=== 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: 78096bc50ec3c41c1825630b7cc5ec9938bfc15f ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/01 job id: 41546481 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/02 job id: 41546484 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/03 job id: 41546485 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/04 job id: 41546486 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/05 job id: 41546488 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/06 job id: 41546489 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/07 job id: 41546490 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/08 job id: 41546491 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/09 job id: 41546493 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/10 job id: 41546494 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/11 job id: 41546495 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/12 job id: 41546496 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/13 job id: 41546497 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/14 job id: 41546498 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/15 job id: 41546499 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/16 job id: 41546500 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/17 job id: 41546501 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/18 job id: 41546502 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/19 job id: 41546503 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/20 job id: 41546505 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/21 job id: 41546506 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/22 job id: 41546507 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/23 job id: 41546508 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/24 job id: 41546509 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/25 job id: 41546510 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/26 job id: 41546511 --- 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 0.023 0.038 136.898 136.899 farming_run 1 2.0 136.190 136.202 136.860 136.864 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.450643E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.032 117.856 117.857 qs_energies 1 2.0 0.000 0.000 117.566 117.569 mp2_main 1 3.0 0.000 0.000 115.382 115.385 mp2_gpw_main 1 4.0 0.026 0.033 114.318 114.321 mp2_ri_gpw_compute_in 1 5.0 0.172 0.186 95.252 95.636 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 56.250 56.637 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.168 42.580 47.897 get_2c_integrals 1 6.0 0.000 0.000 38.064 38.828 integrate_v_rspace 273 8.0 0.434 0.448 25.222 30.236 pw_transfer 6555 10.6 0.375 0.392 27.602 27.981 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.255 26.661 grid_integrate_task_list 273 9.0 20.977 26.512 20.977 26.512 fft_wrap_pw1pw2_100 2178 12.4 1.199 1.383 23.815 24.225 compute_2c_integrals 1 7.0 0.003 0.003 20.416 20.422 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.666 20.100 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.432 19.663 20.098 rpa_ri_compute_en 1 5.0 0.000 0.000 18.950 19.130 cp_fm_cholesky_decompose 12 8.2 17.714 18.511 17.714 18.511 cholesky_decomp 1 7.0 0.000 0.000 16.493 17.251 fft3d_s 5443 13.4 16.256 16.852 16.278 16.873 ao_to_mo_and_store_B_mult_1 272 7.0 10.841 15.554 10.841 15.554 calculate_wavefunction 272 8.0 5.402 5.560 12.580 13.234 rpa_num_int 1 6.0 0.000 0.000 10.782 10.782 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.657 10.695 calc_potential_gpw 544 9.5 0.004 0.005 9.669 10.022 calc_mat_Q 8 8.0 0.000 0.000 9.397 9.558 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.945 9.236 contract_S_to_Q 8 9.0 0.000 0.000 8.816 8.978 potential_pw2rs 545 10.0 0.106 0.108 7.765 8.476 parallel_gemm_fm 14 9.1 0.000 0.000 8.389 8.470 parallel_gemm_fm_cosma 14 10.1 8.389 8.470 8.389 8.470 collocate_single_gaussian 272 10.0 0.039 0.042 7.516 7.767 create_integ_mat 1 6.0 0.021 0.028 7.687 7.687 array2fm 1 7.0 0.000 0.000 6.744 7.097 pw_scatter_s 2720 13.7 4.434 4.596 4.434 4.596 pw_gather_s 2722 13.2 3.995 4.447 3.995 4.447 pw_poisson_solve 545 10.5 2.491 2.568 3.542 3.771 array2fm_buffer_send 1 8.0 3.050 3.173 3.050 3.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=114.320832, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2722.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 0.104 0.168 401.423 401.437 farming_run 1 2.0 400.352 400.439 401.244 401.340 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223963E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 8828893. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.053 211.411 211.411 qs_energies 1 2.0 0.000 0.000 211.162 211.176 scf_env_do_scf 1 3.0 0.000 0.000 106.465 106.466 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.485 105.492 rebuild_ks_matrix 4 6.0 0.000 0.000 105.483 105.490 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.063 105.483 105.490 hfx_ks_matrix 4 8.0 0.001 0.001 105.091 105.094 integrate_four_center 4 9.0 0.143 0.451 105.090 105.093 mp2_main 1 3.0 0.000 0.001 104.397 104.411 mp2_gpw_main 1 4.0 0.032 0.044 103.332 103.349 integrate_four_center_main 4 10.0 0.113 0.499 96.768 99.208 integrate_four_center_bin 265 11.0 96.655 99.022 96.655 99.022 init_scf_loop 1 4.0 0.000 0.000 92.232 92.232 mp2_ri_gpw_compute_in 1 5.0 0.070 0.123 76.170 77.323 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.010 55.130 56.287 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.158 42.660 47.598 integrate_v_rspace 95 8.0 0.396 0.560 28.583 33.487 pw_transfer 2240 10.6 0.147 0.182 30.197 30.673 ao_to_mo_and_store_B_mult_1 91 7.0 10.783 29.785 10.783 29.785 fft_wrap_pw1pw2 1868 11.4 0.018 0.023 29.173 29.647 grid_integrate_task_list 95 9.0 23.838 28.878 23.838 28.878 mp2_ri_gpw_compute_en 1 5.0 0.067 0.082 27.012 28.782 fft_wrap_pw1pw2_100 730 12.4 1.285 1.478 26.886 27.297 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.854 2.037 25.200 25.211 get_2c_integrals 1 6.0 0.013 0.112 20.863 20.971 compute_2c_integrals 1 7.0 0.022 0.088 19.823 19.828 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 19.332 19.604 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.876 19.331 19.603 fft3d_s 1823 13.4 18.541 18.940 18.555 18.953 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.229 14.229 calculate_wavefunction 91 8.0 2.039 2.066 9.826 10.148 potential_pw2rs 186 10.0 0.033 0.034 8.742 9.425 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.568 8.857 9.187 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.641 9.132 calc_potential_gpw 182 9.5 0.002 0.002 8.620 8.910 local_gemm 172 8.0 8.302 8.627 8.302 8.627 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.522 8.095 8.522 collocate_single_gaussian 91 10.0 0.017 0.034 7.947 8.275 mp_sync 38 10.4 3.072 7.133 3.072 7.133 mp_sendrecv_dm3 2068 8.0 6.125 6.563 6.125 6.563 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.433 6.349 6.433 pw_gather_s 912 13.2 5.010 5.524 5.010 5.524 pw_scatter_s 910 13.7 3.950 4.278 3.950 4.278 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=103.321106, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 450.580480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 885409. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.202 0.798 56.934 56.944 qs_mol_dyn_low 1 2.0 0.004 0.006 54.922 55.158 qs_forces 11 3.9 0.004 0.012 54.566 54.567 qs_energies 11 4.9 0.010 0.070 53.000 53.038 scf_env_do_scf 11 5.9 0.006 0.044 45.654 45.654 scf_env_do_scf_inner_loop 108 6.5 0.007 0.027 43.322 43.329 dbcsr_multiply_generic 2286 12.5 0.115 0.120 34.687 35.210 qs_scf_new_mos 108 7.5 0.000 0.001 32.823 33.123 qs_scf_loop_do_ot 108 8.5 0.001 0.001 32.823 33.123 ot_scf_mini 108 9.5 0.003 0.013 31.148 31.332 multiply_cannon 2286 13.5 0.185 0.192 26.541 28.193 multiply_cannon_loop 2286 14.5 1.508 1.581 25.788 27.394 velocity_verlet 10 3.0 0.005 0.016 26.952 26.958 ot_mini 108 10.5 0.004 0.015 19.902 20.155 qs_ot_get_derivative 108 11.5 0.001 0.001 16.937 17.138 mp_waitall_1 267858 16.1 9.004 15.638 9.004 15.638 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 6.152 13.412 multiply_cannon_multrec 54864 15.5 4.188 6.504 7.933 11.400 mp_sum_l 7207 12.9 6.592 8.587 6.592 8.587 rebuild_ks_matrix 119 8.3 0.000 0.000 8.186 8.336 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.033 8.186 8.336 multiply_cannon_sync_h2d 54864 15.5 5.864 7.637 5.864 7.637 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.231 7.362 qs_ot_get_p 119 10.4 0.001 0.001 6.524 6.826 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.785 6.205 init_scf_run 11 5.9 0.000 0.001 5.821 5.821 scf_env_initial_rho_setup 11 6.9 0.003 0.021 5.821 5.821 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.649 5.794 dbcsr_mm_accdrv_process 76910 16.1 1.185 1.832 3.668 5.038 sum_up_and_integrate 119 10.3 0.008 0.011 4.771 4.779 integrate_v_rspace 119 11.3 0.002 0.002 4.763 4.772 qs_rho_update_rho_low 119 7.7 0.006 0.024 4.405 4.539 calculate_rho_elec 119 8.7 0.011 0.016 4.399 4.539 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.482 3.544 calculate_first_density_matrix 1 7.0 0.005 0.033 3.513 3.527 calculate_dm_sparse 119 9.5 0.001 0.001 3.203 3.346 rs_pw_transfer 974 11.9 0.012 0.013 3.217 3.325 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.854 3.315 jit_kernel_multiply 13 15.8 2.421 3.117 2.421 3.117 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.915 3.106 apply_single 119 13.6 0.000 0.000 2.915 3.106 ot_diis_step 108 11.5 0.014 0.052 2.678 2.682 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.584 2.649 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.623 2.624 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.584 2.588 density_rs2pw 119 9.7 0.004 0.004 2.416 2.529 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.390 2.391 cp_fm_redistribute_end 50 14.0 2.170 2.367 2.175 2.369 cp_fm_diag_elpa_base 50 14.0 0.192 2.324 0.193 2.333 init_scf_loop 11 6.9 0.003 0.020 2.307 2.315 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.199 2.253 acc_transpose_blocks 54864 15.5 0.225 0.243 1.784 2.229 wfi_extrapolate 11 7.9 0.001 0.008 2.199 2.199 grid_integrate_task_list 119 12.3 2.022 2.133 2.022 2.133 potential_pw2rs 119 12.3 0.004 0.004 2.111 2.131 mp_sum_d 4129 12.0 1.461 2.051 1.461 2.051 make_m2s 4572 13.5 0.054 0.056 1.833 1.890 pw_transfer 1439 11.6 0.051 0.055 1.781 1.864 make_images 4572 14.5 0.133 0.139 1.750 1.806 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.706 1.791 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.699 1.727 mp_alltoall_d11v 2130 13.8 1.465 1.647 1.465 1.647 fft3d_ps 1201 14.6 0.359 0.460 1.482 1.559 mp_waitany 12084 13.8 1.258 1.412 1.258 1.412 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.302 1.385 grid_collocate_task_list 119 9.7 1.287 1.367 1.287 1.367 parallel_gemm_fm 81 9.0 0.000 0.000 1.326 1.331 parallel_gemm_fm_cosma 81 10.0 1.326 1.330 1.326 1.330 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=56.944000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.174144E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 71945. