.5 0.091 0.095 13.836 13.961 qs_scf_new_mos 108 7.5 0.001 0.001 13.169 13.190 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.168 13.190 ot_scf_mini 108 9.5 0.002 0.002 12.431 12.453 multiply_cannon 2286 13.5 0.232 0.240 10.645 11.251 multiply_cannon_loop 2286 14.5 0.329 0.342 9.560 9.850 rebuild_ks_matrix 119 8.3 0.000 0.000 6.994 7.027 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.994 7.027 ot_mini 108 10.5 0.001 0.001 6.877 6.908 multiply_cannon_multrec 9144 15.5 1.872 2.128 6.253 6.555 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.279 6.309 qs_ot_get_derivative 108 11.5 0.001 0.002 5.444 5.465 dbcsr_mm_accdrv_process 12550 15.8 3.345 4.362 4.280 4.402 sum_up_and_integrate 119 10.3 0.042 0.055 4.199 4.230 integrate_v_rspace 119 11.3 0.003 0.003 4.157 4.192 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.009 4.019 calculate_rho_elec 119 8.7 0.060 0.061 4.008 4.019 init_scf_run 11 5.9 0.000 0.001 3.711 3.711 scf_env_initial_rho_setup 11 6.9 0.017 0.021 3.711 3.711 qs_ot_get_p 119 10.4 0.001 0.001 3.624 3.664 init_scf_loop 11 6.9 0.000 0.000 3.176 3.177 mp_waitall_1 115863 16.7 2.230 2.871 2.230 2.871 pw_transfer 1439 11.6 0.066 0.069 2.462 2.498 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.458 2.460 make_m2s 4572 13.5 0.034 0.036 2.156 2.452 calculate_first_density_matrix 1 7.0 0.000 0.000 2.403 2.420 density_rs2pw 119 9.7 0.004 0.004 2.285 2.419 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.367 2.404 rs_pw_transfer 974 11.9 0.008 0.008 2.249 2.365 make_images 4572 14.5 0.268 0.300 2.066 2.362 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.235 2.235 fft3d_ps 1201 14.6 0.569 0.580 2.026 2.061 jit_kernel_multiply 10 15.6 0.897 2.047 0.897 2.047 prepare_preconditioner 11 7.9 0.000 0.000 2.029 2.034 make_preconditioner 11 8.9 0.000 0.000 2.029 2.034 calculate_dm_sparse 119 9.5 0.000 0.000 1.940 1.963 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.946 1.956 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.911 1.942 grid_integrate_task_list 119 12.3 1.852 1.936 1.852 1.936 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.875 1.912 potential_pw2rs 119 12.3 0.010 0.010 1.865 1.894 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.865 1.873 cp_fm_diag_elpa_base 50 14.0 1.835 1.852 1.863 1.871 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.733 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.522 1.542 grid_collocate_task_list 119 9.7 1.280 1.411 1.280 1.411 ot_diis_step 108 11.5 0.013 0.013 1.406 1.406 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.388 1.403 make_images_data 4572 15.5 0.039 0.042 1.025 1.388 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.961 1.363 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.317 1.322 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.249 1.279 apply_single 119 13.6 0.000 0.000 1.248 1.278 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.233 1.234 wfi_extrapolate 11 7.9 0.001 0.001 1.233 1.233 acc_transpose_blocks 9144 15.5 0.038 0.040 1.090 1.099 mp_alltoall_d11v 2130 13.8 1.001 1.095 1.001 1.095 cp_fm_cholesky_invert 11 10.9 1.052 1.055 1.052 1.055 mp_alltoall_z22v 1201 16.6 0.905 0.952 0.905 0.952 mp_allgather_i34 2286 14.5 0.415 0.950 0.415 0.950 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.922 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.850 0.856 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.407 0.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.841 0.844 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.813 0.826 mp_sum_l 7207 12.9 0.614 0.784 0.614 0.784 multiply_cannon_sync_h2d 9144 15.5 0.710 0.768 0.710 0.768 yz_to_x 606 15.1 0.263 0.275 0.731 0.757 jit_kernel_transpose 5 15.6 0.733 0.737 0.733 0.737 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.686 0.735 x_to_yz 595 16.2 0.277 0.289 0.714 0.719 mp_sum_d 4131 12.0 0.409 0.699 0.409 0.699 dbcsr_complete_redistribute 329 12.2 0.163 0.173 0.655 0.696 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.496 0.673 make_images_sizes 4572 15.5 0.005 0.005 0.467 0.664 mp_alltoall_i44 4572 16.5 0.463 0.659 0.463 0.659 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.686000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=574.272727, yerr=3.518546 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 756.203520E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.035 44.595 44.596 qs_mol_dyn_low 1 2.0 0.003 0.004 44.340 44.347 qs_forces 11 3.9 0.002 0.002 44.275 44.277 qs_energies 11 4.9 0.001 0.001 42.239 42.242 scf_env_do_scf 11 5.9 0.001 0.001 35.815 35.815 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.380 27.381 velocity_verlet 10 3.0 0.001 0.001 24.783 24.789 dbcsr_multiply_generic 2286 12.5 0.098 0.100 18.624 18.917 qs_scf_new_mos 108 7.5 0.001 0.001 16.956 17.056 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.955 17.055 ot_scf_mini 108 9.5 0.002 0.002 15.851 15.954 multiply_cannon 2286 13.5 0.305 0.314 14.376 15.363 multiply_cannon_loop 2286 14.5 0.345 0.359 13.123 14.136 ot_mini 108 10.5 0.001 0.001 9.337 9.464 multiply_cannon_multrec 9144 15.5 3.387 4.734 9.035 9.360 init_scf_loop 11 6.9 0.000 0.000 8.403 8.404 rebuild_ks_matrix 119 8.3 0.000 0.000 7.691 7.853 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.691 7.853 prepare_preconditioner 11 7.9 0.000 0.000 7.376 7.391 make_preconditioner 11 8.9 0.000 0.000 7.376 7.391 qs_ot_get_derivative 108 11.5 0.001 0.002 7.220 7.325 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.836 7.265 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.962 7.109 dbcsr_mm_accdrv_process 12550 15.8 4.410 5.977 5.525 7.095 cp_fm_upper_to_full 72 14.2 3.269 4.756 3.269 4.756 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.596 4.604 calculate_rho_elec 119 8.7 0.118 0.122 4.596 4.603 init_scf_run 11 5.9 0.000 0.001 4.356 4.356 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.356 4.356 sum_up_and_integrate 119 10.3 0.066 0.068 4.300 4.309 integrate_v_rspace 119 11.3 0.004 0.004 4.234 4.244 qs_ot_get_p 119 10.4 0.001 0.001 3.712 3.860 mp_waitall_1 94719 16.7 2.897 3.725 2.897 3.725 pw_transfer 1439 11.6 0.069 0.069 3.059 3.065 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.150 3.030 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.619 2.990 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.960 2.967 calculate_first_density_matrix 1 7.0 0.000 0.000 2.793 2.794 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.795 2.662 make_m2s 4572 13.5 0.038 0.039 2.498 2.631 fft3d_ps 1201 14.6 0.600 0.608 2.582 2.587 calculate_dm_sparse 119 9.5 0.000 0.000 2.498 2.518 make_images 4572 14.5 0.356 0.384 2.377 2.510 density_rs2pw 119 9.7 0.004 0.004 2.465 2.501 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.178 2.461 apply_single 119 13.6 0.000 0.000 2.178 2.461 fft_wrap_pw1pw2_140 487 13.2 0.096 0.097 2.449 2.460 mp_alltoall_i22 627 13.8 1.535 2.455 1.535 2.455 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.534 2.391 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.420 2.331 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.300 2.302 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.231 2.232 jit_kernel_multiply 8 15.2 1.088 2.124 1.088 2.124 grid_integrate_task_list 119 12.3 2.082 2.101 2.082 2.101 ot_diis_step 108 11.5 0.014 0.015 2.075 2.076 mp_sum_l 7207 12.9 1.208 2.023 1.208 2.023 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.013 2.014 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.944 1.999 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.855 1.903 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.784 1.785 rs_pw_transfer 974 11.9 0.009 0.010 1.709 1.740 cp_fm_cholesky_invert 11 10.9 1.710 1.714 1.710 1.714 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.683 1.683 cp_fm_diag_elpa_base 50 14.0 1.526 1.584 1.680 1.681 potential_pw2rs 119 12.3 0.014 0.014 1.634 1.638 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.548 1.557 grid_collocate_task_list 119 9.7 1.507 1.535 1.507 1.535 wfi_extrapolate 11 7.9 0.001 0.001 1.492 1.492 hybrid_alltoall_any 4725 16.4 0.088 0.147 1.231 1.473 make_images_data 4572 15.5 0.043 0.046 1.194 1.410 mp_alltoall_d11v 2130 13.8 1.211 1.229 1.211 1.229 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.179 1.208 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.121 1.145 mp_alltoall_z22v 1201 16.6 1.030 1.074 1.030 1.074 yz_to_x 606 15.1 0.460 0.470 1.051 1.064 multiply_cannon_sync_h2d 9144 15.5 1.041 1.045 1.041 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.982 1.038 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.945 0.960 qs_create_task_list 11 7.9 0.000 0.000 0.947 0.958 generate_qs_task_list 11 8.9 0.369 0.388 0.946 0.958 x_to_yz 595 16.2 0.480 0.488 0.919 0.922 acc_transpose_blocks 9144 15.5 0.039 0.039 0.908 0.912 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.596000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=710.818182, yerr=13.394831 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 502.611968E+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 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 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 0.021 0.047 86.027 86.030 qs_mol_dyn_low 1 2.0 0.003 0.006 85.676 85.685 qs_forces 11 3.9 0.003 0.006 85.504 85.505 qs_energies 11 4.9 0.002 0.007 82.606 82.617 scf_env_do_scf 11 5.9 0.001 0.003 72.581 72.584 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 66.737 66.738 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.774 52.008 qs_scf_new_mos 99 7.5 0.000 0.001 48.232 48.367 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.231 48.366 ot_scf_mini 99 9.5 0.002 0.003 45.808 45.912 velocity_verlet 10 3.0 0.001 0.002 44.095 44.097 multiply_cannon 2055 13.4 0.183 0.188 42.755 43.553 multiply_cannon_loop 2055 14.4 1.493 1.531 41.710 42.524 ot_mini 99 10.5 0.001 0.002 26.775 26.872 qs_ot_get_derivative 99 11.5 0.001 0.003 20.003 20.118 multiply_cannon_multrec 49320 15.4 12.419 13.160 17.438 18.190 rebuild_ks_matrix 110 8.3 0.000 0.000 15.123 15.214 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.123 