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.067 40.926 40.934 qs_mol_dyn_low 1 2.0 0.007 0.039 40.460 40.467 qs_forces 11 3.9 0.002 0.007 40.342 40.344 qs_energies 11 4.9 0.014 0.018 38.610 38.614 scf_env_do_scf 11 5.9 0.000 0.001 32.769 32.770 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 30.126 30.127 dbcsr_multiply_generic 2286 12.5 0.108 0.112 22.163 22.544 qs_scf_new_mos 108 7.5 0.001 0.001 20.429 20.681 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.429 20.681 ot_scf_mini 108 9.5 0.002 0.003 19.509 19.685 velocity_verlet 10 3.0 0.004 0.016 18.919 18.920 multiply_cannon 2286 13.5 0.208 0.217 16.806 18.385 multiply_cannon_loop 2286 14.5 0.906 0.977 15.608 17.034 ot_mini 108 10.5 0.001 0.001 12.247 12.482 mp_waitall_1 217478 16.2 6.602 11.986 6.602 11.986 qs_ot_get_derivative 108 11.5 0.002 0.012 9.767 9.945 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.315 9.717 multiply_cannon_multrec 27432 15.5 1.976 4.632 6.058 9.144 rebuild_ks_matrix 119 8.3 0.000 0.000 7.487 7.632 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.487 7.632 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.614 6.744 dbcsr_mm_accdrv_process 47894 16.0 3.245 5.834 4.013 5.963 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.783 4.674 qs_ot_get_p 119 10.4 0.001 0.001 4.337 4.568 init_scf_run 11 5.9 0.000 0.001 4.511 4.512 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.511 4.512 sum_up_and_integrate 119 10.3 0.015 0.016 4.473 4.480 integrate_v_rspace 119 11.3 0.002 0.002 4.458 4.464 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.105 4.219 apply_single 119 13.6 0.000 0.000 3.105 4.218 mp_sum_l 7207 12.9 2.237 4.194 2.237 4.194 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.050 4.106 calculate_rho_elec 119 8.7 0.021 0.024 4.050 4.106 rs_pw_transfer 974 11.9 0.010 0.011 3.045 3.117 make_m2s 4572 13.5 0.052 0.054 2.722 2.950 calculate_first_density_matrix 1 7.0 0.003 0.021 2.939 2.942 make_images 4572 14.5 0.200 0.237 2.633 2.861 multiply_cannon_sync_h2d 27432 15.5 2.179 2.854 2.179 2.854 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.766 2.784 init_scf_loop 11 6.9 0.000 0.000 2.620 2.621 ot_diis_step 108 11.5 0.011 0.011 2.430 2.431 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.258 2.349 density_rs2pw 119 9.7 0.004 0.004 2.266 2.328 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.318 2.319 cp_dbcsr_sm_fm_multiply 37 9.5 0.012 0.091 2.269 2.274 calculate_dm_sparse 119 9.5 0.001 0.002 2.149 2.227 potential_pw2rs 119 12.3 0.006 0.006 2.190 2.200 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.911 1.950 pw_transfer 1439 11.6 0.063 0.066 1.904 1.937 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.931 1.932 grid_integrate_task_list 119 12.3 1.835 1.917 1.835 1.917 cp_fm_redistribute_end 50 14.0 1.588 1.900 1.592 1.903 cp_fm_diag_elpa_base 50 14.0 0.301 1.836 0.308 1.866 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.815 1.849 acc_transpose_blocks 27432 15.5 0.107 0.111 1.299 1.709 prepare_preconditioner 11 7.9 0.000 0.000 1.660 1.687 make_preconditioner 11 8.9 0.000 0.000 1.660 1.687 mp_alltoall_d11v 2130 13.8 1.433 1.659 1.433 1.659 make_images_data 4572 15.5 0.045 0.051 1.219 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.547 1.606 hybrid_alltoall_any 4725 16.4 0.050 0.111 1.055 1.555 fft3d_ps 1201 14.6 0.501 0.555 1.520 1.550 jit_kernel_multiply 7 16.1 0.715 1.533 0.715 1.533 wfi_extrapolate 11 7.9 0.001 0.001 1.521 1.521 fft_wrap_pw1pw2_140 487 13.2 0.077 0.083 1.383 1.417 mp_allgather_i34 2286 14.5 0.639 1.401 0.639 1.401 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.349 1.357 grid_collocate_task_list 119 9.7 1.242 1.324 1.242 1.324 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.223 1.272 mp_sum_d 4129 12.0 0.641 1.097 0.641 1.097 acc_transpose_blocks_kernels 27432 16.5 0.184 0.275 0.769 1.091 make_images_sizes 4572 15.5 0.005 0.005 0.688 0.986 mp_alltoall_i44 4572 16.5 0.683 0.982 0.683 0.982 qs_energies_init_hamiltonians 11 5.9 0.004 0.012 0.973 0.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.935 0.951 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.607 0.855 0.913 mp_alltoall_z22v 1201 16.6 0.776 0.852 0.776 0.852 dbcsr_complete_redistribute 329 12.2 0.082 0.109 0.741 0.828 jit_kernel_transpose 5 15.5 0.586 0.821 0.586 0.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.934000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.636364, yerr=1.772144 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 521.981952E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.054 35.524 35.525 qs_mol_dyn_low 1 2.0 0.005 0.018 35.161 35.169 qs_forces 11 3.9 0.010 0.022 35.074 35.077 qs_energies 11 4.9 0.005 0.007 33.469 33.480 scf_env_do_scf 11 5.9 0.000 0.001 27.317 27.317 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.276 24.276 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.173 17.257 velocity_verlet 10 3.0 0.010 0.014 16.450 16.452 qs_scf_new_mos 108 7.5 0.001 0.001 15.605 15.632 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.605 15.632 ot_scf_mini 108 9.5 0.002 0.002 14.845 14.868 multiply_cannon 2286 13.5 0.195 0.201 13.404 14.082 multiply_cannon_loop 2286 14.5 0.640 0.673 12.504 13.227 ot_mini 108 10.5 0.001 0.001 9.222 9.245 qs_ot_get_derivative 108 11.5 0.001 0.001 7.659 7.682 multiply_cannon_multrec 18288 15.5 1.952 2.871 6.862 7.171 rebuild_ks_matrix 119 8.3 0.000 0.000 6.823 6.850 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.823 6.849 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.062 6.084 dbcsr_mm_accdrv_process 38222 16.0 4.020 5.366 4.826 5.811 mp_waitall_1 169478 16.3 3.613 4.651 3.613 4.651 init_scf_run 11 5.9 0.000 0.001 4.498 4.498 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.498 4.498 sum_up_and_integrate 119 10.3 0.017 0.018 4.333 4.339 integrate_v_rspace 119 11.3 0.002 0.003 4.316 4.323 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.614 3.621 calculate_rho_elec 119 8.7 0.030 0.031 3.613 3.620 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.881 3.522 qs_ot_get_p 119 10.4 0.001 0.001 3.394 3.432 rs_pw_transfer 974 11.9 0.009 0.010 3.129 3.280 calculate_first_density_matrix 1 7.0 0.000 0.002 3.142 3.143 init_scf_loop 11 6.9 0.000 0.000 3.019 3.020 jit_kernel_multiply 13 16.2 0.754 2.526 0.754 2.526 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.475 2.450 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.070 2.415 apply_single 119 13.6 0.000 0.000 2.070 2.415 density_rs2pw 119 9.7 0.004 0.004 2.187 2.313 make_m2s 4572 13.5 0.045 0.046 2.113 2.274 make_images 4572 14.5 0.190 0.202 2.027 2.187 potential_pw2rs 119 12.3 0.007 0.008 2.089 2.108 prepare_preconditioner 11 7.9 0.000 0.000 2.092 2.094 make_preconditioner 11 8.9 0.000 0.000 2.092 2.094 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.062 2.066 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.930 2.020 calculate_dm_sparse 119 9.5 0.001 0.001 1.919 1.933 grid_integrate_task_list 119 12.3 1.798 1.903 1.798 1.903 pw_transfer 1439 11.6 0.063 0.066 1.884 1.896 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.843 1.844 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.794 1.806 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.777 1.778 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.725 1.738 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.663 1.672 mp_sum_l 7207 12.9 1.277 1.672 1.277 1.672 multiply_cannon_sync_h2d 18288 15.5 1.372 1.557 1.372 1.557 ot_diis_step 108 11.5 0.011 0.011 1.535 1.536 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.513 1.513 fft3d_ps 1201 14.6 0.507 0.522 1.481 1.496 cp_fm_redistribute_end 50 14.0 1.117 1.480 1.118 1.481 cp_fm_diag_elpa_base 50 14.0 0.346 1.402 0.360 1.442 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.359 1.370 acc_transpose_blocks 18288 15.5 0.073 0.075 1.300 1.356 fft_wrap_pw1pw2_140 487 13.2 0.086 0.090 1.340 1.353 wfi_extrapolate 11 7.9 0.001 0.001 1.299 1.299 grid_collocate_task_list 119 9.7 1.203 1.292 1.203 1.292 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.261 parallel_gemm_fm 81 9.0 0.000 0.000 1.148 1.152 parallel_gemm_fm_cosma 81 10.0 1.148 1.152 1.148 1.152 make_images_data 4572 15.5 0.044 0.049 0.927 1.130 make_basis_sm 11 9.8 0.001 0.002 1.046 1.048 rs_pw_transfer_RS2PW_140 130 11.5 0.177 0.189 0.862 1.012 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.796 1.002 mp_alltoall_d11v 2130 13.8 0.838 0.987 0.838 0.987 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.898 0.976 mp_waitany 9880 13.7 0.829 0.969 0.829 0.969 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.939 0.966 cp_fm_cholesky_invert 11 10.9 0.929 0.934 0.929 0.934 acc_transpose_blocks_kernels 18288 16.5 0.213 0.221 0.861 0.902 mp_alltoall_z22v 1201 16.6 0.793 0.852 0.793 0.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.838 mp_allgather_i34 2286 14.5 0.359 0.761 0.359 0.761 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.657 0.712 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.525000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.454545, yerr=2.675262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.259008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 38.148 38.149 qs_mol_dyn_low 1 2.0 0.003 0.004 37.357 37.795 qs_forces 11 3.9 0.002 0.002 37.274 37.275 qs_energies 11 4.9 0.001 0.002 35.542 35.547 scf_env_do_scf 11 5.9 0.000 0.001 29.550 29.552 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.852 25.853 dbcsr_multiply_generic 2286 12.5 0.101 0.103 19.417 19.534 velocity_verlet 10 3.0 0.004 0.014 18.833 18.836 qs_scf_new_mos 108 7.5 0.001 0.001 17.126 17.190 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.125 17.189 ot_scf_mini 108 9.5 0.002 0.003 16.153 16.206 multiply_cannon 2286 13.5 0.229 0.268 15.316 15.712 multiply_cannon_loop 2286 14.5 0.937 0.971 14.277 14.556 ot_mini 108 10.5 0.001 0.001 9.797 9.864 multiply_cannon_multrec 27432 15.5 2.382 3.076 9.274 9.665 dbcsr_mm_accdrv_process 47916 15.9 5.425 7.467 6.797 8.132 qs_ot_get_derivative 108 11.5 0.001 0.001 7.959 8.015 rebuild_ks_matrix 119 8.3 0.000 0.000 6.813 6.863 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.813 6.863 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.076 6.119 init_scf_run 11 5.9 0.000 0.001 4.612 4.613 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.612 4.612 sum_up_and_integrate 119 10.3 0.020 0.022 4.045 4.053 integrate_v_rspace 119 11.3 0.002 0.002 4.025 4.033 qs_ot_get_p 119 10.4 0.001 0.001 3.660 3.735 init_scf_loop 11 6.9 0.000 0.000 3.673 3.673 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.638 3.673 calculate_rho_elec 119 8.7 0.040 0.046 3.637 3.672 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.061 3.445 mp_waitall_1 145218 16.4 2.726 3.357 2.726 3.357 calculate_first_density_matrix 1 7.0 0.000 0.000 3.043 3.045 prepare_preconditioner 11 7.9 0.000 0.000 2.788 2.796 make_preconditioner 11 8.9 0.000 0.000 2.788 2.796 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.379 2.722 make_m2s 4572 13.5 0.054 0.056 2.600 2.705 rs_pw_transfer 974 11.9 0.009 0.010 2.517 2.664 make_images 4572 14.5 0.272 0.331 2.492 2.594 jit_kernel_multiply 10 15.9 1.309 2.508 1.309 2.508 calculate_dm_sparse 119 9.5 0.000 0.000 2.294 2.345 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.335 2.337 density_rs2pw 119 9.7 0.003 0.004 2.112 2.257 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.222 2.232 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.043 2.175 apply_single 119 13.6 0.000 0.000 2.042 2.175 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.081 2.096 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.059 2.087 pw_transfer 1439 11.6 0.064 0.066 1.885 1.916 grid_integrate_task_list 119 12.3 1.825 1.898 1.825 1.898 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.883 1.883 potential_pw2rs 119 12.3 0.009 0.010 1.829 1.833 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.795 1.829 ot_diis_step 108 11.5 0.012 0.012 1.796 1.797 acc_transpose_blocks 27432 15.5 0.111 0.113 1.569 1.608 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.586 1.588 cp_fm_redistribute_end 50 14.0 1.045 1.551 1.047 1.553 cp_fm_diag_elpa_base 50 14.0 0.481 1.472 0.502 1.512 wfi_extrapolate 11 7.9 0.001 0.001 1.511 1.511 fft3d_ps 1201 14.6 0.531 0.581 1.475 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.441 1.449 fft_wrap_pw1pw2_140 487 13.2 0.086 0.094 1.405 1.440 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.768 1.392 mp_sum_l 7207 12.9 1.081 1.381 1.081 1.381 grid_collocate_task_list 119 9.7 1.220 1.313 1.220 1.313 dbcsr_complete_redistribute 329 12.2 0.125 0.175 0.926 1.209 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.160 1.175 make_images_data 4572 15.5 0.045 0.049 1.014 1.160 cp_fm_upper_to_full 72 13.5 0.824 1.149 0.824 1.149 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.093 1.094 hybrid_alltoall_any 4725 16.4 0.061 0.149 0.858 1.055 multiply_cannon_sync_h2d 27432 15.5 0.980 1.040 0.980 1.040 mp_alltoall_d11v 2130 13.8 0.869 1.024 0.869 1.024 acc_transpose_blocks_kernels 27432 16.5 0.271 0.277 0.943 0.980 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.663 0.939 cp_fm_cholesky_invert 11 10.9 0.873 0.877 0.873 0.877 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.870 mp_alltoall_z22v 1201 16.6 0.827 0.851 0.827 0.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.829 0.834 mp_alltoall_i22 627 13.8 0.457 0.764 0.457 0.764 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.149000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.545455, yerr=2.807899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 593.125376E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.092 0.123 31.407 31.407 qs_mol_dyn_low 1 2.0 0.005 0.033 30.744 30.751 qs_forces 11 3.9 0.039 0.049 30.585 30.587 qs_energies 11 4.9 0.036 0.038 28.820 28.834 scf_env_do_scf 11 5.9 0.000 0.001 22.719 22.719 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.917 19.918 velocity_verlet 10 3.0 0.002 0.002 15.061 15.070 dbcsr_multiply_generic 2286 12.5 0.112 0.114 13.554 13.683 qs_scf_new_mos 108 7.5 0.001 0.001 11.974 11.998 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.974 11.998 ot_scf_mini 108 9.5 0.002 0.002 11.270 11.299 multiply_cannon 2286 13.5 0.231 0.243 10.601 11.271 multiply_cannon_loop 2286 14.5 0.331 0.344 9.620 9.845 multiply_cannon_multrec 9144 15.5 1.638 1.839 6.182 6.416 ot_mini 108 10.5 0.001 0.001 6.320 6.355 rebuild_ks_matrix 119 8.3 0.000 0.000 6.096 6.122 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.096 6.122 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.448 5.472 qs_ot_get_derivative 108 11.5 0.001 0.001 4.929 4.956 dbcsr_mm_accdrv_process 12550 15.8 3.518 4.511 4.444 4.553 init_scf_run 11 5.9 0.000 0.001 4.429 4.429 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.428 4.429 sum_up_and_integrate 119 10.3 0.022 0.023 3.733 3.757 integrate_v_rspace 119 11.3 0.003 0.003 3.711 3.735 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.404 3.411 calculate_rho_elec 119 8.7 0.059 0.061 3.403 3.411 calculate_first_density_matrix 1 7.0 0.000 0.000 3.253 3.254 qs_ot_get_p 119 10.4 0.001 0.001 3.126 3.172 init_scf_loop 11 6.9 0.000 0.000 2.780 2.781 mp_waitall_1 121218 16.5 2.255 2.737 2.255 2.737 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.374 2.375 make_m2s 4572 13.5 0.034 0.035 2.042 2.242 make_images 4572 14.5 0.266 0.301 1.952 2.151 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.050 2.052 rs_pw_transfer 974 11.9 0.008 0.009 1.917 2.010 calculate_dm_sparse 119 9.5 0.000 0.000 1.977 1.998 jit_kernel_multiply 9 15.9 0.887 1.963 0.887 1.963 prepare_preconditioner 11 7.9 0.000 0.000 1.955 1.961 make_preconditioner 11 8.9 0.000 0.000 1.955 1.961 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.912 1.924 grid_integrate_task_list 119 12.3 1.854 1.908 1.854 1.908 density_rs2pw 119 9.7 0.003 0.004 1.804 1.888 pw_transfer 1439 11.6 0.064 0.065 1.849 1.875 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.837 1.873 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.844 1.845 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.759 1.784 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.676 1.690 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.528 1.529 potential_pw2rs 119 12.3 0.010 0.011 1.501 1.525 cp_fm_redistribute_end 50 14.0 0.759 1.493 0.761 1.494 cp_fm_diag_elpa_base 50 14.0 0.686 1.410 0.731 1.469 fft3d_ps 1201 14.6 0.541 0.553 1.425 1.453 fft_wrap_pw1pw2_140 487 13.2 0.084 0.087 1.402 1.425 qs_energies_init_hamiltonians 11 5.9 0.035 0.057 1.376 1.388 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.370 1.375 ot_diis_step 108 11.5 0.012 0.013 1.368 1.368 grid_collocate_task_list 119 9.7 1.277 1.357 1.277 1.357 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.297 1.311 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.222 1.252 apply_single 119 13.6 0.000 0.000 1.222 1.252 acc_transpose_blocks 9144 15.5 0.038 0.039 1.199 1.213 make_images_data 4572 15.5 0.039 0.042 0.939 1.197 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.902 1.196 wfi_extrapolate 11 7.9 0.001 0.001 1.130 1.130 cp_fm_cholesky_invert 11 10.9 1.043 1.047 1.043 1.047 mp_alltoall_d11v 2130 13.8 0.868 0.991 0.868 0.991 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.957 0.969 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.922 dbcsr_complete_redistribute 329 12.2 0.255 0.446 0.846 0.876 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.429 0.867 jit_kernel_transpose 5 15.6 0.840 0.853 0.840 0.853 mp_alltoall_z22v 1201 16.6 0.757 0.815 0.757 0.815 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.758 0.807 multiply_cannon_sync_h2d 9144 15.5 0.714 0.801 0.714 0.801 mp_allgather_i34 2286 14.5 0.293 0.793 0.293 0.793 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.771 0.774 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.760 0.768 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.498 0.710 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.407000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=564.454545, yerr=2.571297 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 730.927104E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.064 43.642 43.643 qs_mol_dyn_low 1 2.0 0.003 0.004 43.381 43.389 qs_forces 11 3.9 0.001 0.002 42.967 42.968 qs_energies 11 4.9 0.002 0.002 40.966 40.970 scf_env_do_scf 11 5.9 0.001 0.001 34.368 34.368 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.247 26.248 velocity_verlet 10 3.0 0.015 0.021 23.913 23.962 dbcsr_multiply_generic 2286 12.5 0.102 0.104 18.525 18.755 qs_scf_new_mos 108 7.5 0.001 0.001 16.933 17.027 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.932 17.026 ot_scf_mini 108 9.5 0.002 0.002 15.806 15.900 multiply_cannon 2286 13.5 0.306 0.313 14.160 15.216 multiply_cannon_loop 2286 14.5 0.344 0.348 12.845 13.852 ot_mini 108 10.5 0.001 0.001 9.409 9.524 multiply_cannon_multrec 9144 15.5 3.456 4.799 8.675 8.807 init_scf_loop 11 6.9 0.000 0.000 8.091 8.092 qs_ot_get_derivative 108 11.5 0.001 0.001 7.354 7.451 rebuild_ks_matrix 119 8.3 0.000 0.000 7.120 7.257 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.119 7.257 prepare_preconditioner 11 7.9 0.000 0.000 7.114 7.126 make_preconditioner 11 8.9 0.000 0.000 7.114 7.126 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.671 6.999 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.440 6.565 dbcsr_mm_accdrv_process 12550 15.8 4.115 5.622 5.096 6.373 init_scf_run 11 5.9 0.000 0.001 4.548 4.548 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.547 4.547 cp_fm_upper_to_full 72 14.2 3.161 4.542 3.161 4.542 mp_waitall_1 97218 16.6 3.099 3.984 3.099 3.984 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.891 3.896 calculate_rho_elec 119 8.7 0.118 0.121 3.891 3.896 sum_up_and_integrate 119 10.3 0.038 0.039 3.853 3.864 integrate_v_rspace 119 11.3 0.003 0.004 3.815 3.826 qs_ot_get_p 119 10.4 0.001 0.001 3.553 3.687 calculate_first_density_matrix 1 7.0 0.000 0.000 3.112 3.113 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.602 3.108 dbcsr_complete_redistribute 329 12.2 0.288 0.294 2.039 2.846 make_m2s 4572 13.5 0.038 0.038 2.576 2.771 make_images 4572 14.5 0.352 0.383 2.455 2.651 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.285 2.606 apply_single 119 13.6 0.000 0.000 2.285 2.605 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.691 2.497 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.506 2.354 calculate_dm_sparse 119 9.5 0.000 0.000 2.252 2.269 mp_alltoall_i22 627 13.8 1.427 2.262 1.427 2.262 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.438 2.238 pw_transfer 1439 11.6 0.067 0.068 2.184 2.195 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.174 2.177 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.089 2.099 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.086 2.088 grid_integrate_task_list 119 12.3 2.046 2.063 2.046 2.063 ot_diis_step 108 11.5 0.014 0.014 2.026 2.027 mp_sum_l 7207 12.9 1.259 2.027 1.259 2.027 density_rs2pw 119 9.7 0.003 0.003 1.977 1.994 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.844 1.889 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.875 1.876 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.822 1.871 jit_kernel_multiply 10 15.5 0.954 1.772 0.954 1.772 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.768 fft3d_ps 1201 14.6 0.568 0.579 1.723 1.734 fft_wrap_pw1pw2_140 487 13.2 0.086 0.088 1.713 1.723 cp_fm_cholesky_invert 11 10.9 1.663 1.667 1.663 1.667 rs_pw_transfer 974 11.9 0.009 0.009 1.615 1.631 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.545 1.547 cp_fm_diag_elpa_base 50 14.0 1.398 1.452 1.543 1.545 make_images_data 4572 15.5 0.043 0.045 1.259 1.527 grid_collocate_task_list 119 9.7 1.479 1.504 1.479 1.504 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.210 1.491 wfi_extrapolate 11 7.9 0.001 0.001 1.378 1.378 potential_pw2rs 119 12.3 0.014 0.014 1.368 1.377 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.279 1.288 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.199 1.225 acc_transpose_blocks 9144 15.5 0.038 0.039 1.099 1.208 mp_alltoall_d11v 2130 13.8 1.110 1.138 1.110 1.138 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.098 1.118 make_basis_sm 11 9.8 0.014 0.029 1.051 1.052 multiply_cannon_sync_h2d 9144 15.5 1.041 1.044 1.041 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.988 1.036 mp_alltoall_z22v 1201 16.6 1.018 1.032 1.018 1.032 qs_create_task_list 11 7.9 0.000 0.001 0.949 0.962 generate_qs_task_list 11 8.9 0.373 0.393 0.949 0.961 acc_transpose_blocks_kernels 9144 16.5 0.119 0.120 0.849 0.958 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.892 0.904 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.643000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=689.363636, yerr=11.795783 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 504.197120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1941519. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.044 84.262 84.263 qs_mol_dyn_low 1 2.0 0.003 0.007 83.776 83.786 qs_forces 11 3.9 0.013 0.026 83.707 83.708 qs_energies 11 4.9 0.002 0.002 80.834 80.853 scf_env_do_scf 11 5.9 0.000 0.001 70.815 70.817 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.153 65.154 dbcsr_multiply_generic 2055 12.4 0.134 0.145 52.616 52.861 qs_scf_new_mos 99 7.5 0.000 0.001 48.023 48.133 qs_scf_loop_do_ot 99 8.5 0.000 0.001 48.023 48.132 ot_scf_mini 99 9.5 0.002 0.002 45.611 45.734 multiply_cannon 2055 13.4 0.190 0.210 43.088 44.002 multiply_cannon_loop 2055 14.4 1.572 1.605 42.098 43.029 velocity_verlet 10 3.0 0.001 0.001 41.981 41.982 ot_mini 99 10.5 0.001 0.001 27.627 27.766 qs_ot_get_derivative 99 11.5 0.001 0.001 20.731 20.857 multiply_cannon_multrec 49320 15.4 12.079 12.829 17.449 18.458 rebuild_ks_matrix 110 8.3 0.000 0.001 14.586 14.705 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.585 14.704 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.813 12.907 mp_waitall_1 241148 16.1 11.713 12.576 11.713 12.576 multiply_cannon_sync_h2d 49320 15.4 9.960 10.453 9.960 10.453 qs_ot_get_p 110 10.4 0.001 0.001 9.500 9.666 multiply_cannon_metrocomm3 49320 15.4 0.075 0.078 6.838 7.933 init_scf_run 11 5.9 0.000 0.001 7.899 7.900 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.899 7.899 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.154 7.601 apply_single 110 13.6 0.000 0.000 7.153 7.601 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.053 7.571 sum_up_and_integrate 110 10.3 0.023 0.028 7.165 7.180 integrate_v_rspace 110 11.3 0.002 0.003 7.142 7.163 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.685 6.747 ot_diis_step 99 11.5 0.006 0.006 6.572 6.572 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.299 6.435 calculate_rho_elec 110 8.6 0.020 0.024 6.299 6.435 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.165 6.188 mp_sum_l 6514 12.8 4.983 5.957 4.983 5.957 init_scf_loop 11 6.9 0.000 0.000 5.632 5.633 dbcsr_mm_accdrv_process 87628 16.1 2.100 2.185 5.250 5.529 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.093 5.094 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.580 4.582 cp_fm_redistribute_end 48 14.0 3.988 4.553 3.992 4.554 cp_fm_diag_elpa_base 48 14.0 0.555 4.453 0.559 4.480 rs_pw_transfer 902 11.9 0.012 0.014 4.190 4.337 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.176 4.180 make_m2s 4110 13.4 0.061 0.066 4.064 4.162 calculate_dm_sparse 110 9.5 0.000 0.001 3.980 4.086 wfi_extrapolate 11 7.9 0.001 0.001 4.082 4.082 make_images 4110 14.4 0.179 0.192 3.968 4.069 multiply_cannon_metrocomm1 49320 15.4 0.060 0.064 2.701 3.879 calculate_first_density_matrix 1 7.0 0.000 0.000 3.729 3.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.656 3.695 density_rs2pw 110 9.6 0.004 0.004 3.448 3.599 prepare_preconditioner 11 7.9 0.000 0.000 3.473 3.494 make_preconditioner 11 8.9 0.000 0.000 3.473 3.494 grid_integrate_task_list 110 12.3 3.254 3.395 3.254 3.395 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.257 3.324 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.241 3.278 jit_kernel_multiply 13 15.9 2.865 2.945 2.865 2.945 pw_transfer 1331 11.6 0.054 0.062 2.886 2.943 potential_pw2rs 110 12.3 0.006 0.006 2.835 2.853 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.794 2.853 acc_transpose_blocks 49320 15.4 0.213 0.223 2.502 2.594 mp_alltoall_d11v 2046 13.8 1.977 2.485 1.977 2.485 fft3d_ps 1111 14.6 0.756 0.843 2.333 2.382 fft_wrap_pw1pw2_140 451 13.1 0.168 0.186 2.302 2.363 grid_collocate_task_list 110 9.6 2.083 2.195 2.083 2.195 mp_sum_d 3883 11.9 1.538 2.083 1.538 2.083 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.913 1.939 make_images_data 4110 15.4 0.043 0.047 1.785 1.929 cp_fm_cholesky_invert 11 10.9 1.898 1.902 1.898 1.902 mp_waitany 14300 13.8 1.716 1.893 1.716 1.893 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.847 1.885 hybrid_alltoall_any 4261 16.3 0.081 0.476 1.557 1.797 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.263000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.923261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 584.445952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2085285. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 69.200 69.201 qs_mol_dyn_low 1 2.0 0.003 0.004 68.924 68.933 qs_forces 11 3.9 0.002 0.002 68.126 68.126 qs_energies 11 4.9 0.002 0.002 64.828 64.832 scf_env_do_scf 11 5.9 0.000 0.001 55.787 55.791 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.238 48.238 dbcsr_multiply_generic 2055 12.4 0.123 0.130 38.035 38.190 velocity_verlet 10 3.0 0.001 0.002 36.507 36.524 multiply_cannon 2055 13.4 0.220 0.248 31.429 32.648 qs_scf_new_mos 99 7.5 0.001 0.001 32.392 32.519 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.392 32.518 multiply_cannon_loop 2055 14.4 0.932 0.948 30.144 31.057 ot_scf_mini 99 9.5 0.003 0.004 30.738 30.860 ot_mini 99 10.5 0.001 0.001 18.206 18.333 multiply_cannon_multrec 24660 15.4 7.616 9.309 14.262 15.782 rebuild_ks_matrix 110 8.3 0.000 0.001 13.631 13.723 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.631 13.723 qs_ot_get_derivative 99 11.5 0.001 0.001 12.435 12.557 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.020 12.106 mp_waitall_1 186928 16.3 7.753 10.320 7.753 10.320 multiply_cannon_sync_h2d 24660 15.4 6.963 8.293 6.963 8.293 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.178 7.794 init_scf_loop 11 6.9 0.000 0.000 7.514 7.514 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.530 7.353 apply_single 110 13.6 0.000 0.001 6.530 7.353 dbcsr_mm_accdrv_process 52282 16.1 4.828 6.214 6.485 6.772 init_scf_run 11 5.9 0.000 0.001 6.687 6.687 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.687 6.687 sum_up_and_integrate 110 10.3 0.028 0.033 6.554 6.567 integrate_v_rspace 110 11.3 0.002 0.003 6.525 6.538 qs_ot_get_p 110 10.4 0.001 0.001 5.944 6.088 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.768 5.777 calculate_rho_elec 110 8.6 0.039 0.047 5.768 5.776 ot_diis_step 99 11.5 0.010 0.010 5.722 5.722 prepare_preconditioner 11 7.9 0.000 0.000 5.499 5.518 make_preconditioner 11 8.9 0.000 0.000 5.499 5.518 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.711 5.431 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.084 5.247 make_m2s 4110 13.4 0.057 0.060 4.144 4.636 make_images 4110 14.4 0.395 0.441 4.036 4.524 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.060 4.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.700 3.703 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.634 3.634 wfi_extrapolate 11 7.9 0.001 0.001 3.499 3.499 pw_transfer 1331 11.6 0.066 0.071 3.251 3.385 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.286 3.337 grid_integrate_task_list 110 12.3 3.155 3.333 3.155 3.333 density_rs2pw 110 9.6 0.004 0.004 3.072 3.279 rs_pw_transfer 902 11.9 0.012 0.013 3.068 3.278 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.140 3.276 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.157 3.219 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.115 3.117 calculate_first_density_matrix 1 7.0 0.000 0.001 3.104 3.105 cp_fm_redistribute_end 48 14.0 2.325 3.091 2.327 3.092 calculate_dm_sparse 110 9.5 0.001 0.001 3.040 3.068 cp_fm_diag_elpa_base 48 14.0 0.731 2.967 0.762 3.050 make_images_data 4110 15.4 0.047 0.051 2.328 2.828 fft_wrap_pw1pw2_140 451 13.1 0.201 0.217 2.658 2.792 hybrid_alltoall_any 4261 16.3 0.101 0.443 2.023 2.792 fft3d_ps 1111 14.6 1.069 1.266 2.496 2.621 cp_fm_cholesky_invert 11 10.9 2.557 2.565 2.557 2.565 mp_sum_l 6514 12.8 1.824 2.448 1.824 2.448 potential_pw2rs 110 12.3 0.008 0.009 2.439 2.447 grid_collocate_task_list 110 9.6 2.119 2.244 2.119 2.244 jit_kernel_multiply 9 16.4 1.303 2.084 1.303 2.084 mp_alltoall_d11v 2046 13.8 1.738 1.993 1.738 1.993 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.938 1.957 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.851 1.852 multiply_cannon_metrocomm4 22605 15.4 0.074 0.079 0.804 1.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.773 1.785 mp_irecv_dv 57340 16.2 0.679 1.709 0.679 1.709 acc_transpose_blocks 24660 15.4 0.105 0.109 1.563 1.657 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.618 1.628 mp_allgather_i34 2055 14.4 0.615 1.626 0.615 1.626 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.477 dbcsr_complete_redistribute 325 12.2 0.294 0.371 1.197 1.467 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.201000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.454545, yerr=5.990350 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 658.472960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.194 0.433 74.356 74.357 qs_mol_dyn_low 1 2.0 0.009 0.049 72.395 72.405 qs_forces 11 3.9 0.025 0.062 72.246 72.275 qs_energies 11 4.9 0.057 0.192 68.985 69.041 scf_env_do_scf 11 5.9 0.003 0.012 51.480 51.481 scf_env_do_scf_inner_loop 99 6.5 0.016 0.087 41.929 41.932 dbcsr_multiply_generic 2055 12.4 0.152 0.305 34.119 34.588 velocity_verlet 10 3.0 0.011 0.017 34.067 34.069 multiply_cannon 2055 13.4 0.228 0.252 25.976 27.299 qs_scf_new_mos 99 7.5 0.001 0.001 26.667 26.800 qs_scf_loop_do_ot 99 8.5 0.006 0.021 26.666 26.799 multiply_cannon_loop 2055 14.4 0.619 0.634 24.510 25.697 ot_scf_mini 99 9.5 0.014 0.069 25.407 25.557 ot_mini 99 10.5 0.001 0.001 14.779 14.944 rebuild_ks_matrix 110 8.3 0.000 0.000 12.654 12.814 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.654 12.813 init_scf_run 11 5.9 0.000 0.001 11.714 11.718 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.714 11.718 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.193 11.340 multiply_cannon_multrec 16440 15.4 3.893 4.947 10.071 11.286 mp_waitall_1 146766 16.3 7.863 11.083 7.863 11.083 qs_ot_get_derivative 99 11.5 0.001 0.001 10.215 10.379 init_scf_loop 11 6.9 0.009 0.035 9.507 9.508 calculate_first_density_matrix 1 7.0 0.007 0.036 8.420 8.425 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.498 7.706 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 7.507 7.510 prepare_preconditioner 11 7.9 0.000 0.000 7.381 7.400 make_preconditioner 11 8.9 0.010 0.040 7.381 7.400 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.654 7.019 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.624 6.762 sum_up_and_integrate 110 10.3 0.033 0.034 6.741 6.756 integrate_v_rspace 110 11.3 0.002 0.003 6.708 6.723 dbcsr_mm_accdrv_process 34862 16.1 5.225 5.915 6.033 6.284 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.062 6.086 calculate_rho_elec 110 8.6 0.058 0.058 6.061 6.085 qs_ot_get_p 110 10.4 0.001 0.001 5.878 6.040 rs_pw_transfer 902 11.9 0.010 0.011 5.040 5.732 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.041 5.471 apply_single 110 13.6 0.000 0.000 5.040 5.471 make_m2s 4110 13.4 0.050 0.053 5.007 5.447 make_images 4110 14.4 0.411 0.680 4.891 5.331 qs_energies_init_hamiltonians 11 5.9 0.033 0.119 4.866 4.901 ot_diis_step 99 11.5 0.011 0.011 4.520 4.521 multiply_cannon_sync_h2d 16440 15.4 3.680 4.259 3.680 4.259 acc_transpose_blocks 16440 15.4 0.071 0.074 4.171 4.247 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.526 4.187 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.825 3.834 acc_transpose_blocks_kernels 16440 16.4 0.209 0.213 3.749 3.826 mp_waitany 17072 13.8 3.009 3.683 3.009 3.683 jit_kernel_transpose 5 15.6 3.539 3.616 3.539 3.616 qs_env_update_s_mstruct 11 6.9 0.004 0.027 3.193 3.576 rs_pw_transfer_RS2PW_140 121 11.5 0.273 0.323 2.784 3.473 pw_transfer 1331 11.6 0.066 0.073 3.452 3.463 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.437 3.437 density_rs2pw 110 9.6 0.004 0.004 3.220 3.427 grid_integrate_task_list 110 12.3 3.190 3.425 3.190 3.425 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.341 3.354 make_images_data 4110 15.4 0.044 0.049 2.602 3.214 mp_sum_l 6514 12.8 2.539 3.180 2.539 3.180 wfi_extrapolate 11 7.9 0.001 0.001 3.113 3.114 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.273 3.000 dbcsr_complete_redistribute 325 12.2 0.412 0.843 2.135 2.994 cp_fm_cholesky_invert 11 10.9 2.962 2.969 2.962 2.969 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.815 2.818 fft_wrap_pw1pw2_140 451 13.1 0.212 0.216 2.774 2.786 cp_fm_redistribute_end 48 14.0 1.736 2.762 1.738 2.763 calculate_dm_sparse 110 9.5 0.001 0.003 2.702 2.730 cp_fm_diag_elpa_base 48 14.0 0.959 2.611 1.017 2.716 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.634 2.699 fft3d_ps 1111 14.6 1.063 1.075 2.654 2.667 calculate_rho_core 11 7.9 0.087 0.196 2.129 2.601 build_core_hamiltonian_matrix 11 6.9 0.032 0.123 1.067 2.589 potential_pw2rs 110 12.3 0.011 0.011 2.558 2.581 copy_fm_to_dbcsr 174 11.2 0.003 0.017 1.578 2.560 mp_alltoall_d11v 2046 13.8 2.060 2.486 2.060 2.486 grid_collocate_task_list 110 9.6 2.193 2.425 2.193 2.425 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.885 2.425 mp_irecv_dv 48980 15.7 0.816 2.299 0.816 2.299 mp_sum_d 3881 11.9 1.617 2.134 1.617 2.134 cp_fm_upper_to_full 70 13.6 1.395 1.849 1.395 1.849 cp_fm_cholesky_decompose 22 10.9 1.803 1.846 1.803 1.846 mp_allgather_i34 2055 14.4 0.718 1.840 0.718 1.840 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.677 1.698 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.638 1.652 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.377 1.499 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=74.357000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.909091, yerr=8.543520 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 740.114432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.136 0.259 67.526 67.531 qs_mol_dyn_low 1 2.0 0.003 0.005 66.640 66.649 qs_forces 11 3.9 0.005 0.023 66.568 66.570 qs_energies 11 4.9 0.017 0.048 63.166 63.172 scf_env_do_scf 11 5.9 0.001 0.001 54.042 54.045 scf_env_do_scf_inner_loop 99 6.5 0.005 0.014 41.940 41.942 velocity_verlet 10 3.0 0.008 0.011 37.438 37.440 dbcsr_multiply_generic 2055 12.4 0.118 0.121 30.438 30.629 qs_scf_new_mos 99 7.5 0.001 0.001 27.210 27.315 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.210 27.315 ot_scf_mini 99 9.5 0.003 0.003 25.573 25.681 multiply_cannon 2055 13.4 0.258 0.301 23.203 24.316 multiply_cannon_loop 2055 14.4 0.883 0.901 21.571 22.127 ot_mini 99 10.5 0.002 0.009 14.635 14.762 multiply_cannon_multrec 24660 15.4 4.220 6.996 13.091 14.396 rebuild_ks_matrix 110 8.3 0.000 0.000 12.209 12.323 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.208 12.322 init_scf_loop 11 6.9 0.001 0.002 12.060 12.063 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.813 10.914 qs_ot_get_derivative 99 11.5 0.004 0.023 10.413 10.527 prepare_preconditioner 11 7.9 0.000 0.000 10.239 10.258 make_preconditioner 11 8.9 0.000 0.001 10.239 10.258 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.430 9.906 dbcsr_mm_accdrv_process 52304 16.0 7.392 8.953 8.725 9.814 mp_waitall_1 126806 16.4 4.603 6.463 4.603 6.463 sum_up_and_integrate 110 10.3 0.037 0.039 6.446 6.460 integrate_v_rspace 110 11.3 0.003 0.003 6.409 6.423 init_scf_run 11 5.9 0.000 0.001 6.160 6.161 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.160 6.161 qs_ot_get_p 110 10.4 0.001 0.001 5.875 6.024 make_m2s 4110 13.4 0.059 0.061 5.463 5.959 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.809 5.820 calculate_rho_elec 110 8.6 0.077 0.081 5.808 5.819 make_images 4110 14.4 0.574 0.697 5.322 5.814 cp_fm_upper_to_full 70 13.8 3.343 4.764 3.343 4.764 apply_preconditioner_dbcsr 110 12.6 0.007 0.058 4.147 4.242 apply_single 110 13.6 0.000 0.000 4.140 4.242 ot_diis_step 99 11.5 0.012 0.015 4.175 4.177 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.026 4.041 dbcsr_complete_redistribute 325 12.2 0.430 0.513 2.767 3.918 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.581 3.644 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.539 3.539 grid_integrate_task_list 110 12.3 3.267 3.416 3.267 3.416 copy_fm_to_dbcsr 174 11.2 0.003 0.017 2.236 3.376 multiply_cannon_sync_h2d 24660 15.4 3.180 3.364 3.180 3.364 pw_transfer 1331 11.6 0.066 0.075 3.325 3.351 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.531 3.309 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.215 3.243 make_images_data 4110 15.4 0.046 0.049 2.745 3.235 hybrid_alltoall_any 4261 16.3 0.119 0.457 2.368 3.232 calculate_dm_sparse 110 9.5 0.001 0.002 3.113 3.131 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.063 3.124 density_rs2pw 110 9.6 0.004 0.004 2.929 3.120 wfi_extrapolate 11 7.9 0.001 0.001 3.023 3.023 calculate_first_density_matrix 1 7.0 0.000 0.002 3.015 3.018 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.004 3.005 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.973 2.975 rs_pw_transfer 902 11.9 0.010 0.011 2.726 2.954 cp_fm_redistribute_end 48 14.0 1.483 2.943 1.485 2.945 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.800 2.918 cp_fm_diag_elpa_base 48 14.0 1.375 2.807 1.456 2.917 cp_fm_cholesky_invert 11 10.9 2.886 2.895 2.886 2.895 mp_alltoall_i22 605 13.7 1.689 2.892 1.689 2.892 fft_wrap_pw1pw2_140 451 13.1 0.206 0.217 2.718 2.