15.214 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.299 13.381 mp_waitall_1 220248 16.4 11.312 12.383 11.312 12.383 multiply_cannon_sync_h2d 49320 15.4 10.266 10.913 10.266 10.913 qs_ot_get_p 110 10.4 0.001 0.001 10.456 10.584 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.484 8.016 init_scf_run 11 5.9 0.000 0.001 7.818 7.819 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.818 7.819 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.583 7.761 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 7.193 7.760 apply_single 110 13.6 0.000 0.001 7.193 7.760 sum_up_and_integrate 110 10.3 0.037 0.044 7.513 7.530 integrate_v_rspace 110 11.3 0.003 0.004 7.476 7.502 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.335 7.460 calculate_rho_elec 110 8.6 0.021 0.026 7.335 7.460 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.210 7.244 ot_diis_step 99 11.5 0.006 0.008 6.573 6.574 cp_dbcsr_syevd 48 12.0 0.003 0.005 6.168 6.168 init_scf_loop 11 6.9 0.001 0.004 5.816 5.817 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.724 5.784 cp_fm_diag_elpa 48 13.0 0.000 0.001 5.486 5.512 cp_fm_diag_elpa_base 48 14.0 5.468 5.495 5.482 5.508 dbcsr_mm_accdrv_process 87628 16.1 1.901 1.994 4.894 5.343 mp_sum_l 6514 12.8 4.319 5.290 4.319 5.290 rs_pw_transfer 902 11.9 0.012 0.014 4.174 4.775 density_rs2pw 110 9.6 0.005 0.006 4.110 4.705 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.376 4.380 make_m2s 4110 13.4 0.061 0.066 4.252 4.364 make_images 4110 14.4 0.178 0.192 4.155 4.270 wfi_extrapolate 11 7.9 0.001 0.002 4.089 4.089 calculate_dm_sparse 110 9.5 0.000 0.001 3.793 3.930 pw_transfer 1331 11.6 0.055 0.066 3.841 3.919 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.753 3.834 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 3.628 3.680 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.576 3.669 calculate_first_density_matrix 1 7.0 0.000 0.003 3.615 3.619 prepare_preconditioner 11 7.9 0.000 0.001 3.580 3.601 make_preconditioner 11 8.9 0.000 0.002 3.580 3.601 grid_integrate_task_list 110 12.3 3.250 3.421 3.250 3.421 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.355 3.400 fft3d_ps 1111 14.6 0.779 0.882 3.276 3.353 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.299 3.341 fft_wrap_pw1pw2_140 451 13.1 0.175 0.227 2.926 3.011 potential_pw2rs 110 12.3 0.006 0.007 2.972 3.007 jit_kernel_multiply 13 15.9 2.717 2.909 2.717 2.909 mp_alltoall_d11v 2046 13.8 2.264 2.676 2.264 2.676 acc_transpose_blocks 49320 15.4 0.211 0.218 2.256 2.495 mp_waitany 14300 13.8 1.880 2.480 1.880 2.480 grid_collocate_task_list 110 9.6 2.081 2.282 2.081 2.282 mp_alltoall_z22v 1111 16.6 1.801 2.099 1.801 2.099 make_images_data 4110 15.4 0.042 0.045 1.905 2.059 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.967 1.982 hybrid_alltoall_any 4261 16.3 0.082 0.471 1.686 1.929 cp_fm_cholesky_invert 11 10.9 1.915 1.920 1.915 1.920 mp_sum_d 3887 11.9 1.389 1.913 1.389 1.913 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=86.030000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=2.339033 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.718080E+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 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2937936. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 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 0.056 0.400 75.722 76.019 qs_mol_dyn_low 1 2.0 0.003 0.004 74.803 74.838 qs_forces 11 3.9 0.003 0.004 74.649 74.650 qs_energies 11 4.9 0.001 0.001 71.225 71.229 scf_env_do_scf 11 5.9 0.000 0.001 62.307 62.310 scf_env_do_scf_inner_loop 99 6.5 0.015 0.228 53.655 53.656 dbcsr_multiply_generic 2055 12.4 0.116 0.120 39.007 39.240 velocity_verlet 10 3.0 0.001 0.001 39.075 39.076 qs_scf_new_mos 99 7.5 0.001 0.001 35.872 36.011 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.871 36.011 ot_scf_mini 99 9.5 0.003 0.003 34.142 34.273 multiply_cannon 2055 13.4 0.222 0.247 31.483 33.037 multiply_cannon_loop 2055 14.4 0.924 0.955 29.895 31.018 ot_mini 99 10.5 0.001 0.001 19.328 19.472 multiply_cannon_multrec 24660 15.4 7.558 9.309 13.770 15.457 rebuild_ks_matrix 110 8.3 0.000 0.000 14.589 14.717 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.588 14.717 qs_ot_get_derivative 99 11.5 0.001 0.002 13.401 13.542 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.884 12.998 mp_waitall_1 176588 16.5 8.413 11.623 8.413 11.623 init_scf_loop 11 6.9 0.000 0.000 8.602 8.603 multiply_cannon_metrocomm3 24660 15.4 0.070 0.073 5.433 8.601 qs_ot_get_p 110 10.4 0.001 0.001 8.011 8.190 multiply_cannon_sync_h2d 24660 15.4 6.958 7.751 6.958 7.751 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.613 7.351 apply_single 110 13.6 0.000 0.001 6.612 7.351 sum_up_and_integrate 110 10.3 0.054 0.068 7.233 7.246 integrate_v_rspace 110 11.3 0.003 0.003 7.180 7.194 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.856 6.867 calculate_rho_elec 110 8.6 0.040 0.048 6.856 6.866 init_scf_run 11 5.9 0.000 0.001 6.489 6.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.488 6.489 prepare_preconditioner 11 7.9 0.000 0.000 6.448 6.466 make_preconditioner 11 8.9 0.000 0.000 6.448 6.466 dbcsr_mm_accdrv_process 52282 16.1 4.577 5.479 6.055 6.357 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.022 6.178 ot_diis_step 99 11.5 0.010 0.010 5.859 5.859 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.792 5.813 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.983 5.676 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.319 5.319 make_m2s 4110 13.4 0.057 0.060 4.681 5.299 make_images 4110 14.4 0.402 0.447 4.572 5.187 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.616 4.628 cp_fm_diag_elpa_base 48 14.0 4.562 4.580 4.613 4.625 density_rs2pw 110 9.6 0.004 0.005 3.956 4.446 pw_transfer 1331 11.6 0.066 0.074 4.257 4.415 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.150 4.311 rs_pw_transfer 902 11.9 0.012 0.014 3.678 4.176 wfi_extrapolate 11 7.9 0.001 0.001 3.812 3.812 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.590 3.659 fft3d_ps 1111 14.6 1.108 1.338 3.501 3.647 fft_wrap_pw1pw2_140 451 13.1 0.201 0.219 3.335 3.503 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.381 3.383 grid_integrate_task_list 110 12.3 3.147 3.352 3.147 3.352 make_images_data 4110 15.4 0.047 0.050 2.558 3.189 mp_sum_l 6514 12.8 2.237 3.134 2.237 3.134 calculate_dm_sparse 110 9.5 0.001 0.001 3.052 3.091 hybrid_alltoall_any 4261 16.3 0.102 0.444 2.245 3.058 cp_fm_cholesky_invert 11 10.9 3.031 3.039 3.031 3.039 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.970 3.014 potential_pw2rs 110 12.3 0.008 0.009 2.978 3.002 calculate_first_density_matrix 1 7.0 0.000 0.000 2.576 2.580 grid_collocate_task_list 110 9.6 2.052 2.472 2.052 2.472 mp_alltoall_d11v 2046 13.8 2.032 2.421 2.032 2.421 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.087 2.116 mp_allgather_i34 2055 14.4 0.915 1.998 0.915 1.998 jit_kernel_multiply 10 16.4 1.124 1.975 1.124 1.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.873 1.883 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.875 1.875 mp_waitany 10164 13.8 1.293 1.824 1.293 1.824 mp_alltoall_z22v 1111 16.6 1.678 1.789 1.678 1.789 cp_fm_cholesky_decompose 22 10.9 1.699 1.706 1.699 1.706 dbcsr_complete_redistribute 325 12.2 0.234 0.296 1.369 1.645 multiply_cannon_metrocomm4 22605 15.4 0.074 0.078 0.805 1.585 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.569 1.581 acc_transpose_blocks 24660 15.4 0.112 0.116 1.540 1.575 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=76.019000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.000000, yerr=7.096734 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 659.554304E+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 3473 66428. MP_Allreduce 9694 567. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 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 0.045 0.062 63.042 63.043 qs_mol_dyn_low 1 2.0 0.003 0.003 62.110 62.119 qs_forces 11 3.9 0.007 0.040 62.010 62.011 qs_energies 11 4.9 0.001 0.002 58.705 58.707 scf_env_do_scf 11 5.9 0.000 0.001 50.702 50.702 scf_env_do_scf_inner_loop 99 6.5 0.021 0.053 41.632 41.634 velocity_verlet 10 3.0 0.001 0.001 33.817 33.820 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.672 29.015 qs_scf_new_mos 99 7.5 0.001 0.001 25.828 25.918 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.828 25.917 ot_scf_mini 99 9.5 0.002 0.003 24.605 24.717 multiply_cannon 2055 13.4 0.213 0.223 22.058 23.268 multiply_cannon_loop 2055 14.4 0.615 0.631 20.861 21.814 ot_mini 99 10.5 0.001 0.001 13.772 13.886 rebuild_ks_matrix 110 8.3 0.000 0.000 12.878 13.021 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.877 13.021 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.349 11.482 multiply_cannon_multrec 16440 15.4 3.968 4.966 9.713 10.691 mp_waitall_1 139946 16.5 7.185 10.427 7.185 10.427 qs_ot_get_derivative 99 11.5 0.001 0.002 9.323 9.433 init_scf_loop 11 6.9 0.000 0.000 9.032 9.032 multiply_cannon_metrocomm3 16440 15.4 0.044 0.046 4.202 7.347 prepare_preconditioner 11 7.9 0.000 0.000 7.155 7.173 make_preconditioner 11 8.9 0.000 0.000 7.155 7.173 sum_up_and_integrate 110 10.3 0.061 0.062 6.924 6.939 integrate_v_rspace 110 11.3 0.003 0.004 6.863 6.880 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.494 6.841 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.319 6.328 calculate_rho_elec 110 8.6 0.059 0.059 6.319 6.327 qs_ot_get_p 110 10.4 0.001 0.001 6.052 6.192 dbcsr_mm_accdrv_process 34862 16.1 4.503 5.291 5.599 5.732 init_scf_run 11 5.9 0.000 0.001 5.594 5.594 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.594 5.594 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.943 5.358 apply_single 110 13.6 0.000 0.000 4.942 5.357 make_m2s 4110 13.4 0.050 0.051 4.415 4.759 make_images 4110 14.4 0.394 0.514 4.300 4.645 density_rs2pw 110 9.6 0.004 0.004 3.394 4.621 ot_diis_step 99 11.5 0.011 0.011 4.416 4.416 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.274 4.277 multiply_cannon_sync_h2d 16440 15.4 3.691 4.270 3.691 4.270 rs_pw_transfer 902 11.9 0.011 0.012 2.991 4.189 