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.588 2.625 fft3d_ps 1111 14.6 1.063 1.094 2.513 2.529 grid_collocate_task_list 110 9.6 2.228 2.394 2.228 2.394 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 2.370 2.371 mp_alltoall_d11v 2046 13.8 1.927 2.366 1.927 2.366 jit_kernel_multiply 10 15.9 0.999 2.319 0.999 2.319 potential_pw2rs 110 12.3 0.013 0.013 2.225 2.236 cp_fm_cholesky_decompose 22 10.9 1.747 1.799 1.747 1.799 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.729 1.769 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.732 1.745 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.603 1.703 mp_allgather_i34 2055 14.4 0.714 1.676 0.714 1.676 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.641 1.652 acc_transpose_blocks 24660 15.4 0.103 0.106 1.577 1.601 mp_sum_l 6514 12.8 1.132 1.567 1.132 1.567 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.846 1.469 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.299 1.412 mp_irecv_dv 62702 16.1 0.749 1.392 0.749 1.392 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.531000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=697.181818, yerr=12.430302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 829.296640E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.057 0.093 59.395 59.395 qs_mol_dyn_low 1 2.0 0.003 0.004 58.771 58.780 qs_forces 11 3.9 0.002 0.002 58.691 58.693 qs_energies 11 4.9 0.011 0.037 55.037 55.042 scf_env_do_scf 11 5.9 0.000 0.001 45.059 45.060 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 36.681 36.682 velocity_verlet 10 3.0 0.008 0.019 31.830 31.833 dbcsr_multiply_generic 2055 12.4 0.128 0.164 24.606 24.714 qs_scf_new_mos 99 7.5 0.001 0.001 22.185 22.226 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.184 22.226 ot_scf_mini 99 9.5 0.002 0.002 20.899 20.920 multiply_cannon 2055 13.4 0.267 0.289 18.630 20.038 multiply_cannon_loop 2055 14.4 0.323 0.335 17.191 17.486 rebuild_ks_matrix 110 8.3 0.000 0.000 11.828 11.851 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.828 11.851 ot_mini 99 10.5 0.001 0.001 11.092 11.112 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.549 10.570 multiply_cannon_multrec 8220 15.4 3.304 4.653 8.453 9.370 init_scf_loop 11 6.9 0.000 0.000 8.325 8.326 mp_waitall_1 106626 16.5 6.609 8.199 6.609 8.199 qs_ot_get_derivative 99 11.5 0.001 0.001 7.220 7.243 init_scf_run 11 5.9 0.000 0.001 6.714 6.714 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.713 6.714 prepare_preconditioner 11 7.9 0.000 0.000 6.507 6.522 make_preconditioner 11 8.9 0.000 0.000 6.507 6.522 sum_up_and_integrate 110 10.3 0.048 0.049 6.401 6.416 integrate_v_rspace 110 11.3 0.003 0.003 6.353 6.369 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.070 6.150 dbcsr_mm_accdrv_process 17442 15.9 3.091 4.323 5.017 6.098 qs_ot_get_p 110 10.4 0.001 0.001 5.871 5.891 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.786 5.800 calculate_rho_elec 110 8.6 0.115 0.116 5.786 5.799 make_m2s 4110 13.4 0.039 0.040 4.470 4.740 make_images 4110 14.4 0.631 0.689 4.339 4.607 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 3.121 4.447 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.280 4.285 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.961 3.962 ot_diis_step 99 11.5 0.012 0.012 3.843 3.844 calculate_first_density_matrix 1 7.0 0.000 0.000 3.805 3.806 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.734 3.764 apply_single 110 13.6 0.000 0.000 3.734 3.764 grid_integrate_task_list 110 12.3 3.387 3.505 3.387 3.505 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.435 3.437 pw_transfer 1331 11.6 0.066 0.071 3.371 3.399 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.372 3.373 cp_fm_redistribute_end 48 14.0 0.853 3.329 0.859 3.331 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.259 3.291 cp_fm_diag_elpa_base 48 14.0 2.285 3.118 2.461 3.290 cp_fm_cholesky_invert 11 10.9 3.256 3.261 3.256 3.261 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.029 3.041 multiply_cannon_sync_h2d 8220 15.4 2.904 3.014 2.904 3.014 density_rs2pw 110 9.6 0.004 0.004 2.834 2.985 make_images_data 4110 15.4 0.038 0.043 2.476 2.860 wfi_extrapolate 11 7.9 0.001 0.001 2.815 2.815 hybrid_alltoall_any 4261 16.3 0.199 0.858 2.367 2.792 fft_wrap_pw1pw2_140 451 13.1 0.212 0.215 2.743 2.779 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.755 2.757 calculate_dm_sparse 110 9.5 0.001 0.001 2.677 2.726 rs_pw_transfer 902 11.9 0.010 0.010 2.555 2.714 fft3d_ps 1111 14.6 1.115 1.146 2.518 2.539 grid_collocate_task_list 110 9.6 2.327 2.523 2.327 2.523 potential_pw2rs 110 12.3 0.015 0.016 2.166 2.185 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.166 2.177 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.776 1.999 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.946 1.959 mp_alltoall_d11v 2046 13.8 1.742 1.906 1.742 1.906 jit_kernel_multiply 6 15.7 1.619 1.894 1.619 1.894 cp_fm_cholesky_decompose 22 10.9 1.774 1.795 1.774 1.795 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.593 1.714 mp_allgather_i34 2055 14.4 0.573 1.698 0.573 1.698 dbcsr_complete_redistribute 325 12.2 0.563 0.581 1.573 1.660 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.616 1.624 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.600 1.610 qs_create_task_list 11 7.9 0.007 0.019 1.252 1.348 generate_qs_task_list 11 8.9 0.379 0.447 1.245 1.330 mp_waitany 9240 13.8 1.122 1.300 1.122 1.300 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.226 1.246 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.893 1.201 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.395000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=776.636364, yerr=13.082068 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 1.196700E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.155 87.050 87.051 qs_mol_dyn_low 1 2.0 0.003 0.003 86.139 86.148 qs_forces 11 3.9 0.027 0.092 86.060 86.061 qs_energies 11 4.9 0.010 0.019 81.985 82.012 scf_env_do_scf 11 5.9 0.001 0.001 71.425 71.425 velocity_verlet 10 3.0 0.002 0.002 54.786 54.792 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.705 42.706 dbcsr_multiply_generic 2055 12.4 0.120 0.122 28.688 28.763 init_scf_loop 11 6.9 0.000 0.000 28.639 28.640 prepare_preconditioner 11 7.9 0.000 0.000 26.592 26.599 make_preconditioner 11 8.9 0.000 0.000 26.592 26.599 qs_scf_new_mos 99 7.5 0.001 0.001 26.323 26.370 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.323 26.370 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.816 26.066 ot_scf_mini 99 9.5 0.002 0.002 24.615 24.664 multiply_cannon 2055 13.4 0.364 0.395 21.797 22.602 multiply_cannon_loop 2055 14.4 0.340 0.343 19.988 20.416 cp_fm_upper_to_full 70 14.2 12.722 18.214 12.722 18.214 ot_mini 99 10.5 0.001 0.001 13.610 13.649 rebuild_ks_matrix 110 8.3 0.001 0.001 13.279 13.296 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.279 13.296 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.092 12.107 dbcsr_complete_redistribute 325 12.2 1.055 1.076 7.449 10.709 multiply_cannon_multrec 8220 15.4 4.422 4.565 9.682 9.791 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.361 9.622 qs_ot_get_derivative 99 11.5 0.001 0.001 9.119 9.171 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.761 9.001 mp_waitall_1 87304 16.6 7.845 8.742 7.845 8.742 mp_alltoall_i22 605 13.7 5.395 8.716 5.395 8.716 sum_up_and_integrate 110 10.3 0.091 0.092 6.658 6.702 integrate_v_rspace 110 11.3 0.003 0.003 6.567 6.611 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.409 6.443 calculate_rho_elec 110 8.6 0.226 0.227 6.408 6.442 init_scf_run 11 5.9 0.000 0.001 6.178 6.178 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.178 6.178 qs_ot_get_p 110 10.4 0.001 0.001 5.867 5.913 make_m2s 4110 13.4 0.043 0.043 5.210 5.681 make_images 4110 14.4 0.878 0.930 5.024 5.494 cp_fm_cholesky_invert 11 10.9 5.444 5.449 5.444 5.449 dbcsr_mm_accdrv_process 11614 15.7 3.287 3.681 5.117 5.415 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.818 5.352 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.576 5.054 apply_single 110 13.6 0.000 0.000 4.576 5.053 ot_diis_step 99 11.5 0.015 0.015 4.465 4.465 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.261 4.268 multiply_cannon_sync_h2d 8220 15.4 3.949 3.957 3.949 3.957 pw_transfer 1331 11.6 0.073 0.074 3.874 3.908 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.821 3.822 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.750 3.782 qs_energies_init_hamiltonians 11 5.9 0.007 0.013 3.740 3.766 grid_integrate_task_list 110 12.3 3.657 3.713 3.657 3.713 hybrid_alltoall_any 4261 16.3 0.255 0.551 2.776 3.512 make_images_data 4110 15.4 0.041 0.044 2.839 3.437 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.983 3.392 fft_wrap_pw1pw2_140 451 13.1 0.218 0.221 3.246 3.279 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.259 3.260 cp_fm_diag_elpa_base 48 14.0 2.727 2.920 3.256 3.257 calculate_dm_sparse 110 9.5 0.001 0.001 3.197 3.232 wfi_extrapolate 11 7.9 0.001 0.001 3.190 3.190 density_rs2pw 110 9.6 0.004 0.004 3.031 3.048 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.017 3.023 fft3d_ps 1111 14.6 1.265 1.277 2.957 2.990 calculate_first_density_matrix 1 7.0 0.000 0.000 2.882 2.883 grid_collocate_task_list 110 9.6 2.626 2.651 2.626 2.651 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.500 2.519 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.392 2.416 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.255 2.353 rs_pw_transfer 902 11.9 0.011 0.011 2.263 2.343 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.191 potential_pw2rs 110 12.3 0.021 0.022 2.155 2.188 cp_fm_cholesky_decompose 22 10.9 2.079 2.098 2.079 2.098 qs_create_task_list 11 7.9 0.001 0.001 1.911 1.959 generate_qs_task_list 11 8.9 0.740 0.795 1.910 1.958 mp_alltoall_d11v 2046 13.8 1.848 1.897 1.848 1.897 jit_kernel_multiply 10 15.3 1.628 1.887 1.628 1.887 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.762 1.825 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.786 1.802 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.793 1.794 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.051000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1115.727273, yerr=26.233535 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.169600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2177654. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.121 208.942 208.944 qs_mol_dyn_low 1 2.0 0.003 0.004 208.330 208.345 qs_forces 11 3.9 0.012 0.039 208.204 208.205 qs_energies 11 4.9 0.006 0.041 202.576 202.597 scf_env_do_scf 11 5.9 0.001 0.002 185.019 185.022 scf_env_do_scf_inner_loop 117 6.6 0.009 0.054 162.941 162.944 dbcsr_multiply_generic 2507 12.6 0.175 0.179 125.965 126.990 velocity_verlet 10 3.0 0.007 0.050 124.531 124.532 qs_scf_new_mos 117 7.6 0.001 0.001 122.838 123.040 qs_scf_loop_do_ot 117 8.6 0.001 0.002 122.837 123.040 ot_scf_mini 117 9.6 0.004 0.012 116.212 116.405 multiply_cannon 2507 13.6 0.239 0.250 102.080 104.266 multiply_cannon_loop 2507 14.6 2.120 2.196 99.754 101.690 ot_mini 117 10.6 0.001 0.002 67.036 67.208 multiply_cannon_multrec 60168 15.6 33.230 35.189 42.203 44.123 qs_ot_get_derivative 117 11.6 0.001 0.002 41.781 42.010 rebuild_ks_matrix 128 8.3 0.001 0.001 34.460 34.713 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.044 34.459 34.712 mp_waitall_1 291448 16.2 29.592 33.656 29.592 33.656 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.097 31.359 multiply_cannon_sync_h2d 60168 15.6 27.385 30.695 27.385 30.695 qs_ot_get_p 128 10.4 0.001 0.001 27.533 27.884 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.576 25.640 apply_single 128 13.6 0.001 0.001 24.576 25.640 ot_diis_step 117 11.6 0.008 0.010 24.859 24.860 init_scf_loop 11 6.9 0.017 0.133 21.990 21.992 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.322 20.447 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.424 19.582 multiply_cannon_metrocomm3 60168 15.6 0.110 0.115 15.851 17.964 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.587 17.589 prepare_preconditioner 11 7.9 0.000 0.000 16.722 16.765 make_preconditioner 11 8.9 0.000 0.002 16.722 16.765 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.913 16.114 sum_up_and_integrate 128 10.3 0.052 0.063 14.740 14.754 integrate_v_rspace 128 11.3 0.003 0.005 14.688 14.706 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.634 14.639 cp_fm_redistribute_end 83 14.4 11.557 14.558 11.569 14.561 make_m2s 5014 13.6 0.104 0.112 14.153 14.448 cp_fm_diag_elpa_base 83 14.4 2.947 14.234 2.978 14.352 make_images 5014 14.6 0.405 0.424 13.971 14.276 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.342 13.458 calculate_rho_elec 128 8.7 0.045 0.063 13.341 13.457 init_scf_run 11 5.9 0.000 0.001 13.041 13.042 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.041 13.042 mp_sum_l 7870 13.0 8.553 9.986 8.553 9.986 cp_fm_cholesky_invert 11 10.9 9.616 9.625 9.616 9.625 wfi_extrapolate 11 7.9 0.001 0.001 9.170 9.170 dbcsr_mm_accdrv_process 124484 16.2 3.315 3.459 8.531 9.085 calculate_dm_sparse 128 9.5 0.001 0.003 8.517 8.621 multiply_cannon_metrocomm1 60168 15.6 0.087 0.092 6.237 8.586 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.953 8.074 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.832 7.951 make_images_data 5014 15.6 0.069 0.075 7.010 7.892 density_rs2pw 128 9.7 0.006 0.007 7.054 7.517 grid_integrate_task_list 128 12.3 7.046 7.505 7.046 7.505 pw_transfer 1547 11.6 0.075 0.106 7.074 7.382 rs_pw_transfer 1046 11.9 0.017 0.020 6.889 7.362 hybrid_alltoall_any 5200 16.5 0.289 2.253 6.122 7.348 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.860 7.144 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 7.022 7.031 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.140 6.280 fft_wrap_pw1pw2_140 523 13.2 0.435 0.486 5.801 5.986 fft3d_ps 1291 14.7 2.094 2.595 5.664 5.897 mp_alltoall_d11v 2415 14.1 4.523 5.694 4.523 5.694 potential_pw2rs 128 12.3 0.009 0.010 5.196 5.228 grid_collocate_task_list 128 9.7 4.730 5.105 4.730 5.105 mp_sum_d 4459 12.1 4.252 4.948 4.252 4.948 cp_fm_cholesky_decompose 22 10.9 4.849 4.866 4.849 4.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=208.944000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.090909, yerr=8.073014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 826.572800E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 4718955. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.126 197.930 197.931 qs_mol_dyn_low 1 2.0 0.003 0.004 197.303 197.317 qs_forces 11 3.9 0.005 0.009 197.189 197.194 qs_energies 11 4.9 0.022 0.050 190.388 190.407 scf_env_do_scf 11 5.9 0.001 0.001 172.011 172.021 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 138.122 138.125 velocity_verlet 10 3.0 0.012 0.028 122.853 122.855 dbcsr_multiply_generic 2507 12.6 0.214 0.300 99.585 100.688 qs_scf_new_mos 117 7.6 0.001 0.001 98.205 98.572 qs_scf_loop_do_ot 117 8.6 0.001 0.001 98.204 98.571 ot_scf_mini 117 9.6 0.004 0.004 93.366 93.803 multiply_cannon 2507 13.6 0.476 0.521 78.849 82.788 multiply_cannon_loop 2507 14.6 1.261 1.303 75.428 77.607 ot_mini 117 10.6 0.001 0.001 50.733 51.183 mp_waitall_1 226760 16.4 25.439 38.488 25.439 38.488 multiply_cannon_multrec 30084 15.6 22.212 26.318 32.740 37.101 rebuild_ks_matrix 128 8.3 0.001 0.001 33.514 34.040 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.513 34.039 init_scf_loop 11 6.9 0.000 0.000 33.796 33.798 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.194 30.689 prepare_preconditioner 11 7.9 0.000 0.000 29.301 29.372 make_preconditioner 11 8.9 0.000 0.000 29.301 29.372 qs_ot_get_derivative 117 11.6 0.001 0.002 28.820 29.247 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.978 28.532 multiply_cannon_metrocomm3 30084 15.6 0.095 0.099 15.702 27.798 qs_ot_get_p 128 10.4 0.001 0.001 24.268 24.789 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.100 23.195 apply_single 128 13.6 0.001 0.001 22.100 23.195 ot_diis_step 117 11.6 0.014 0.015 21.739 21.741 multiply_cannon_sync_h2d 30084 15.6 19.481 21.698 19.481 21.698 qs_ot_p2m_diag 83 11.4 0.188 0.216 19.321 19.365 cp_dbcsr_syevd 83 12.4 0.005 0.005 18.125 18.126 cp_fm_cholesky_invert 11 10.9 17.136 17.150 17.136 17.150 make_m2s 5014 13.6 0.089 0.093 14.689 16.275 make_images 5014 14.6 1.146 1.335 14.480 16.066 sum_up_and_integrate 128 10.3 0.061 0.070 14.720 14.753 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.697 14.707 integrate_v_rspace 128 11.3 0.003 0.003 14.659 14.695 cp_fm_redistribute_end 83 14.4 8.573 14.609 8.586 14.613 cp_fm_diag_elpa_base 83 14.4 5.781 14.129 6.006 14.484 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.680 13.719 calculate_rho_elec 128 8.7 0.088 0.105 13.679 13.718 init_scf_run 11 5.9 0.000 0.001 13.377 13.378 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.376 13.378 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.310 11.637 multiply_cannon_metrocomm4 27577 15.6 0.096 0.112 3.670 10.813 dbcsr_mm_accdrv_process 62242 16.2 4.586 5.337 9.993 10.658 make_images_data 5014 15.6 0.066 0.073 8.745 10.599 mp_irecv_dv 69486 16.3 3.478 10.420 3.478 10.420 hybrid_alltoall_any 5200 16.5 0.342 1.515 7.568 10.135 wfi_extrapolate 11 7.9 0.001 0.001 8.602 8.602 pw_transfer 1547 11.6 0.084 0.104 8.045 8.119 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.816 7.878 density_rs2pw 128 9.7 0.006 0.006 7.252 7.649 grid_integrate_task_list 128 12.3 7.149 7.587 7.149 7.587 cp_fm_cholesky_decompose 22 10.9 7.272 7.346 7.272 7.346 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.458 7.308 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.260 7.268 rs_pw_transfer 1046 11.9 0.014 0.015 6.329 6.813 calculate_dm_sparse 128 9.5 0.001 0.001 6.641 6.812 fft_wrap_pw1pw2_140 523 13.2 0.458 0.498 6.748 6.805 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.251 6.431 fft3d_ps 1291 14.7 2.774 2.936 6.199 6.244 mp_sum_l 7870 13.0 4.244 6.098 4.244 6.098 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.369 5.429 grid_collocate_task_list 128 9.7 4.911 5.298 4.911 5.298 mp_alltoall_d11v 2415 14.1 4.874 5.296 4.874 5.296 potential_pw2rs 128 12.3 0.016 0.018 5.202 5.223 mp_allgather_i34 2507 14.6 1.980 4.971 1.980 4.971 calculate_first_density_matrix 1 7.0 0.000 0.000 4.595 4.601 dbcsr_complete_redistribute 395 12.7 0.792 0.905 3.358 4.213 mp_sum_d 4459 12.1 2.723 4.034 2.723 4.034 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=197.931000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=786.727273, yerr=2.894852 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 967.839744E+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 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_ISendRecv 15872 75008. MP_Wait 29756 MP_comm_split 83 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.101 0.282 178.660 178.712 qs_mol_dyn_low 1 2.0 0.003 0.005 177.579 177.592 qs_forces 11 3.9 0.004 0.015 177.460 177.466 qs_energies 11 4.9 0.034 0.087 170.911 170.920 scf_env_do_scf 11 5.9 0.001 0.001 154.324 154.325 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 118.687 118.688 velocity_verlet 10 3.0 0.019 0.026 113.251 113.253 dbcsr_multiply_generic 2507 12.6 0.194 0.224 81.941 83.051 qs_scf_new_mos 117 7.6 0.001 0.001 82.059 82.406 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.058 82.405 ot_scf_mini 117 9.6 0.004 0.005 77.883 78.270 multiply_cannon 2507 13.6 0.501 0.527 62.054 66.364 multiply_cannon_loop 2507 14.6 0.861 0.896 58.930 61.596 ot_mini 117 10.6 0.001 0.002 42.854 43.253 init_scf_loop 11 6.9 0.001 0.003 35.528 35.530 mp_waitall_1 178456 16.5 25.479 34.730 25.479 34.730 prepare_preconditioner 11 7.9 0.000 0.000 31.459 31.502 make_preconditioner 11 8.9 0.000 0.002 31.459 31.502 rebuild_ks_matrix 128 8.3 0.001 0.001 30.170 30.643 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 30.169 30.643 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.122 30.519 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.230 27.661 multiply_cannon_multrec 20056 15.6 13.548 16.748 22.395 25.536 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 15.317 24.780 qs_ot_get_derivative 117 11.6 0.001 0.002 23.149 23.539 qs_ot_get_p 128 10.4 0.001 0.001 20.414 20.884 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.847 20.843 apply_single 128 13.6 0.001 0.001 19.847 20.843 ot_diis_step 117 11.6 0.018 0.020 19.595 19.596 qs_ot_p2m_diag 83 11.4 0.266 0.273 16.108 16.125 multiply_cannon_sync_h2d 20056 15.6 14.215 15.816 14.215 15.816 make_m2s 5014 13.6 0.081 0.086 14.780 15.692 make_images 5014 14.6 1.164 1.271 14.547 15.458 cp_dbcsr_syevd 83 12.4 0.005 0.005 15.066 15.067 cp_fm_cholesky_invert 11 10.9 14.721 14.730 14.721 14.730 sum_up_and_integrate 128 10.3 0.071 0.076 14.156 14.180 integrate_v_rspace 128 11.3 0.003 0.003 14.085 14.111 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.112 13.145 calculate_rho_elec 128 8.7 0.132 0.146 13.111 13.144 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.885 11.887 cp_fm_redistribute_end 83 14.4 4.504 11.814 4.517 11.817 cp_fm_diag_elpa_base 83 14.4 6.867 11.199 7.275 11.709 init_scf_run 11 5.9 0.000 0.001 11.064 11.064 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.064 11.064 make_images_data 5014 15.6 0.060 0.067 9.067 10.310 hybrid_alltoall_any 5200 16.5 0.431 1.966 8.008 9.427 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.025 9.314 multiply_cannon_metrocomm4 17549 15.6 0.063 0.072 3.444 9.227 mp_irecv_dv 50230 16.2 3.321 8.984 3.321 8.984 dbcsr_mm_accdrv_process 41502 16.2 4.544 5.469 8.299 8.547 pw_transfer 1547 11.6 0.083 0.102 7.676 7.782 grid_integrate_task_list 128 12.3 7.233 7.648 7.233 7.648 cp_fm_cholesky_decompose 22 10.9 7.595 7.641 7.595 7.641 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.450 7.562 cp_fm_upper_to_full 105 14.5 5.767 7.540 5.767 7.540 wfi_extrapolate 11 7.9 0.001 0.001 7.359 7.359 density_rs2pw 128 9.7 0.005 0.006 6.622 7.101 fft_wrap_pw1pw2_140 523 13.2 0.470 0.510 6.503 6.620 dbcsr_complete_redistribute 395 12.7 1.186 1.293 4.670 6.468 rs_pw_transfer 1046 11.9 0.013 0.014 5.549 6.027 calculate_dm_sparse 128 9.5 0.001 0.002 5.825 5.943 fft3d_ps 1291 14.7 2.697 2.914 5.728 5.805 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.658 5.664 grid_collocate_task_list 128 9.7 5.066 5.415 5.066 5.415 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.694 5.395 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.474 5.269 mp_alltoall_d11v 2415 14.1 4.247 5.001 4.247 5.001 mp_sum_l 7870 13.0 3.377 4.810 3.377 4.810 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.672 4.799 potential_pw2rs 128 12.3 0.021 0.023 4.644 4.662 mp_allgather_i34 2507 14.6 1.649 4.429 1.649 4.429 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.059 4.095 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.318 4.095 qs_energies_init_hamiltonians 11 5.9 0.008 0.029 3.937 3.941 mp_alltoall_i22 716 14.1 1.924 3.907 1.924 3.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.712000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=906.818182, yerr=22.114652 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.134481E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922351760 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4125 56653. MP_Allreduce 11351 1064. MP_Sync 172 MP_Alltoall 1724 12509574. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.256 0.731 193.515 193.516 qs_mol_dyn_low 1 2.0 0.003 0.012 192.228 192.241 qs_forces 11 3.9 0.007 0.030 192.098 192.104 qs_energies 11 4.9 0.008 0.054 184.883 184.901 scf_env_do_scf 11 5.9 0.001 0.001 166.097 166.107 velocity_verlet 10 3.0 0.011 0.021 126.110 126.113 scf_env_do_scf_inner_loop 118 6.6 0.015 0.104 119.290 119.295 qs_scf_new_mos 118 7.6 0.001 0.001 82.866 83.112 qs_scf_loop_do_ot 118 8.6 0.001 0.002 82.866 83.111 dbcsr_multiply_generic 2534 12.6 0.203 0.220 81.379 82.355 ot_scf_mini 118 9.6 0.003 0.004 78.348 78.653 multiply_cannon 2534 13.6 0.565 0.598 55.711 58.638 multiply_cannon_loop 2534 14.6 1.189 1.223 51.813 54.052 init_scf_loop 11 6.9 0.002 0.013 46.680 46.703 ot_mini 118 10.6 0.001 0.001 43.209 43.491 prepare_preconditioner 11 7.9 0.000 0.000 42.510 42.532 make_preconditioner 11 8.9 0.000 0.002 42.510 42.532 make_full_inverse_cholesky 11 9.9 0.000 0.000 36.196 41.194 multiply_cannon_multrec 30408 15.6 14.165 19.710 26.588 31.598 rebuild_ks_matrix 129 8.3 0.001 0.001 29.575 29.881 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.042 29.574 29.881 mp_waitall_1 155386 16.5 18.141 28.721 18.141 28.721 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.780 27.057 qs_ot_get_derivative 118 11.6 0.001 0.002 23.418 23.727 make_m2s 5068 13.6 0.098 0.116 20.744 21.850 qs_ot_get_p 129 10.4 0.001 0.001 21.289 21.595 make_images 5068 14.6 1.953 2.269 20.435 21.548 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.169 19.848 apply_single 129 13.6 0.001 0.001 19.169 19.847 ot_diis_step 118 11.6 0.018 0.019 19.662 19.663 qs_ot_p2m_diag 84 11.4 0.348 0.395 16.889 16.941 cp_fm_upper_to_full 106 14.7 11.129 16.478 11.129 16.478 cp_fm_cholesky_invert 11 10.9 16.386 16.395 16.386 16.395 multiply_cannon_metrocomm3 30408 15.6 0.047 0.049 6.499 15.972 cp_dbcsr_syevd 84 12.4 0.005 0.005 15.589 15.590 sum_up_and_integrate 129 10.3 0.080 0.087 14.251 14.276 integrate_v_rspace 129 11.3 0.003 0.004 14.170 14.198 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.530 13.586 calculate_rho_elec 129 8.7 0.176 0.192 13.529 13.585 make_images_data 5068 15.6 0.064 0.069 11.220 13.095 dbcsr_complete_redistribute 397 12.7 1.567 1.665 9.307 13.050 multiply_cannon_sync_h2d 30408 15.6 11.782 12.876 11.782 12.876 init_scf_run 11 5.9 0.000 0.001 12.466 12.467 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.466 12.467 dbcsr_mm_accdrv_process 62840 16.2 7.485 8.415 11.996 12.408 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.316 12.318 cp_fm_redistribute_end 84 14.4 2.123 12.223 2.138 12.229 cp_fm_diag_elpa_base 84 14.4 9.463 11.580 10.055 12.108 hybrid_alltoall_any 5255 16.5 0.530 2.236 9.848 12.010 copy_fm_to_dbcsr 210 11.7 0.002 0.002 7.907 11.597 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.292 9.926 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.680 9.906 mp_alltoall_i22 720 14.1 5.537 9.313 5.537 9.313 pw_transfer 1559 11.6 0.084 0.098 7.983 8.044 grid_integrate_task_list 129 12.3 7.547 7.945 7.547 7.945 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.753 7.819 cp_fm_cholesky_decompose 22 10.9 7.617 7.700 7.617 7.700 multiply_cannon_metrocomm4 25340 15.6 0.075 0.087 2.817 7.654 wfi_extrapolate 11 7.9 0.001 0.001 7.606 7.606 mp_irecv_dv 76866 16.2 2.672 7.374 2.672 7.374 density_rs2pw 129 9.7 0.005 0.006 6.516 6.970 fft_wrap_pw1pw2_140 527 13.2 0.476 0.484 6.821 6.911 calculate_dm_sparse 129 9.5 0.001 0.002 6.341 6.440 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.366 6.420 fft3d_ps 1301 14.7 2.819 2.870 5.981 6.045 grid_collocate_task_list 129 9.7 5.250 5.668 5.250 5.668 rs_pw_transfer 1054 12.0 0.013 0.014 5.093 5.536 mp_alltoall_d11v 2429 14.1 5.110 5.528 5.110 5.528 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.054 5.196 calculate_first_density_matrix 1 7.0 0.000 0.001 4.664 4.669 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.536 4.626 qs_energies_init_hamiltonians 11 5.9 0.005 0.023 4.600 4.608 potential_pw2rs 129 12.3 0.024 0.024 4.463 4.481 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.224 4.291 mp_sum_l 7951 13.1 3.108 4.108 3.108 4.108 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=193.516000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1070.454545, yerr=18.879347 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.510277E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58222. MP_Allreduce 11000 1175. MP_Sync 86 MP_Alltoall 1700 18828153. MP_ISendRecv 7620 122880. MP_Wait 19810 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.109 0.157 173.333 173.334 qs_mol_dyn_low 1 2.0 0.003 0.004 172.165 172.178 qs_forces 11 3.9 0.003 0.003 172.063 172.067 qs_energies 11 4.9 0.042 0.224 164.679 164.683 scf_env_do_scf 11 5.9 0.001 0.001 145.220 145.233 velocity_verlet 10 3.0 0.014 0.017 110.973 110.977 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 109.236 109.236 dbcsr_multiply_generic 2485 12.5 0.209 0.223 74.437 74.899 qs_scf_new_mos 116 7.6 0.001 0.001 73.734 73.834 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.733 73.833 ot_scf_mini 116 9.6 0.003 0.004 69.296 69.353 multiply_cannon 2485 13.5 0.591 0.627 54.821 59.133 multiply_cannon_loop 2485 14.5 0.446 0.454 50.168 50.994 ot_mini 116 10.6 0.001 0.001 39.241 39.301 init_scf_loop 11 6.9 0.000 0.000 35.832 35.833 mp_waitall_1 128490 16.6 25.938 32.765 25.938 32.765 prepare_preconditioner 11 7.9 0.000 0.000 31.853 31.879 make_preconditioner 11 8.9 0.000 0.000 31.853 31.879 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.738 29.998 rebuild_ks_matrix 127 8.3 0.001 0.001 28.628 28.734 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 28.627 28.733 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.097 26.195 multiply_cannon_multrec 9940 15.5 10.300 14.309 18.112 20.962 ot_diis_step 116 11.6 0.020 0.022 19.664 19.664 qs_ot_get_derivative 116 11.6 0.002 0.002 19.509 19.570 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.146 19.511 apply_single 127 13.6 0.001 0.001 19.145 19.511 multiply_cannon_metrocomm3 9940 15.5 0.022 0.023 12.198 19.247 make_m2s 4970 13.5 0.066 0.072 16.061 18.361 cp_fm_cholesky_invert 11 10.9 18.155 18.161 18.155 18.161 make_images 4970 14.5 2.280 2.854 15.757 18.050 qs_ot_get_p 127 10.4 0.001 0.001 17.330 17.417 sum_up_and_integrate 127 10.3 0.102 0.108 13.929 13.978 integrate_v_rspace 127 11.3 0.003 0.004 13.827 13.882 qs_ot_p2m_diag 82 11.4 0.490 0.496 13.587 13.603 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.097 13.136 calculate_rho_elec 127 8.7 0.255 0.266 13.097 13.135 make_images_data 4970 15.5 0.051 0.059 9.779 12.475 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.457 12.457 init_scf_run 11 5.9 0.000 0.001 12.310 12.310 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.310 12.310 multiply_cannon_sync_h2d 9940 15.5 11.563 12.131 11.563 12.131 hybrid_alltoall_any 5155 16.4 0.770 3.509 9.580 11.932 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.482 9.485 cp_fm_diag_elpa_base 82 14.4 9.241 9.318 9.478 9.482 cp_fm_cholesky_decompose 22 10.9 8.038 8.150 8.038 8.150 grid_integrate_task_list 127 12.3 7.676 8.120 7.676 8.120 dbcsr_mm_accdrv_process 20590 16.1 2.688 3.665 7.443 8.111 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.748 7.790 pw_transfer 1535 11.6 0.082 0.088 7.713 7.726 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.488 7.508 multiply_cannon_metrocomm1 9940 15.5 0.029 0.029 4.435 7.453 wfi_extrapolate 11 7.9 0.001 0.001 7.241 7.241 mp_allgather_i34 2485 14.5 2.819 7.115 2.819 7.115 calculate_dm_sparse 127 9.5 0.001 0.001 6.535 6.604 fft_wrap_pw1pw2_140 519 13.2 0.490 0.506 6.584 6.604 density_rs2pw 127 9.7 0.005 0.005 5.979 6.292 grid_collocate_task_list 127 9.7 5.439 5.715 5.439 5.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 5.687 5.694 fft3d_ps 1281 14.7 2.705 2.771 5.668 5.693 dbcsr_complete_redistribute 393 12.7 2.165 2.216 5.272 5.639 mp_alltoall_d11v 2401 14.1 4.718 5.410 4.718 5.410 qs_energies_init_hamiltonians 11 5.9 0.006 0.010 5.371 5.373 calculate_first_density_matrix 1 7.0 0.000 0.000 4.870 4.873 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.564 4.576 rs_pw_transfer 1038 11.9 0.012 0.013 4.240 4.568 potential_pw2rs 127 12.3 0.026 0.027 4.119 4.132 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.906 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.838 3.880 mp_irecv_dv 28618 15.9 1.801 3.813 1.801 3.813 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.442 3.761 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.655 3.675 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.563 3.630 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.550 3.582 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.334000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1404.818182, yerr=52.842507 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.289447E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342272E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59820. MP_Allreduce 10885 1518. MP_Sync 86 MP_Alltoall 1700 36954383. MP_ISendRecv 3556 218624. MP_Wait 11506 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.058 289.537 289.538 qs_mol_dyn_low 1 2.0 0.003 0.005 288.622 288.633 qs_forces 11 3.9 0.005 0.007 288.490 288.494 qs_energies 11 4.9 0.007 0.007 279.900 279.915 scf_env_do_scf 11 5.9 0.001 0.001 257.801 257.814 velocity_verlet 10 3.0 0.041 0.045 208.117 208.125 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 132.632 132.634 init_scf_loop 11 6.9 0.000 0.000 124.913 124.915 prepare_preconditioner 11 7.9 0.000 0.000 120.161 120.184 make_preconditioner 11 8.9 0.000 0.000 120.161 120.184 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.521 117.359 qs_scf_new_mos 116 7.6 0.001 0.001 91.051 91.119 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.050 91.118 ot_scf_mini 116 9.6 0.004 0.004 86.326 86.423 dbcsr_multiply_generic 2485 12.5 0.212 0.218 82.765 82.898 cp_fm_upper_to_full 104 14.8 53.306 75.964 53.306 75.964 multiply_cannon 2485 13.5 0.684 0.731 58.574 59.968 multiply_cannon_loop 2485 14.5 0.468 0.474 54.933 55.321 ot_mini 116 10.6 0.001 0.001 44.553 44.667 dbcsr_complete_redistribute 393 12.7 3.995 4.044 29.345 42.128 copy_fm_to_dbcsr 208 11.6 0.001 0.001 25.932 38.731 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.594 36.256 mp_alltoall_i22 712 14.1 21.463 34.505 21.463 34.505 cp_fm_cholesky_invert 11 10.9 33.954 33.960 33.954 33.960 rebuild_ks_matrix 127 8.3 0.001 0.001 33.346 33.493 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 33.345 33.493 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.076 31.213 mp_waitall_1 104546 16.7 27.767 31.081 27.767 31.081 qs_ot_get_p 127 10.4 0.001 0.001 26.563 26.725 qs_ot_get_derivative 116 11.6 0.002 0.002 24.710 24.796 qs_ot_p2m_diag 82 11.4 0.867 0.873 22.449 22.479 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.716 20.718 make_m2s 4970 13.5 0.076 0.079 19.269 20.189 ot_diis_step 116 11.6 0.022 0.022 19.791 19.791 make_images 4970 14.5 3.723 3.856 18.799 19.720 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.126 19.271 apply_single 127 13.6 0.001 0.001 19.126 19.271 multiply_cannon_metrocomm3 9940 15.5 0.023 0.023 18.670 18.982 multiply_cannon_multrec 9940 15.5 10.533 12.429 18.186 18.274 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.500 17.504 cp_fm_diag_elpa_base 82 14.4 13.119 14.749 17.496 17.500 sum_up_and_integrate 127 10.3 0.193 0.194 15.459 15.552 multiply_cannon_sync_h2d 9940 15.5 15.535 15.549 15.535 15.549 integrate_v_rspace 127 11.3 0.004 0.004 15.266 15.359 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.761 14.799 calculate_rho_elec 127 8.7 0.478 0.479 14.760 14.799 init_scf_run 11 5.9 0.000 0.001 12.149 12.150 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.149 12.150 make_images_data 4970 15.5 0.059 0.064 10.150 11.933 hybrid_alltoall_any 5155 16.4 1.294 3.009 10.215 11.717 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.369 9.436 dbcsr_mm_accdrv_process 20590 16.0 3.976 6.062 7.417 9.363 cp_fm_cholesky_decompose 22 10.9 9.057 9.109 9.057 9.109 wfi_extrapolate 11 7.9 0.001 0.001 8.698 8.698 grid_integrate_task_list 127 12.3 8.479 8.640 8.479 8.640 pw_transfer 1535 11.6 0.093 0.094 8.176 8.186 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 7.938 7.947 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.904 7.907 fft_wrap_pw1pw2_140 519 13.2 0.520 0.523 7.022 7.034 mp_alltoall_d11v 2401 14.1 6.817 6.958 6.817 6.958 calculate_dm_sparse 127 9.5 0.001 0.001 6.561 6.674 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.415 6.489 grid_collocate_task_list 127 9.7 6.287 6.319 6.287 6.319 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.105 6.200 fft3d_ps 1281 14.7 2.742 2.748 6.010 6.017 density_rs2pw 127 9.7 0.005 0.005 5.964 5.