pw_transfer 1331 11.6 0.066 0.072 3.977 3.988 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.922 3.922 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.870 3.884 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.237 3.837 grid_integrate_task_list 110 12.3 3.188 3.421 3.188 3.421 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.330 3.340 cp_fm_diag_elpa_base 48 14.0 3.257 3.296 3.328 3.338 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 3.285 3.300 fft3d_ps 1111 14.6 1.090 1.099 3.169 3.181 wfi_extrapolate 11 7.9 0.001 0.001 3.073 3.073 make_images_data 4110 15.4 0.043 0.047 2.572 3.038 cp_fm_cholesky_invert 11 10.9 2.909 2.916 2.909 2.916 hybrid_alltoall_any 4261 16.3 0.105 0.371 2.241 2.871 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.812 2.814 potential_pw2rs 110 12.3 0.010 0.011 2.590 2.606 calculate_dm_sparse 110 9.5 0.001 0.001 2.510 2.538 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.459 2.526 mp_waitany 17072 13.8 1.233 2.510 1.233 2.510 grid_collocate_task_list 110 9.6 2.080 2.478 2.080 2.478 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.400 2.458 calculate_first_density_matrix 1 7.0 0.000 0.001 2.432 2.434 multiply_cannon_metrocomm4 14385 15.4 0.045 0.047 0.871 2.417 mp_irecv_dv 48980 15.7 0.801 2.295 0.801 2.295 mp_sum_l 6514 12.8 1.592 2.204 1.592 2.204 mp_alltoall_d11v 2046 13.8 1.867 2.203 1.867 2.203 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.189 0.918 2.120 dbcsr_complete_redistribute 325 12.2 0.324 0.368 1.505 1.974 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.963 1.964 cp_fm_upper_to_full 70 14.2 1.401 1.779 1.401 1.779 cp_fm_cholesky_decompose 22 10.9 1.745 1.763 1.745 1.763 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.731 1.743 mp_allgather_i34 2055 14.4 0.519 1.641 0.519 1.641 jit_kernel_multiply 8 16.4 0.705 1.508 0.705 1.508 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.354 1.479 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.468 1.478 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.989 1.454 mp_alltoall_z22v 1111 16.6 1.310 1.342 1.310 1.342 rs_gather_matrices 110 12.3 0.235 0.264 1.000 1.308 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.265 1.272 acc_transpose_blocks 16440 15.4 0.073 0.075 1.234 1.268 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.043000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.272727, yerr=8.697867 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 728.809472E+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 3473 66426. MP_Allreduce 9694 608. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 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 0.018 0.029 68.870 68.870 qs_mol_dyn_low 1 2.0 0.003 0.003 68.631 68.640 qs_forces 11 3.9 0.003 0.003 68.540 68.541 qs_energies 11 4.9 0.001 0.001 65.051 65.055 scf_env_do_scf 11 5.9 0.000 0.001 55.888 55.891 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.616 43.617 velocity_verlet 10 3.0 0.001 0.001 38.629 38.632 dbcsr_multiply_generic 2055 12.4 0.113 0.117 30.421 30.566 qs_scf_new_mos 99 7.5 0.001 0.001 27.745 27.884 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.745 27.884 ot_scf_mini 99 9.5 0.003 0.003 26.100 26.218 multiply_cannon 2055 13.4 0.243 0.267 23.053 24.057 multiply_cannon_loop 2055 14.4 0.885 0.905 21.578 22.061 multiply_cannon_multrec 24660 15.4 4.215 6.819 13.289 14.557 ot_mini 99 10.5 0.001 0.001 14.292 14.436 rebuild_ks_matrix 110 8.3 0.000 0.000 12.719 12.847 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.719 12.846 init_scf_loop 11 6.9 0.000 0.000 12.227 12.227 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.235 11.349 prepare_preconditioner 11 7.9 0.000 0.000 10.400 10.416 make_preconditioner 11 8.9 0.000 0.000 10.400 10.416 qs_ot_get_derivative 99 11.5 0.001 0.002 10.156 10.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.603 10.086 dbcsr_mm_accdrv_process 52304 16.0 7.233 9.048 8.929 9.832 qs_ot_get_p 110 10.4 0.001 0.001 6.743 6.907 sum_up_and_integrate 110 10.3 0.068 0.071 6.877 6.890 integrate_v_rspace 110 11.3 0.003 0.003 6.808 6.823 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.574 6.589 calculate_rho_elec 110 8.6 0.078 0.082 6.573 6.588 mp_waitall_1 121746 16.5 4.545 6.477 4.545 6.477 init_scf_run 11 5.9 0.000 0.001 6.386 6.386 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.385 6.386 make_m2s 4110 13.4 0.060 0.063 5.684 6.032 make_images 4110 14.4 0.578 0.700 5.542 5.886 cp_fm_upper_to_full 70 14.2 3.370 4.901 3.370 4.901 qs_ot_p2m_diag 48 11.0 0.054 0.064 4.755 4.769 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.262 4.263 pw_transfer 1331 11.6 0.065 0.075 4.175 4.213 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.069 4.111 ot_diis_step 99 11.5 0.011 0.011 4.097 4.098 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.010 4.089 apply_single 110 13.6 0.000 0.000 4.010 4.088 dbcsr_complete_redistribute 325 12.2 0.417 0.472 2.799 3.953 density_rs2pw 110 9.6 0.004 0.004 3.428 3.942 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.638 3.647 cp_fm_diag_elpa_base 48 14.0 3.481 3.546 3.636 3.645 fft_wrap_pw1pw2_140 451 13.1 0.202 0.215 3.438 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.399 3.460 grid_integrate_task_list 110 12.3 3.268 3.436 3.268 3.436 fft3d_ps 1111 14.6 1.096 1.132 3.359 3.387 multiply_cannon_sync_h2d 24660 15.4 3.190 3.375 3.190 3.375 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.211 3.349 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.422 3.288 make_images_data 4110 15.4 0.047 0.050 2.889 3.277 calculate_dm_sparse 110 9.5 0.001 0.001 3.234 3.262 rs_pw_transfer 902 11.9 0.010 0.011 2.737 3.256 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.193 3.195 wfi_extrapolate 11 7.9 0.001 0.001 3.186 3.186 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.103 3.174 calculate_first_density_matrix 1 7.0 0.000 0.000 3.100 3.103 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.358 3.064 cp_fm_cholesky_invert 11 10.9 2.987 2.996 2.987 2.996 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.788 2.910 jit_kernel_multiply 12 15.8 1.363 2.847 1.363 2.847 mp_alltoall_i22 605 13.7 1.683 2.846 1.683 2.846 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.783 2.813 potential_pw2rs 110 12.3 0.013 0.013 2.462 2.476 grid_collocate_task_list 110 9.6 2.183 2.421 2.183 2.421 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.258 2.260 mp_alltoall_d11v 2046 13.8 1.965 2.174 1.965 2.174 cp_fm_cholesky_decompose 22 10.9 1.852 1.908 1.852 1.908 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.832 1.847 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.730 1.766 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.731 1.744 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.702 mp_sum_l 6514 12.8 0.995 1.682 0.995 1.682 mp_allgather_i34 2055 14.4 0.579 1.613 0.579 1.613 mp_waitany 13376 13.8 1.085 1.543 1.085 1.543 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.833 1.542 acc_transpose_blocks 24660 15.4 0.107 0.109 1.507 1.524 mp_alltoall_z22v 1111 16.6 1.404 1.510 1.404 1.510 mp_irecv_dv 62702 16.1 0.733 1.465 0.733 1.465 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.870000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=688.909091, yerr=8.447211 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 827.695104E+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 3473 66424. MP_Allreduce 9694 649. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 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 0.022 0.030 60.816 60.817 qs_mol_dyn_low 1 2.0 0.003 0.003 60.471 60.480 qs_forces 11 3.9 0.006 0.007 60.399 60.400 qs_energies 11 4.9 0.001 0.001 56.619 56.625 scf_env_do_scf 11 5.9 0.000 0.001 47.147 47.147 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 38.307 38.307 velocity_verlet 10 3.0 0.001 0.001 33.754 33.757 dbcsr_multiply_generic 2055 12.4 0.103 0.108 24.841 24.998 qs_scf_new_mos 99 7.5 0.001 0.001 22.347 22.419 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.347 22.419 ot_scf_mini 99 9.5 0.002 0.002 21.079 21.137 multiply_cannon 2055 13.4 0.246 0.271 18.799 20.160 multiply_cannon_loop 2055 14.4 0.319 0.331 17.383 17.679 rebuild_ks_matrix 110 8.3 0.000 0.000 12.571 12.632 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.571 12.631 ot_mini 99 10.5 0.001 0.001 11.298 11.359 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.174 11.227 multiply_cannon_multrec 8220 15.4 3.235 4.546 8.399 9.416 init_scf_loop 11 6.9 0.000 0.000 8.784 8.785 mp_waitall_1 103326 16.6 6.856 8.661 6.856 8.661 qs_ot_get_derivative 99 11.5 0.002 0.003 7.393 7.450 prepare_preconditioner 11 7.9 0.000 0.000 7.011 7.021 make_preconditioner 11 8.9 0.000 0.000 7.011 7.021 sum_up_and_integrate 110 10.3 0.079 0.081 6.956 6.967 integrate_v_rspace 110 11.3 0.003 0.003 6.876 6.888 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.745 6.760 calculate_rho_elec 110 8.6 0.116 0.116 6.744 6.760 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.544 6.629 init_scf_run 11 5.9 0.000 0.001 6.239 6.239 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.239 6.239 dbcsr_mm_accdrv_process 17442 15.9 3.246 4.576 5.034 5.968 qs_ot_get_p 110 10.4 0.001 0.001 5.641 5.720 make_m2s 4110 13.4 0.038 0.040 4.612 4.886 make_images 4110 14.4 0.640 0.697 4.483 4.758 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.224 4.754 pw_transfer 1331 11.6 0.066 0.071 4.483 4.492 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.376 4.389 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.036 4.042 ot_diis_step 99 11.5 0.012 0.012 3.878 3.878 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.778 3.829 apply_single 110 13.6 0.000 0.000 3.778 3.829 density_rs2pw 110 9.6 0.004 0.004 3.439 3.814 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.718 3.719 fft_wrap_pw1pw2_140 451 13.1 0.216 0.220 3.700 3.714 fft3d_ps 1111 14.6 1.155 1.175 3.616 3.625 cp_fm_cholesky_invert 11 10.9 3.603 3.608 3.603 3.608 grid_integrate_task_list 110 12.3 3.354 3.544 3.354 3.544 make_images_data 4110 15.4 0.038 0.043 2.682 3.156 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.124 3.132 cp_fm_diag_elpa_base 48 14.0 3.064 3.094 3.122 3.129 calculate_first_density_matrix 1 7.0 0.000 0.000 3.122 3.124 multiply_cannon_sync_h2d 8220 15.4 2.892 3.101 2.892 3.101 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.054 3.056 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.557 3.038 wfi_extrapolate 11 7.9 0.001 0.001 3.017 3.017 rs_pw_transfer 902 11.9 0.010 0.010 2.537 2.940 calculate_dm_sparse 110 9.5 0.001 0.001 2.890 2.928 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.713 2.716 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.632 2.661 grid_collocate_task_list 110 9.6 2.280 2.541 2.280 2.541 potential_pw2rs 110 12.3 0.015 0.016 2.435 2.442 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.324 2.353 mp_alltoall_d11v 2046 13.8 1.930 2.221 1.930 2.221 jit_kernel_multiply 9 15.8 1.477 2.201 1.477 2.201 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.773 1.987 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.950 1.976 cp_fm_cholesky_decompose 22 10.9 1.912 1.930 1.912 1.930 mp_allgather_i34 2055 14.4 0.598 1.868 0.598 1.868 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.854 1.866 dbcsr_complete_redistribute 325 12.2 0.552 0.572 1.655 1.766 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.738 1.748 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.547 1.670 mp_waitany 9240 13.8 1.159 1.567 1.159 1.567 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.974 1.498 mp_alltoall_z22v 1111 16.6 1.352 1.375 1.352 1.375 qs_create_task_list 11 7.9 0.001 0.001 1.231 1.336 generate_qs_task_list 11 8.9 0.378 0.447 1.230 1.335 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.263 1.313 rs_gather_matrices 110 12.3 0.324 0.365 0.993 1.269 multiply_cannon_metrocomm4 6165 15.4 0.018 0.020 0.474 1.267 x_to_yz 550 16.1 0.572 0.582 1.246 1.255 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.164 0.826 1.227 mp_irecv_dv 24056 15.7 0.449 1.221 0.449 1.221 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=60.817000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.363636, yerr=12.152611 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 1.374052E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.057 93.536 93.537 qs_mol_dyn_low 1 2.0 0.003 0.003 93.018 93.027 qs_forces 11 3.9 0.003 0.003 92.948 92.950 qs_energies 11 4.9 0.001 0.001 88.661 88.664 scf_env_do_scf 11 5.9 0.001 0.001 77.721 77.721 velocity_verlet 10 3.0 0.001 0.001 58.668 58.674 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 47.631 47.633 dbcsr_multiply_generic 2055 12.4 0.119 0.124 30.308 30.414 init_scf_loop 11 6.9 0.000 0.000 30.005 30.007 qs_scf_new_mos 99 7.5 0.001 0.001 28.501 28.543 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.500 28.542 prepare_preconditioner 11 7.9 0.000 0.000 27.884 27.892 make_preconditioner 11 8.9 0.000 0.000 27.884 27.892 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.817 27.352 ot_scf_mini 99 9.5 0.002 0.002 26.687 26.725 multiply_cannon 2055 13.4 0.349 0.364 22.764 23.617 multiply_cannon_loop 2055 14.4 0.342 0.343 20.869 21.196 cp_fm_upper_to_full 70 14.2 13.053 18.848 13.053 18.848 rebuild_ks_matrix 110 8.3 0.000 0.001 14.751 14.821 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.750 14.820 ot_mini 99 10.5 0.001 0.001 14.727 14.768 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.338 13.400 dbcsr_complete_redistribute 325 12.2 1.012 1.036 7.720 11.065 multiply_cannon_multrec 8220 15.4 4.380 4.579 10.105 10.264 qs_ot_get_derivative 99 11.5 0.159 0.184 10.038 10.075 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.658 9.996 mp_waitall_1 84994 16.7 8.382 9.501 8.382 9.501 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.051 9.352 mp_alltoall_i22 605 13.7 5.690 9.050 5.690 9.050 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.157 8.184 calculate_rho_elec 110 8.6 0.228 0.228 8.156 8.184 sum_up_and_integrate 110 10.3 0.151 0.152 7.721 7.733 integrate_v_rspace 110 11.3 0.004 0.005 7.570 7.582 qs_ot_get_p 110 10.4 0.001 0.001 6.579 6.633 init_scf_run 11 5.9 0.000 0.001 6.360 6.360 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.360 6.360 make_m2s 4110 13.4 0.043 0.044 5.577 6.132 cp_fm_cholesky_invert 11 10.9 6.027 6.032 6.027 6.032 make_images 4110 14.4 0.886 0.939 5.388 5.943 dbcsr_mm_accdrv_process 11614 15.7 3.400 3.875 5.582 5.862 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.308 5.744 pw_transfer 1331 11.6 0.075 0.075 5.648 5.653 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.531 5.537 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.772 5.347 apply_single 110 13.6 0.000 0.000 4.772 5.346 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.817 4.823 fft_wrap_pw1pw2_140 451 13.1 0.222 0.224 4.760 4.767 fft3d_ps 1111 14.6 1.301 1.309 4.724 4.731 ot_diis_step 99 11.5 0.016 0.016 4.658 4.658 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.357 4.358 density_rs2pw 110 9.6 0.004 0.004 4.093 4.116 qs_energies_init_hamiltonians 11 5.9 0.238 0.243 3.956 3.958 multiply_cannon_sync_h2d 8220 15.4 3.945 3.950 3.945 3.950 make_images_data 4110 15.4 0.042 0.045 3.054 3.835 hybrid_alltoall_any 4261 16.3 0.257 0.554 3.013 3.823 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.330 3.802 grid_integrate_task_list 110 12.3 3.669 3.709 3.669 3.709 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.676 3.676 cp_fm_diag_elpa_base 48 14.0 3.109 3.321 3.673 3.674 wfi_extrapolate 11 7.9 0.001 0.001 3.487 3.487 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.364 3.367 calculate_dm_sparse 110 9.5 0.001 0.001 3.264 3.286 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.823 2.847 potential_pw2rs 110 12.3 0.021 0.021 2.777 2.788 calculate_first_density_matrix 1 7.0 0.000 0.000 2.745 2.747 grid_collocate_task_list 110 9.6 2.640 2.663 2.640 2.663 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.603 2.623 rs_pw_transfer 902 11.9 0.011 0.011 2.562 2.601 jit_kernel_multiply 10 15.3 1.980 2.358 1.980 2.358 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.245 2.301 cp_fm_cholesky_decompose 22 10.9 2.243 2.285 2.243 2.285 mp_alltoall_d11v 2046 13.8 2.198 2.274 2.198 2.274 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.171 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.020 2.027 qs_create_task_list 11 7.9 0.001 0.001 1.908 1.957 generate_qs_task_list 11 8.9 0.736 0.787 1.908 1.956 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.846 1.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=93.537000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1229.636364, yerr=63.505580 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.326976E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 2520936. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.031 221.121 221.123 qs_mol_dyn_low 1 2.0 0.006 0.024 220.081 220.095 qs_forces 11 3.9 0.005 0.006 219.943 219.946 qs_energies 11 4.9 0.002 0.010 214.078 214.090 scf_env_do_scf 11 5.9 0.001 0.001 196.666 196.670 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 174.007 174.009 velocity_verlet 10 3.0 0.001 0.001 131.229 131.231 qs_scf_new_mos 117 7.6 0.001 0.001 130.573 130.906 qs_scf_loop_do_ot 117 8.6 0.001 0.001 130.573 130.905 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.387 126.346 ot_scf_mini 117 9.6 0.003 0.003 123.960 124.273 multiply_cannon 2507 13.6 0.240 0.253 101.169 102.582 multiply_cannon_loop 2507 14.6 2.044 2.112 98.700 99.936 ot_mini 117 10.6 0.001 0.001 66.271 66.575 multiply_cannon_multrec 60168 15.6 33.366 36.227 41.508 43.424 qs_ot_get_derivative 117 11.6 0.001 0.003 41.371 41.684 rebuild_ks_matrix 128 8.3 0.001 0.001 35.794 36.005 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 35.793 36.005 qs_ot_get_p 128 10.4 0.001 0.001 35.562 35.977 qs_ks_update_qs_env 128 7.6 0.001 0.001 32.138 32.354 mp_waitall_1 267128 16.5 29.568 31.584 29.568 31.584 multiply_cannon_sync_h2d 60168 15.6 27.760 30.698 27.760 30.698 qs_ot_p2m_diag 83 11.4 0.079 0.095 28.362 28.445 cp_dbcsr_syevd 83 12.4 0.004 0.005 25.827 25.829 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.414 25.574 apply_single 128 13.6 0.001 0.001 24.413 25.573 ot_diis_step 117 11.6 0.007 0.008 24.652 24.654 init_scf_loop 11 6.9 0.000 0.000 22.552 22.554 cp_fm_diag_elpa 83 13.4 0.000 0.000 22.290 22.331 cp_fm_diag_elpa_base 83 14.4 22.181 22.223 22.284 22.327 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.581 19.798 prepare_preconditioner 11 7.9 0.000 0.000 17.681 17.742 make_preconditioner 11 8.9 0.000 0.000 17.681 17.742 multiply_cannon_metrocomm3 60168 15.6 0.114 0.120 15.292 17.575 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.808 16.997 sum_up_and_integrate 128 10.3 0.090 0.109 15.862 15.878 integrate_v_rspace 128 11.3 0.004 0.005 15.771 15.790 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.636 15.772 calculate_rho_elec 128 8.7 0.046 0.065 15.635 15.771 make_m2s 5014 13.6 0.104 0.110 15.191 15.642 make_images 5014 14.6 0.404 0.423 15.010 15.470 init_scf_run 11 5.9 0.000 0.001 13.017 13.018 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.017 13.017 density_rs2pw 128 9.7 0.006 0.007 8.710 11.918 rs_pw_transfer 1046 11.9 0.017 0.019 7.288 10.485 cp_fm_cholesky_invert 11 10.9 9.999 10.008 9.999 10.008 wfi_extrapolate 11 7.9 0.001 0.001 9.714 9.714 pw_transfer 1547 11.6 0.077 0.106 9.374 9.684 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 9.167 9.483 mp_sum_l 7870 13.0 7.767 8.887 7.767 8.887 calculate_dm_sparse 128 9.5 0.001 0.001 8.434 8.524 make_images_data 5014 15.6 0.066 0.072 7.483 8.397 dbcsr_mm_accdrv_process 124484 16.2 3.065 3.253 7.685 8.263 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.078 8.234 fft3d_ps 1291 14.7 2.104 2.864 7.957 8.187 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.853 7.937 fft_wrap_pw1pw2_140 523 13.2 0.439 0.508 7.567 7.919 hybrid_alltoall_any 5200 16.5 0.290 2.238 6.479 7.862 multiply_cannon_metrocomm1 60168 15.6 0.090 0.095 6.328 7.846 grid_integrate_task_list 128 12.3 7.035 7.563 7.035 7.563 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 6.878 6.887 mp_alltoall_d11v 2415 14.1 5.307 6.670 5.307 6.670 potential_pw2rs 128 12.3 0.009 0.011 5.940 6.019 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.908 5.985 mp_waitany 16020 13.9 2.729 5.911 2.729 5.911 grid_collocate_task_list 128 9.7 4.567 5.777 4.567 5.777 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.294 2.121 5.364 cp_fm_cholesky_decompose 22 10.9 5.287 5.302 5.287 5.302 mp_sum_d 4468 12.1 3.606 4.557 3.606 4.557 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=221.123000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.545455, yerr=6.169774 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.166473E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.582784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069299E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842592400 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58319. MP_Allreduce 10986 968. MP_Sync 86 MP_Alltoall 1955 5616396. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.061 210.713 210.715 qs_mol_dyn_low 1 2.0 0.003 0.004 209.904 209.951 qs_forces 11 3.9 0.004 0.005 209.678 209.679 qs_energies 11 4.9 0.001 0.001 202.566 202.582 scf_env_do_scf 11 5.9 0.001 0.001 185.249 185.259 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 147.583 147.586 velocity_verlet 10 3.0 0.001 0.001 131.579 131.581 qs_scf_new_mos 116 7.6 0.001 0.001 104.791 105.216 qs_scf_loop_do_ot 116 8.6 0.001 0.001 104.791 105.215 dbcsr_multiply_generic 2485 12.5 0.187 0.192 100.443 101.893 ot_scf_mini 116 9.6 0.004 0.004 99.866 100.307 multiply_cannon 2485 13.5 0.474 0.526 78.362 83.822 multiply_cannon_loop 2485 14.5 1.238 1.283 74.434 77.719 ot_mini 116 10.6 0.001 0.001 52.390 52.882 mp_waitall_1 212858 16.6 27.613 42.706 27.613 42.706 init_scf_loop 11 6.9 0.000 0.000 37.551 37.553 rebuild_ks_matrix 127 8.3 0.001 0.001 34.975 35.450 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 34.975 35.449 multiply_cannon_multrec 29820 15.5 21.206 25.145 30.753 35.153 prepare_preconditioner 11 7.9 0.000 0.000 32.865 32.940 make_preconditioner 11 8.9 0.000 0.000 32.864 32.939 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.478 32.040 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.469 31.907 multiply_cannon_metrocomm3 29820 15.5 0.094 0.102 17.555 30.913 qs_ot_get_derivative 116 11.6 0.003 0.006 29.976 30.408 qs_ot_get_p 127 10.4 0.001 0.001 28.884 29.543 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.324 23.624 apply_single 127 13.6 0.001 0.001 22.324 23.624 qs_ot_p2m_diag 82 11.4 0.185 0.213 23.396 23.447 ot_diis_step 116 11.6 0.014 0.015 22.218 22.220 cp_dbcsr_syevd 82 12.4 0.005 0.005 22.072 22.074 multiply_cannon_sync_h2d 29820 15.5 18.569 21.090 18.569 21.090 cp_fm_cholesky_invert 11 10.9 19.358 19.373 19.358 19.373 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.288 18.314 cp_fm_diag_elpa_base 82 14.4 18.001 18.107 18.280 18.309 make_m2s 4970 13.5 0.088 0.092 15.466 17.310 make_images 4970 14.5 1.161 1.354 15.261 17.109 sum_up_and_integrate 127 10.3 0.116 0.132 16.075 16.106 integrate_v_rspace 127 11.3 0.004 0.004 15.959 15.993 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.564 15.598 calculate_rho_elec 127 8.7 0.088 0.105 15.563 15.597 init_scf_run 11 5.9 0.000 0.001 12.323 12.325 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.323 12.325 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.950 12.270 make_images_data 4970 15.5 0.064 0.071 9.261 11.328 density_rs2pw 127 9.7 0.006 0.006 8.857 11.268 multiply_cannon_metrocomm4 27335 15.5 0.096 0.110 3.807 10.682 hybrid_alltoall_any 5155 16.4 0.341 1.492 7.919 10.437 mp_irecv_dv 68888 16.3 3.614 10.286 3.614 10.286 pw_transfer 1535 11.6 0.087 0.105 10.089 10.189 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 9.864 9.967 rs_pw_transfer 1038 11.9 0.015 0.016 7.198 9.666 dbcsr_mm_accdrv_process 61726 16.2 4.522 5.511 9.041 9.632 wfi_extrapolate 11 7.9 0.001 0.001 9.045 9.045 cp_fm_cholesky_decompose 22 10.9 8.431 8.543 8.431 8.543 fft_wrap_pw1pw2_140 519 13.2 0.467 0.522 8.261 8.366 fft3d_ps 1281 14.7 2.734 2.909 8.244 8.299 grid_integrate_task_list 127 12.3 7.117 7.565 7.117 7.565 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.604 7.396 mp_sum_l 7804 13.0 4.884 6.923 4.884 6.923 calculate_dm_sparse 127 9.5 0.001 0.001 6.625 6.766 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.461 6.470 potential_pw2rs 127 12.3 0.015 0.018 6.200 6.220 mp_alltoall_d11v 2401 14.1 5.030 5.846 5.030 5.846 grid_collocate_task_list 127 9.7 4.670 5.830 4.670 5.830 mp_allgather_i34 2485 14.5 2.506 5.814 2.506 5.814 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.618 5.687 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.396 5.571 mp_waitany 11660 13.9 2.604 5.042 2.604 5.042 dbcsr_complete_redistribute 393 12.7 0.766 0.853 3.837 4.744 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.370 2.099 4.500 mp_sum_d 4446 12.1 2.954 4.349 2.954 4.349 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=210.715000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.090909, yerr=2.193152 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 942.288896E+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 4020 57958. MP_Allreduce 11047 1005. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.045 0.092 185.715 185.717 qs_mol_dyn_low 1 2.0 0.003 0.005 185.131 185.144 qs_forces 11 3.9 0.011 0.064 185.015 185.018 qs_energies 11 4.9 0.004 0.020 178.094 178.104 scf_env_do_scf 11 5.9 0.001 0.002 161.837 161.837 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 124.665 124.666 velocity_verlet 10 3.0 0.003 0.014 118.380 118.382 qs_scf_new_mos 117 7.6 0.001 0.001 84.920 85.219 qs_scf_loop_do_ot 117 8.6 0.001 0.002 84.920 85.218 dbcsr_multiply_generic 2507 12.6 0.180 0.185 83.262 84.391 ot_scf_mini 117 9.6 0.004 0.004 80.775 81.162 multiply_cannon 2507 13.6 0.499 0.522 62.525 67.478 multiply_cannon_loop 2507 14.6 0.853 0.878 59.118 61.919 ot_mini 117 10.6 0.001 0.001 43.641 44.016 init_scf_loop 11 6.9 0.001 0.005 37.055 37.059 mp_waitall_1 170520 16.6 25.776 34.749 25.776 34.749 prepare_preconditioner 11 7.9 0.000 0.001 32.834 32.879 make_preconditioner 11 8.9 0.000 0.002 32.833 32.879 rebuild_ks_matrix 128 8.3 0.001 0.001 31.929 32.435 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 31.929 32.434 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.353 31.845 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.656 29.123 multiply_cannon_multrec 20056 15.6 13.505 16.088 22.460 25.120 multiply_cannon_metrocomm3 20056 15.6 0.059 0.065 15.306 24.663 qs_ot_get_derivative 117 11.6 0.002 0.002 23.820 24.202 qs_ot_get_p 128 10.4 0.001 0.002 22.315 22.793 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.952 20.984 apply_single 128 13.6 0.001 0.001 19.952 20.983 ot_diis_step 117 11.6 0.018 0.033 19.714 19.714 qs_ot_p2m_diag 83 11.4 0.265 0.272 17.711 17.727 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.655 16.656 make_m2s 5014 13.6 0.080 0.085 15.246 16.281 make_images 5014 14.6 1.189 1.266 15.015 16.048 multiply_cannon_sync_h2d 20056 15.6 14.281 15.783 14.281 15.783 cp_fm_cholesky_invert 11 10.9 15.590 15.600 15.590 15.600 sum_up_and_integrate 128 10.3 0.135 0.146 15.441 15.493 integrate_v_rspace 128 11.3 0.004 0.005 15.306 15.357 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.075 15.112 calculate_rho_elec 128 8.7 0.132 0.147 15.075 15.111 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.185 13.200 cp_fm_diag_elpa_base 83 14.4 12.741 12.911 13.181 13.196 init_scf_run 11 5.9 0.000 0.001 11.097 11.097 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.096 11.097 make_images_data 5014 15.6 0.061 0.070 9.391 10.893 hybrid_alltoall_any 5200 16.5 0.432 1.972 8.077 10.205 density_rs2pw 128 9.7 0.006 0.007 7.885 10.070 pw_transfer 1547 11.6 0.087 0.109 9.845 10.005 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.619 9.785 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.450 9.725 multiply_cannon_metrocomm4 17549 15.6 0.062 0.072 3.431 9.160 mp_irecv_dv 50230 16.2 3.308 8.920 3.308 8.920 dbcsr_mm_accdrv_process 41502 16.2 4.467 5.407 8.410 8.589 fft_wrap_pw1pw2_140 523 13.2 0.476 0.519 8.339 8.505 fft3d_ps 1291 14.7 2.674 2.906 7.883 8.002 rs_pw_transfer 1046 11.9 0.014 0.015 5.753 7.928 cp_fm_cholesky_decompose 22 10.9 7.883 7.909 7.883 7.909 grid_integrate_task_list 128 12.3 7.279 7.806 7.279 7.806 wfi_extrapolate 11 7.9 0.001 0.001 7.686 7.686 cp_fm_upper_to_full 105 14.8 5.874 7.518 5.874 7.518 dbcsr_complete_redistribute 395 12.7 1.165 1.198 4.894 6.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.887 5.892 calculate_dm_sparse 128 9.5 0.001 0.001 5.742 5.864 grid_collocate_task_list 128 9.7 4.893 5.782 4.893 5.782 mp_alltoall_d11v 2415 14.1 4.666 5.599 4.666 5.599 potential_pw2rs 128 12.3 0.020 0.023 5.451 5.507 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.625 5.460 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.718 5.435 mp_allgather_i34 2507 14.6 1.921 5.325 1.921 5.325 mp_sum_l 7870 13.0 3.690 5.257 3.690 5.257 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.920 5.067 mp_waitany 11748 13.9 2.420 4.684 2.420 4.684 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.461 4.231 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.098 4.136 rs_pw_transfer_RS2PW_140 139 11.5 0.333 0.357 1.943 4.126 mp_alltoall_i22 716 14.1 2.128 4.088 2.128 4.088 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.789 3.834 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.817 3.818 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=185.717000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.545455, yerr=11.324660 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.125880E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 10977 1091. MP_Sync 87 MP_Alltoall 1712 12503084. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 199.783 199.785 qs_mol_dyn_low 1 2.0 0.003 0.003 199.371 199.383 qs_forces 11 3.9 0.004 0.005 198.448 198.454 qs_energies 11 4.9 0.001 0.002 191.108 191.118 scf_env_do_scf 11 5.9 0.001 0.001 173.615 173.622 velocity_verlet 10 3.0 0.001 0.001 131.588 131.624 scf_env_do_scf_inner_loop 117 6.6 0.004 0.010 124.969 124.970 qs_scf_new_mos 117 7.6 0.001 0.001 86.100 86.394 qs_scf_loop_do_ot 117 8.6 0.001 0.001 86.099 86.393 ot_scf_mini 117 9.6 0.003 0.003 81.546 81.814 dbcsr_multiply_generic 2507 12.6 0.187 0.194 80.826 81.437 multiply_cannon 2507 13.6 0.548 0.582 55.331 57.984 multiply_cannon_loop 2507 14.6 1.180 1.209 51.498 52.998 init_scf_loop 11 6.9 0.000 0.000 48.515 48.517 prepare_preconditioner 11 7.9 0.000 0.000 44.307 44.327 make_preconditioner 11 8.9 0.000 0.000 44.307 44.327 ot_mini 117 10.6 0.001 0.001 43.266 43.537 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.945 42.926 rebuild_ks_matrix 128 8.3 0.001 0.001 30.762 30.956 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 30.761 30.956 multiply_cannon_multrec 30084 15.6 14.332 19.189 26.556 30.878 mp_waitall_1 147882 16.7 17.991 27.969 17.991 27.969 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.692 27.856 qs_ot_get_p 128 10.4 0.001 0.001 24.328 24.601 qs_ot_get_derivative 117 11.6 0.002 0.002 23.592 23.860 make_m2s 5014 13.6 0.096 0.101 20.829 21.701 make_images 5014 14.6 1.963 2.268 20.521 21.390 qs_ot_p2m_diag 83 11.4 0.343 0.390 19.909 19.968 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.067 19.611 apply_single 128 13.6 0.001 0.001 19.066 19.610 ot_diis_step 117 11.6 0.017 0.018 19.539 19.541 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.546 18.548 cp_fm_cholesky_invert 11 10.9 17.492 17.501 17.492 17.501 cp_fm_upper_to_full 105 14.8 11.561 16.951 11.561 16.951 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.564 16.093 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.325 15.386 calculate_rho_elec 128 8.7 0.176 0.191 15.325 15.385 sum_up_and_integrate 128 10.3 0.141 0.152 15.358 15.385 integrate_v_rspace 128 11.3 0.004 0.005 15.216 15.248 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.111 15.126 cp_fm_diag_elpa_base 83 14.4 14.095 14.430 15.104 15.117 dbcsr_complete_redistribute 395 12.7 1.508 1.621 9.173 12.915 make_images_data 5014 15.6 0.064 0.070 11.220 12.578 multiply_cannon_sync_h2d 30084 15.6 11.639 12.368 11.639 12.368 dbcsr_mm_accdrv_process 62264 16.2 7.284 8.172 11.804 12.367 init_scf_run 11 5.9 0.000 0.001 11.500 11.501 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.500 11.501 hybrid_alltoall_any 5200 16.5 0.528 2.211 10.031 11.482 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.655 11.379 pw_transfer 1547 11.6 0.088 0.106 9.974 10.045 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.772 9.962 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.340 9.955 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.745 9.821 mp_alltoall_i22 716 14.1 5.593 9.359 5.593 9.359 density_rs2pw 128 9.7 0.006 0.006 7.657 9.015 fft_wrap_pw1pw2_140 523 13.2 0.482 0.492 8.534 8.630 cp_fm_cholesky_decompose 22 10.9 8.242 8.331 8.242 8.331 wfi_extrapolate 11 7.9 0.001 0.001 8.019 8.019 fft3d_ps 1291 14.7 2.784 2.863 7.970 8.017 grid_integrate_task_list 128 12.3 7.423 7.923 7.423 7.923 multiply_cannon_metrocomm4 25070 15.6 0.077 0.087 2.695 7.080 mp_irecv_dv 76098 16.2 2.551 6.816 2.551 6.816 rs_pw_transfer 1046 11.9 0.013 0.015 5.197 6.621 calculate_dm_sparse 128 9.5 0.001 0.001 6.258 6.330 mp_alltoall_d11v 2415 14.1 5.338 6.190 5.338 6.190 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.884 5.921 grid_collocate_task_list 128 9.7 5.079 5.777 5.079 5.777 potential_pw2rs 128 12.3 0.022 0.022 5.081 5.097 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.748 4.815 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.587 4.674 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 4.470 4.472 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.347 4.414 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=199.785000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1058.818182, yerr=19.338225 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019193401344 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019193401344 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820223E+12 0.0% 0.0% 100.0% flops max/rank 5.821533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705860672 0.0% 0.0% 100.0% number of processed stacks 1948544 0.0% 0.0% 100.0% average stack size 0.0 0.0 3441.5 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 1.530900E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 239616 MPI messages size (bytes): total size 1.321658E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.515734E+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 113856 59693334528 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108831376 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8885 51. MP_Alltoall 9606 802511. MP_ISend 39892 2096308. MP_IRecv 39892 2095413. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 58945. MP_Allreduce 10875 1179. MP_Sync 86 MP_Alltoall 1700 18828234. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.101 184.650 184.651 qs_mol_dyn_low 1 2.0 0.003 0.003 184.194 184.201 qs_forces 11 3.9 0.004 0.004 184.062 184.064 qs_energies 11 4.9 0.001 0.001 176.319 176.329 scf_env_do_scf 11 5.9 0.001 0.001 157.970 157.972 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 120.004 120.005 velocity_verlet 10 3.0 0.001 0.001 119.364 119.369 qs_scf_new_mos 116 7.6 0.001 0.001 80.100 80.194 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.100 80.193 dbcsr_multiply_generic 2496 12.6 0.180 0.190 76.672 77.238 ot_scf_mini 116 9.6 0.003 0.004 75.611 75.703 multiply_cannon 2496 13.6 0.579 0.619 56.156 60.895 multiply_cannon_loop 2496 14.6 0.442 0.457 51.372 52.345 ot_mini 116 10.6 0.001 0.001 41.258 41.359 init_scf_loop 11 6.9 0.000 0.000 37.810 37.812 mp_waitall_1 125208 16.7 27.729 35.502 27.729 35.502 prepare_preconditioner 11 7.9 0.000 0.000 33.758 33.784 make_preconditioner 11 8.9 0.000 0.000 33.758 33.784 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.492 31.797 rebuild_ks_matrix 127 8.3 0.001 0.001 30.764 30.859 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.764 30.858 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.873 27.954 multiply_cannon_multrec 9984 15.6 10.318 15.870 17.789 21.804 qs_ot_get_p 127 10.4 0.001 0.001 21.230 21.378 qs_ot_get_derivative 116 11.6 0.246 0.284 21.048 21.130 multiply_cannon_metrocomm3 9984 15.6 0.022 0.024 13.164 20.638 ot_diis_step 116 11.6 0.019 0.021 20.103 20.104 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.629 19.975 apply_single 127 13.6 0.001 0.001 19.629 19.974 cp_fm_cholesky_invert 11 10.9 19.487 19.494 19.487 19.494 make_m2s 4992 13.6 0.066 0.069 16.613 18.904 make_images 4992 14.6 2.290 2.766 16.311 18.605 qs_ot_p2m_diag 82 11.4 0.489 0.495 17.055 17.073 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.009 16.035 calculate_rho_elec 127 8.7 0.257 0.267 16.009 16.035 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.850 15.853 sum_up_and_integrate 127 10.3 0.180 0.189 15.595 15.650 integrate_v_rspace 127 11.3 0.004 0.005 15.415 15.478 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.619 12.629 make_images_data 4992 15.6 0.052 0.059 10.117 12.626 cp_fm_diag_elpa_base 82 14.4 12.374 12.455 12.612 12.621 hybrid_alltoall_any 5177 16.4 0.781 3.548 9.882 12.425 multiply_cannon_sync_h2d 9984 15.6 11.394 12.096 11.394 12.096 init_scf_run 11 5.9 0.000 0.001 11.417 11.417 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.417 11.417 pw_transfer 1535 11.6 0.085 0.094 10.435 10.467 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.213 10.251 fft_wrap_pw1pw2_140 519 13.2 0.492 0.515 8.928 8.964 multiply_cannon_metrocomm1 9984 15.6 0.028 0.029 4.985 8.622 cp_fm_cholesky_decompose 22 10.9 8.362 8.492 8.362 8.492 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.360 8.427 fft3d_ps 1281 14.7 2.694 2.764 8.386 8.405 density_rs2pw 127 9.7 0.005 0.006 7.511 8.392 grid_integrate_task_list 127 12.3 7.707 8.119 7.707 8.119 dbcsr_mm_accdrv_process 20634 16.1 2.621 3.402 7.110 7.952 wfi_extrapolate 11 7.9 0.001 0.001 7.684 7.684 mp_allgather_i34 2496 14.6 2.994 7.466 2.994 7.466 mp_alltoall_d11v 2401 14.1 5.619 6.622 5.619 6.622 calculate_dm_sparse 127 9.5 0.001 0.001 6.159 6.233 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.035 6.041 grid_collocate_task_list 127 9.7 5.353 6.037 5.353 6.037 dbcsr_complete_redistribute 393 12.7 2.122 2.171 5.498 5.926 rs_pw_transfer 1038 11.9 0.013 0.013 4.527 5.446 multiply_cannon_metrocomm4 7488 15.6 0.024 0.028 1.872 5.394 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.364 5.365 mp_irecv_dv 28706 15.9 1.835 5.301 1.835 5.301 potential_pw2rs 127 12.3 0.026 0.027 4.933 4.952 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.903 4.939 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.591 3.908 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.817 3.862 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.566 3.853 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.794 3.808 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.726 3.808 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.693 3.717 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=184.651000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1431.636364, yerr=51.520998 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410020794368 0.0% 0.0% 100.0% flops 32 x 32 x 32 1947767668736 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957875093504 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544039424 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714619674624 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 11444697722880 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019193401344 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019193401344 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624869085184 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.820223E+12 0.0% 0.0% 100.0% flops max/rank 11.609361E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705860672 0.0% 0.0% 100.0% number of processed stacks 1951856 0.0% 0.0% 100.0% average stack size 0.0 0.0 3435.6 marketing flops 143.531365E+12 ------------------------------------------------------------------------------- # multiplications 2496 max memory usage/rank 2.947584E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99840 MPI messages size (bytes): total size 1.127791E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.295984E+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 45208 35114713088 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108831376 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3951 59878. MP_Allreduce 10875 1518. MP_Sync 86 MP_Alltoall 1700 36954490. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.089 312.648 312.661 qs_mol_dyn_low 1 2.0 0.003 0.004 311.925 311.931 qs_forces 11 3.9 0.004 0.005 311.810 311.813 qs_energies 11 4.9 0.001 0.001 302.196 302.209 scf_env_do_scf 11 5.9 0.001 0.001 279.246 279.259 velocity_verlet 10 3.0 0.001 0.001 224.083 224.092 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 148.251 148.252 init_scf_loop 11 6.9 0.000 0.000 130.728 130.730 prepare_preconditioner 11 7.9 0.000 0.000 124.941 124.973 make_preconditioner 11 8.9 0.000 0.000 124.941 124.973 make_full_inverse_cholesky 11 9.9 0.000 0.000 100.329 122.055 qs_scf_new_mos 116 7.6 0.001 0.001 96.194 96.308 qs_scf_loop_do_ot 116 8.6 0.001 0.001 96.193 96.307 ot_scf_mini 116 9.6 0.004 0.004 91.326 91.358 dbcsr_multiply_generic 2496 12.6 0.212 0.223 86.119 86.769 cp_fm_upper_to_full 104 14.8 54.723 78.285 54.723 78.285 multiply_cannon 2496 13.6 0.705 0.732 60.382 60.962 multiply_cannon_loop 2496 14.6 0.468 0.475 56.575 58.222 ot_mini 116 10.6 0.001 0.001 46.390 46.432 dbcsr_complete_redistribute 393 12.7 3.970 4.004 30.433 43.815 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.930 40.306 rebuild_ks_matrix 127 8.3 0.001 0.001 39.689 39.788 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 39.689 39.787 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.567 37.877 qs_ks_update_qs_env 127 7.6 0.001 0.001 36.485 36.572 cp_fm_cholesky_invert 11 10.9 35.663 35.670 35.663 35.670 mp_alltoall_i22 712 14.1 22.316 35.577 22.316 35.577 mp_waitall_1 103208 16.8 30.822 34.826 30.822 34.826 qs_ot_get_p 127 10.4 0.001 0.001 29.062 29.096 qs_ot_get_derivative 116 11.6 0.005 0.006 25.592 25.632 qs_ot_p2m_diag 82 11.4 0.867 0.872 24.626 24.658 cp_dbcsr_syevd 82 12.4 0.006 0.006 22.840 22.843 multiply_cannon_metrocomm3 9984 15.6 0.023 0.023 20.427 22.002 make_m2s 4992 13.6 0.077 0.079 20.797 21.848 make_images 4992 14.6 3.773 3.877 20.323 21.374 qs_rho_update_rho_low 127 7.7 0.001 0.001 21.346 21.363 calculate_rho_elec 127 8.7 0.479 0.480 21.345 21.362 ot_diis_step 116 11.6 0.022 0.022 20.742 20.743 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.955 20.226 apply_single 127 13.6 0.001 0.001 19.954 20.226 sum_up_and_integrate 127 10.3 0.323 0.325 20.014 20.089 integrate_v_rspace 127 11.3 0.004 0.004 19.691 19.765 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.481 19.483 cp_fm_diag_elpa_base 82 14.4 15.010 16.615 19.476 19.476 multiply_cannon_multrec 9984 15.6 10.478 12.268 17.908 17.992 multiply_cannon_sync_h2d 9984 15.6 15.538 15.554 15.538 15.554 hybrid_alltoall_any 5177 16.4 1.291 3.012 11.433 13.367 make_images_data 4992 15.6 0.060 0.065 11.374 13.322 pw_transfer 1535 11.6 0.092 0.093 13.147 13.155 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 12.913 12.921 init_scf_run 11 5.9 0.000 0.001 12.675 12.675 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.674 12.675 fft_wrap_pw1pw2_140 519 13.2 0.540 0.545 11.371 11.381 fft3d_ps 1281 14.7 2.727 2.736 10.960 10.968 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.980 10.014 cp_fm_cholesky_decompose 22 10.9 9.740 9.783 9.740 9.783 wfi_extrapolate 11 7.9 0.001 0.001 9.572 9.572 mp_alltoall_d11v 2401 14.1 8.419 9.348 8.419 9.348 dbcsr_mm_accdrv_process 20634 16.0 3.835 5.639 7.190 9.003 density_rs2pw 127 9.7 0.005 0.005 8.867 8.991 grid_integrate_task_list 127 12.3 8.520 8.734 8.520 8.734 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.241 8.244 calculate_dm_sparse 127 9.5 0.001 0.001 6.742 6.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.524 6.592 grid_collocate_task_list 127 9.7 6.322 6.363 6.322 6.363 rs_scatter_matrices 138 9.7 3.652 4.506 6.139 6.361 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.246 6.323 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=312.661000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2641.272727, yerr=154.820771 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 1.262232E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.069 86.142 86.143 qs_energies 1 2.0 0.000 0.000 85.471 85.485 ls_scf 1 3.0 0.000 0.000 84.455 84.468 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.086 73.257 multiply_cannon 111 7.7 0.017 0.020 56.133 57.393 multiply_cannon_loop 111 8.7 0.210 0.226 52.730 54.172 ls_scf_main 1 4.0 0.000 0.000 52.705 52.705 density_matrix_trs4 2 5.0 0.002 0.003 47.027 47.107 ls_scf_init_scf 1 4.0 0.000 0.000 28.700 28.701 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.570 27.627 mp_waitall_1 11031 10.9 22.979 25.927 22.979 25.927 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.445 25.462 multiply_cannon_multrec 2664 9.7 8.164 8.899 15.596 17.634 multiply_cannon_sync_h2d 2664 9.7 13.548 15.431 13.548 15.431 make_m2s 222 7.7 0.008 0.011 13.331 13.755 make_images 222 8.7 0.099 0.109 13.309 13.732 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.802 12.385 make_images_data 222 9.7 0.004 0.005 7.912 8.450 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.594 7.051 8.314 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.590 8.245 hybrid_alltoall_any 227 10.6 0.215 1.840 6.798 8.129 dbcsr_mm_accdrv_process_sort 4760 11.4 6.292 7.202 6.292 7.202 calculate_norms 4752 9.8 5.525 6.208 5.525 6.208 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.034 5.195 mp_sum_l 807 5.4 3.113 4.457 3.113 4.457 make_images_sizes 222 9.7 0.000 0.000 0.766 3.936 mp_alltoall_i44 222 10.7 0.766 3.936 0.766 3.936 arnoldi_extremal 4 6.8 0.000 0.000 3.395 3.421 arnoldi_normal_ev 4 7.8 0.001 0.003 3.395 3.421 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.311 3.372 build_subspace 16 8.4 0.009 0.012 3.293 3.295 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.064 3.166 mp_irecv_dv 6231 10.9 2.045 3.143 2.045 3.143 ls_scf_post 1 4.0 0.000 0.000 3.050 3.064 ls_scf_store_result 1 5.0 0.000 0.000 2.869 2.911 dbcsr_special_finalize 555 9.7 0.005 0.005 2.303 2.681 dbcsr_merge_single_wm 555 10.7 0.455 0.608 2.295 2.673 make_images_pack 222 9.7 2.204 2.622 2.206 2.624 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.375 2.621 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.446 2.532 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.459 2.069 2.461 dbcsr_sort_data 658 11.4 2.098 2.425 2.098 2.425 buffer_matrices_ensure_size 222 8.7 1.749 2.105 1.749 2.105 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.854 1.855 rebuild_ks_matrix 3 7.3 0.000 0.000 1.844 1.845 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.844 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.143000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 2.111459E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.130 93.564 93.566 qs_energies 1 2.0 0.000 0.000 92.818 92.821 ls_scf 1 3.0 0.000 0.001 91.132 91.135 dbcsr_multiply_generic 111 6.7 0.016 0.025 76.378 76.749 multiply_cannon 111 7.7 0.029 0.045 53.769 57.659 ls_scf_main 1 4.0 0.000 0.002 56.063 56.063 multiply_cannon_loop 111 8.7 0.116 0.122 50.472 54.397 density_matrix_trs4 2 5.0 0.018 0.113 50.212 50.347 ls_scf_init_scf 1 4.0 0.000 0.001 31.441 31.443 mp_waitall_1 9105 10.9 22.522 31.368 22.522 31.368 ls_scf_init_matrix_S 1 5.0 0.000 0.002 30.071 30.171 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.608 27.626 multiply_cannon_multrec 1332 9.7 12.892 17.063 22.161 27.480 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.560 21.812 make_m2s 222 7.7 0.006 0.008 15.853 16.478 make_images 222 8.7 1.581 1.955 15.823 16.449 dbcsr_mm_accdrv_process 4041 10.4 0.265 0.448 8.872 10.444 make_images_data 222 9.7 0.004 0.005 9.283 10.256 dbcsr_mm_accdrv_process_sort 4041 11.4 8.468 9.996 8.468 9.996 hybrid_alltoall_any 227 10.6 0.521 2.442 8.716 9.846 mp_sum_l 807 5.4 5.475 8.782 5.475 8.782 multiply_cannon_metrocomm4 1221 9.7 0.006 0.009 3.228 7.835 mp_irecv_dv 3311 11.0 3.208 7.781 3.208 7.781 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.220 6.911 calculate_norms 2376 9.8 5.963 6.852 5.963 6.852 multiply_cannon_sync_h2d 1332 9.7 4.670 5.983 4.670 5.983 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.122 5.318 arnoldi_extremal 4 6.8 0.000 0.000 4.982 5.006 arnoldi_normal_ev 4 7.8 0.001 0.004 4.982 5.006 build_subspace 16 8.4 0.014 0.021 4.709 4.712 ls_scf_post 1 4.0 0.001 0.003 3.628 3.633 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.254 3.542 ls_scf_store_result 1 5.0 0.000 0.000 3.303 3.447 dbcsr_matrix_vector_mult_local 304 10.0 2.742 3.224 2.744 3.226 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.676 2.786 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.307 2.623 mp_allgather_i34 111 8.7 0.999 2.512 0.999 2.512 make_images_pack 222 9.7 2.020 2.351 2.022 2.353 dbcsr_sort_data 436 11.2 1.831 2.074 1.831 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.982 1.984 rebuild_ks_matrix 3 7.3 0.000 0.000 1.969 1.971 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.036 1.969 1.971 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.566000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1749.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 2.713543E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.046 95.285 95.286 qs_energies 1 2.0 0.000 0.000 94.538 94.541 ls_scf 1 3.0 0.000 0.000 93.129 93.134 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.566 77.841 ls_scf_main 1 4.0 0.000 0.000 58.379 58.383 multiply_cannon 111 7.7 0.043 0.105 53.607 57.869 multiply_cannon_loop 111 8.7 0.099 0.106 49.985 53.608 density_matrix_trs4 2 5.0 0.002 0.003 52.260 52.431 mp_waitall_1 7281 11.0 24.739 34.341 24.739 34.341 ls_scf_init_scf 1 4.0 0.000 0.000 31.115 31.117 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.910 29.971 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.522 27.533 multiply_cannon_multrec 888 9.7 12.564 15.221 21.165 24.366 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.563 23.309 make_m2s 222 7.7 0.006 0.006 17.265 18.527 make_images 222 8.7 1.975 2.296 17.227 18.488 make_images_data 222 9.7 0.003 0.004 9.993 11.011 hybrid_alltoall_any 227 10.6 0.621 2.861 9.685 10.885 dbcsr_mm_accdrv_process 3754 10.4 0.258 0.423 8.118 9.335 dbcsr_mm_accdrv_process_sort 3754 11.4 7.749 8.912 7.749 8.912 mp_sum_l 807 5.4 5.133 8.738 5.133 8.738 multiply_cannon_sync_h2d 888 9.7 6.019 7.365 6.019 7.365 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.471 7.157 mp_irecv_dv 2335 11.1 2.455 7.093 2.455 7.093 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.898 6.664 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.804 6.306 arnoldi_extremal 4 6.8 0.000 0.000 5.247 5.262 arnoldi_normal_ev 4 7.8 0.001 0.005 5.247 5.262 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.028 5.252 build_subspace 16 8.4 0.014 0.020 4.923 4.930 calculate_norms 1584 9.8 4.266 4.584 4.266 4.584 mp_allgather_i34 111 8.7 1.454 3.873 1.454 3.873 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.526 3.819 ls_scf_post 1 4.0 0.000 0.000 3.636 3.640 dbcsr_matrix_vector_mult_local 304 10.0 3.016 3.596 3.018 3.598 ls_scf_store_result 1 5.0 0.000 0.000 3.370 3.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.927 3.045 dbcsr_sort_data 325 11.1 1.887 2.171 1.887 2.171 make_images_pack 222 9.7 1.815 2.136 1.817 2.139 make_images_sizes 222 9.7 0.000 0.000 0.933 2.123 mp_alltoall_i44 222 10.7 0.933 2.122 0.933 2.122 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.959 1.960 rebuild_ks_matrix 3 7.3 0.000 0.000 1.940 1.942 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.940 1.942 dbcsr_data_release 9322 10.9 1.324 1.931 1.324 1.931 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.286000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2213.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 3.353178E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.065 98.475 98.476 qs_energies 1 2.0 0.000 0.000 97.271 97.276 ls_scf 1 3.0 0.000 0.000 95.626 95.631 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.902 79.184 ls_scf_main 1 4.0 0.000 0.000 59.505 59.506 multiply_cannon 111 7.7 0.049 0.116 51.757 56.644 density_matrix_trs4 2 5.0 0.002 0.003 53.219 53.345 multiply_cannon_loop 111 8.7 0.115 0.126 46.661 50.024 ls_scf_init_scf 1 4.0 0.000 0.000 32.861 32.862 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.632 31.691 mp_waitall_1 6369 11.0 23.114 29.727 23.114 29.727 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.150 29.165 multiply_cannon_multrec 1332 9.7 14.071 17.306 21.982 24.889 make_m2s 222 7.7 0.007 0.008 21.251 22.685 make_images 222 8.7 3.144 3.597 21.201 22.637 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.489 17.716 make_images_data 222 9.7 0.004 0.004 11.926 13.443 hybrid_alltoall_any 227 10.6 0.797 3.809 11.231 13.170 dbcsr_mm_accdrv_process 3641 10.4 0.237 0.408 7.564 9.113 dbcsr_mm_accdrv_process_sort 3641 11.4 7.188 8.696 7.188 8.696 mp_sum_l 807 5.4 4.347 7.868 4.347 7.868 multiply_cannon_sync_h2d 1332 9.7 5.487 6.255 5.487 6.255 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.413 6.207 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.071 5.948 mp_irecv_dv 3229 10.9 2.049 5.901 2.049 5.901 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.516 5.824 arnoldi_extremal 4 6.8 0.000 0.000 5.525 5.543 arnoldi_normal_ev 4 7.8 0.001 0.005 5.525 5.543 build_subspace 16 8.4 0.014 0.021 5.165 5.173 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.554 4.718 mp_allgather_i34 111 8.7 2.216 4.581 2.216 4.581 calculate_norms 2376 9.8 4.183 4.559 4.183 4.559 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.708 4.043 dbcsr_matrix_vector_mult_local 304 10.0 3.182 3.685 3.184 3.687 dbcsr_sort_data 658 11.4 3.042 3.341 3.042 3.341 ls_scf_post 1 4.0 0.000 0.000 3.259 3.264 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.174 3.246 dbcsr_special_finalize 555 9.7 0.006 0.007 2.793 3.212 dbcsr_merge_single_wm 555 10.7 0.531 0.665 2.784 3.203 ls_scf_store_result 1 5.0 0.000 0.000 2.996 3.066 dbcsr_data_release 10477 10.7 1.559 2.374 1.559 2.374 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.991 1.993 dbcsr_finalize 304 7.8 0.049 0.061 1.790 1.984 rebuild_ks_matrix 3 7.3 0.000 0.000 1.968 1.970 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.968 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.476000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 4.619039E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.062 99.192 99.193 qs_energies 1 2.0 0.000 0.000 98.347 98.351 ls_scf 1 3.0 0.000 0.000 96.311 96.315 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.489 77.719 ls_scf_main 1 4.0 0.000 0.000 61.977 61.978 multiply_cannon 111 7.7 0.093 0.154 55.088 59.590 density_matrix_trs4 2 5.0 0.002 0.003 54.756 54.835 multiply_cannon_loop 111 8.7 0.069 0.076 50.512 52.291 mp_waitall_1 5436 11.0 26.054 31.128 26.054 31.128 ls_scf_init_scf 1 4.0 0.000 0.000 30.735 30.737 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.365 29.401 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.110 27.123 multiply_cannon_multrec 444 9.7 14.048 17.075 21.165 23.009 make_m2s 222 7.7 0.004 0.005 17.679 20.225 make_images 222 8.7 3.732 4.438 17.618 20.165 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.731 16.029 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.253 14.367 make_images_data 222 9.7 0.003 0.004 9.870 12.361 hybrid_alltoall_any 227 10.6 0.787 3.771 9.612 12.175 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.341 6.821 7.945 multiply_cannon_sync_h2d 444 9.7 6.565 7.935 6.565 7.935 dbcsr_mm_accdrv_process_sort 3003 11.4 6.495 7.598 6.495 7.598 mp_allgather_i34 111 8.7 2.799 7.061 2.799 7.061 arnoldi_extremal 4 6.8 0.000 0.000 5.944 5.958 arnoldi_normal_ev 4 7.8 0.002 0.005 5.944 5.957 build_subspace 16 8.4 0.015 0.020 5.536 5.546 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.648 4.791 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.250 4.428 mp_sum_l 807 5.4 2.763 4.190 2.763 4.190 dbcsr_matrix_vector_mult_local 304 10.0 3.688 4.177 3.691 4.179 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.614 3.698 calculate_norms 792 9.8 3.571 3.688 3.571 3.688 ls_scf_post 1 4.0 0.000 0.000 3.598 3.603 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.585 3.597 mp_irecv_dv 1241 11.2 1.564 3.543 1.564 3.543 make_images_sizes 222 9.7 0.000 0.000 1.147 3.467 mp_alltoall_i44 222 10.7 1.147 3.467 1.147 3.467 ls_scf_store_result 1 5.0 0.000 0.000 3.377 3.426 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.836 3.124 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.263 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.161 2.162 rebuild_ks_matrix 3 7.3 0.000 0.000 2.128 2.128 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.127 2.128 dbcsr_merge_all 275 8.9 0.472 0.523 2.047 2.100 qs_energies_init_hamiltonians 1 3.0 0.102 0.120 2.018 2.018 dbcsr_data_release 10123 10.8 1.330 1.987 1.330 1.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.193000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3627.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/afe0d05cf1032bb96baeefe089b5bc565e4f53ca_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 8.796344E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.103 0.131 110.531 110.542 qs_energies 1 2.0 0.000 0.000 108.542 108.555 ls_scf 1 3.0 0.000 0.000 105.596 105.608 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.684 78.791 ls_scf_main 1 4.0 0.000 0.000 67.157 67.158 density_matrix_trs4 2 5.0 0.002 0.003 57.717 57.775 multiply_cannon 111 7.7 0.130 0.189 50.996 52.917 multiply_cannon_loop 111 8.7 0.067 0.070 47.460 48.523 ls_scf_init_scf 1 4.0 0.000 0.000 34.739 34.739 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.030 33.043 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.199 30.206 mp_waitall_1 4527 11.1 23.192 26.841 23.192 26.841 make_m2s 222 7.7 0.005 0.005 24.043 25.078 make_images 222 8.7 4.606 5.008 23.937 24.969 multiply_cannon_multrec 444 9.7 17.851 18.648 22.639 23.438 make_images_data 222 9.7 0.003 0.003 13.237 15.722 hybrid_alltoall_any 227 10.6 1.661 3.628 12.998 15.670 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.284 11.962 multiply_cannon_sync_h2d 444 9.7 8.846 8.884 8.846 8.884 arnoldi_extremal 4 6.8 0.000 0.000 7.552 7.559 arnoldi_normal_ev 4 7.8 0.003 0.009 7.551 7.559 build_subspace 16 8.4 0.026 0.036 6.989 7.000 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.543 5.707 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.372 5.461 dbcsr_matrix_vector_mult_local 304 10.0 5.015 5.333 5.017 5.336 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.050 5.319 dbcsr_mm_accdrv_process 1814 10.4 0.318 0.618 4.619 4.875 dbcsr_mm_accdrv_process_sort 1814 11.4 4.172 4.303 4.172 4.303 ls_scf_post 1 4.0 0.000 0.000 3.700 3.713 mp_allgather_i34 111 8.7 1.168 3.646 1.168 3.646 make_images_sizes 222 9.7 0.000 0.000 1.458 3.604 mp_alltoall_i44 222 10.7 1.458 3.604 1.458 3.604 ls_scf_store_result 1 5.0 0.000 0.000 3.444 3.450 calculate_norms 792 9.8 3.244 3.280 3.244 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.084 3.136 dbcsr_complete_redistribute 5 7.6 1.439 1.486 2.793 2.930 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.916 2.916 dbcsr_merge_all 275 8.9 0.891 0.920 2.869 2.911 dbcsr_data_release 12724 10.6 2.329 2.846 2.329 2.846 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.647 2.649 matrix_ls_to_qs 2 6.0 0.000 0.000 2.451 2.599 rebuild_ks_matrix 3 7.3 0.000 0.000 2.581 2.583 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.581 2.583 dbcsr_sort_data 325 11.1 2.442 2.503 2.442 2.503 dbcsr_new_transposed 4 7.5 0.243 0.251 2.279 2.294 dbcsr_frobenius_norm 74 6.6 2.054 2.132 2.203 2.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.542000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7023.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: afe0d05cf1032bb96baeefe089b5bc565e4f53ca Summary: empty Status: OK