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.538000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2145.090909, yerr=49.371421 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 1.260167E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.146 0.564 89.031 89.033 qs_energies 1 2.0 0.000 0.000 87.051 87.073 ls_scf 1 3.0 0.008 0.057 85.705 85.719 dbcsr_multiply_generic 111 6.7 0.016 0.031 73.429 73.635 multiply_cannon 111 7.7 0.022 0.057 56.263 57.128 multiply_cannon_loop 111 8.7 0.211 0.227 52.811 53.976 ls_scf_main 1 4.0 0.004 0.032 52.643 52.657 density_matrix_trs4 2 5.0 0.002 0.003 47.092 47.189 ls_scf_init_scf 1 4.0 0.006 0.027 29.992 29.997 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.265 28.358 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.040 26.079 mp_waitall_1 11316 10.9 22.718 25.187 22.718 25.187 multiply_cannon_multrec 2664 9.7 8.180 9.027 15.507 17.393 multiply_cannon_sync_h2d 2664 9.7 13.886 15.348 13.886 15.348 make_m2s 222 7.7 0.008 0.011 13.383 13.837 make_images 222 8.7 0.099 0.109 13.360 13.816 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.705 12.424 make_images_data 222 9.7 0.004 0.007 7.854 8.422 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.477 8.227 dbcsr_mm_accdrv_process 4760 10.4 0.514 0.641 6.945 8.027 hybrid_alltoall_any 227 10.6 0.215 1.831 6.696 7.779 dbcsr_mm_accdrv_process_sort 4760 11.4 6.210 7.070 6.210 7.070 calculate_norms 4752 9.8 5.560 6.229 5.560 6.229 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.019 5.239 mp_sum_l 807 5.4 3.462 4.939 3.462 4.939 arnoldi_extremal 4 6.8 0.013 0.058 3.708 3.754 arnoldi_normal_ev 4 7.8 0.016 0.062 3.694 3.753 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.066 3.586 mp_irecv_dv 6231 10.9 2.049 3.561 2.049 3.561 make_images_sizes 222 9.7 0.000 0.000 0.829 3.531 mp_alltoall_i44 222 10.7 0.829 3.530 0.829 3.530 build_subspace 16 8.4 0.017 0.047 3.349 3.355 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.391 3.270 ls_scf_post 1 4.0 0.005 0.029 3.061 3.078 ls_scf_store_result 1 5.0 0.000 0.000 2.870 2.913 dbcsr_special_finalize 555 9.7 0.005 0.006 2.355 2.692 dbcsr_merge_single_wm 555 10.7 0.461 0.587 2.347 2.684 make_images_pack 222 9.7 2.199 2.635 2.201 2.636 dbcsr_matrix_vector_mult 304 9.0 0.003 0.009 2.341 2.596 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.472 2.069 2.474 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.264 2.418 dbcsr_sort_data 658 11.4 2.144 2.404 2.144 2.404 buffer_matrices_ensure_size 222 8.7 1.776 2.081 1.776 2.081 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.000 2.036 rebuild_ks_matrix 3 7.3 0.000 0.000 1.990 2.027 qs_ks_build_kohn_sham_matrix 3 8.3 0.021 0.123 1.990 2.026 ls_scf_initial_guess 1 5.0 0.000 0.000 1.720 1.826 ls_scf_qs_atomic_guess 1 6.0 0.004 0.016 1.720 1.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=89.033000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 2.109301E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.165 0.409 92.721 92.722 qs_energies 1 2.0 0.000 0.000 91.625 91.693 ls_scf 1 3.0 0.006 0.045 90.253 90.327 dbcsr_multiply_generic 111 6.7 0.021 0.064 75.763 76.127 multiply_cannon 111 7.7 0.037 0.081 53.422 57.173 ls_scf_main 1 4.0 0.001 0.010 55.225 55.239 multiply_cannon_loop 111 8.7 0.116 0.122 50.155 53.266 density_matrix_trs4 2 5.0 0.002 0.003 49.447 49.625 ls_scf_init_scf 1 4.0 0.000 0.001 31.496 31.498 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.272 30.390 mp_waitall_1 9246 10.9 21.500 29.476 21.500 29.476 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.645 27.660 multiply_cannon_multrec 1332 9.7 13.077 16.914 22.269 27.226 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.988 20.890 make_m2s 222 7.7 0.011 0.042 15.565 16.166 make_images 222 8.7 1.578 1.980 15.530 16.138 dbcsr_mm_accdrv_process 4041 10.4 0.266 0.453 8.801 10.408 dbcsr_mm_accdrv_process_sort 4041 11.4 8.386 9.956 8.386 9.956 make_images_data 222 9.7 0.004 0.005 8.852 9.709 hybrid_alltoall_any 227 10.6 0.519 2.458 8.258 9.154 mp_sum_l 807 5.4 5.582 8.460 5.582 8.460 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.210 7.563 mp_irecv_dv 3311 11.0 3.190 7.514 3.190 7.514 calculate_norms 2376 9.8 6.010 6.689 6.010 6.689 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.201 6.678 multiply_cannon_sync_h2d 1332 9.7 4.772 6.047 4.772 6.047 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.181 5.409 arnoldi_extremal 4 6.8 0.000 0.002 4.979 5.000 arnoldi_normal_ev 4 7.8 0.002 0.007 4.979 5.000 build_subspace 16 8.4 0.016 0.037 4.661 4.687 ls_scf_post 1 4.0 0.014 0.083 3.525 3.605 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.184 3.383 ls_scf_store_result 1 5.0 0.000 0.000 3.262 3.335 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.216 2.741 3.218 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.282 2.968 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.594 2.711 make_images_pack 222 9.7 2.039 2.408 2.041 2.410 mp_allgather_i34 111 8.7 0.959 2.360 0.959 2.360 dbcsr_sort_data 436 11.2 1.834 2.096 1.834 2.096 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.860 1.864 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.722000, 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/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 2.712166E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.117 0.161 95.606 95.607 qs_energies 1 2.0 0.000 0.000 94.557 94.560 ls_scf 1 3.0 0.000 0.001 93.029 93.032 dbcsr_multiply_generic 111 6.7 0.019 0.050 77.559 77.943 ls_scf_main 1 4.0 0.000 0.000 58.178 58.183 multiply_cannon 111 7.7 0.044 0.105 53.174 57.002 multiply_cannon_loop 111 8.7 0.100 0.107 49.637 53.800 density_matrix_trs4 2 5.0 0.002 0.003 52.265 52.443 mp_waitall_1 7374 11.0 24.520 34.554 24.520 34.554 ls_scf_init_scf 1 4.0 0.000 0.001 31.227 31.230 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.078 30.167 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.533 27.555 multiply_cannon_multrec 888 9.7 12.517 15.118 21.008 24.255 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.281 23.731 make_m2s 222 7.7 0.006 0.008 17.583 18.741 make_images 222 8.7 1.977 2.313 17.545 18.703 hybrid_alltoall_any 227 10.6 0.620 2.871 9.628 11.283 make_images_data 222 9.7 0.005 0.018 9.976 11.043 dbcsr_mm_accdrv_process 3754 10.4 0.232 0.419 8.032 9.247 dbcsr_mm_accdrv_process_sort 3754 11.4 7.657 8.828 7.657 8.828 mp_sum_l 807 5.4 5.245 8.641 5.245 8.641 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.880 7.587 multiply_cannon_sync_h2d 888 9.7 6.012 7.485 6.012 7.485 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.521 7.086 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.135 7.076 mp_irecv_dv 2335 11.1 2.505 7.048 2.505 7.048 arnoldi_extremal 4 6.8 0.000 0.003 5.163 5.189 arnoldi_normal_ev 4 7.8 0.010 0.023 5.162 5.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.888 5.080 build_subspace 16 8.4 0.014 0.022 4.771 4.785 calculate_norms 1584 9.8 4.263 4.696 4.263 4.696 mp_allgather_i34 111 8.7 1.372 3.874 1.372 3.874 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.419 3.748 ls_scf_post 1 4.0 0.000 0.000 3.624 3.628 dbcsr_matrix_vector_mult_local 304 10.0 3.007 3.583 3.009 3.585 ls_scf_store_result 1 5.0 0.000 0.000 3.358 3.446 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.803 2.922 make_images_sizes 222 9.7 0.000 0.000 1.301 2.463 mp_alltoall_i44 222 10.7 1.300 2.463 1.300 2.463 dbcsr_sort_data 325 11.1 1.863 2.160 1.863 2.160 make_images_pack 222 9.7 1.817 2.136 1.820 2.138 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.607000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2166.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 3.330855E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.215 0.282 97.837 97.837 qs_energies 1 2.0 0.000 0.000 96.066 96.072 ls_scf 1 3.0 0.000 0.000 94.335 94.340 dbcsr_multiply_generic 111 6.7 0.017 0.017 77.995 78.309 ls_scf_main 1 4.0 0.000 0.000 58.367 58.368 multiply_cannon 111 7.7 0.048 0.097 51.630 56.300 density_matrix_trs4 2 5.0 0.002 0.003 52.330 52.455 multiply_cannon_loop 111 8.7 0.114 0.124 46.540 49.450 ls_scf_init_scf 1 4.0 0.000 0.000 32.755 32.757 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.589 31.648 mp_waitall_1 6438 11.0 22.502 29.386 22.502 29.386 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.007 29.020 multiply_cannon_multrec 1332 9.7 14.142 17.269 22.038 25.210 make_m2s 222 7.7 0.007 0.008 20.816 22.280 make_images 222 8.7 3.134 3.601 20.765 22.232 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.266 16.913 make_images_data 222 9.7 0.004 0.004 11.566 13.215 hybrid_alltoall_any 227 10.6 0.796 3.818 10.939 12.751 dbcsr_mm_accdrv_process 3641 10.4 0.209 0.568 7.531 9.002 dbcsr_mm_accdrv_process_sort 3641 11.4 7.089 8.586 7.089 8.586 mp_sum_l 807 5.4 4.011 7.662 4.011 7.662 multiply_cannon_sync_h2d 1332 9.7 5.524 6.113 5.524 6.113 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.106 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.052 6.080 mp_irecv_dv 3229 10.9 2.058 6.027 2.058 6.027 arnoldi_extremal 4 6.8 0.000 0.000 5.366 5.380 arnoldi_normal_ev 4 7.8 0.002 0.005 5.366 5.380 build_subspace 16 8.4 0.014 0.021 5.003 5.012 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.492 4.792 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.584 4.750 mp_allgather_i34 111 8.7 2.212 4.699 2.212 4.699 calculate_norms 2376 9.8 4.196 4.576 4.196 4.576 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.597 3.933 dbcsr_matrix_vector_mult_local 304 10.0 3.171 3.668 3.173 3.670 dbcsr_sort_data 658 11.4 3.074 3.415 3.074 3.415 ls_scf_post 1 4.0 0.000 0.000 3.213 3.219 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.207 dbcsr_merge_single_wm 555 10.7 0.538 0.657 2.821 3.198 ls_scf_store_result 1 5.0 0.000 0.000 2.967 3.012 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.909 2.958 dbcsr_data_release 10477 10.7 1.581 2.433 1.581 2.433 dbcsr_finalize 304 7.8 0.049 0.061 1.794 1.970 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.837000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 4.621902E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.097 0.108 100.739 100.739 qs_energies 1 2.0 0.000 0.000 99.100 99.106 ls_scf 1 3.0 0.000 0.000 97.117 97.122 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.555 78.750 ls_scf_main 1 4.0 0.000 0.000 61.971 61.971 multiply_cannon 111 7.7 0.087 0.175 55.522 60.410 density_matrix_trs4 2 5.0 0.002 0.003 54.945 55.025 multiply_cannon_loop 111 8.7 0.069 0.075 51.125 52.948 mp_waitall_1 5481 11.0 27.033 32.050 27.033 32.050 ls_scf_init_scf 1 4.0 0.000 0.000 31.444 31.449 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.211 30.271 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.628 27.639 multiply_cannon_multrec 444 9.7 13.884 16.484 21.026 23.223 make_m2s 222 7.7 0.004 0.005 18.107 20.481 make_images 222 8.7 3.713 4.408 18.045 20.420 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.064 16.631 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.497 14.278 hybrid_alltoall_any 227 10.6 0.790 3.772 10.079 12.687 make_images_data 222 9.7 0.003 0.004 10.269 12.656 multiply_cannon_sync_h2d 444 9.7 6.593 8.087 6.593 8.087 dbcsr_mm_accdrv_process 3003 10.4 0.174 0.513 6.844 7.988 dbcsr_mm_accdrv_process_sort 3003 11.4 6.392 7.490 6.392 7.490 mp_allgather_i34 111 8.7 2.634 7.036 2.634 7.036 arnoldi_extremal 4 6.8 0.000 0.000 5.841 5.848 arnoldi_normal_ev 4 7.8 0.027 0.040 5.841 5.848 build_subspace 16 8.4 0.015 0.020 5.393 5.404 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.885 5.029 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.139 4.334 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.696 4.239 mp_irecv_dv 1241 11.2 1.676 4.216 1.676 4.216 mp_sum_l 807 5.4 2.954 4.170 2.954 4.170 dbcsr_matrix_vector_mult_local 304 10.0 3.656 4.127 3.658 4.129 ls_scf_post 1 4.0 0.000 0.000 3.702 3.708 make_images_sizes 222 9.7 0.000 0.000 1.196 3.675 mp_alltoall_i44 222 10.7 1.195 3.675 1.195 3.675 calculate_norms 792 9.8 3.539 3.652 3.539 3.652 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.376 3.492 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.469 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.994 3.141 dbcsr_finalize 304 7.8 0.062 0.078 2.208 2.326 dbcsr_merge_all 275 8.9 0.475 0.544 2.058 2.161 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.739000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3643.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/78096bc50ec3c41c1825630b7cc5ec9938bfc15f_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 8.719766E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.095 0.133 109.558 109.559 qs_energies 1 2.0 0.000 0.000 107.909 107.923 ls_scf 1 3.0 0.000 0.000 104.964 104.978 dbcsr_multiply_generic 111 6.7 0.023 0.025 78.974 79.092 ls_scf_main 1 4.0 0.000 0.000 66.668 66.669 density_matrix_trs4 2 5.0 0.002 0.003 57.747 57.803 multiply_cannon 111 7.7 0.221 0.353 51.189 53.271 multiply_cannon_loop 111 8.7 0.067 0.069 47.614 48.702 ls_scf_init_scf 1 4.0 0.000 0.000 34.586 34.587 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.199 33.210 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.091 30.100 mp_waitall_1 4569 11.1 23.455 27.233 23.455 27.233 make_m2s 222 7.7 0.005 0.005 24.059 25.154 make_images 222 8.7 4.576 4.963 23.954 25.046 multiply_cannon_multrec 444 9.7 17.891 18.814 22.556 23.420 make_images_data 222 9.7 0.003 0.003 13.295 15.839 hybrid_alltoall_any 227 10.6 1.658 3.618 13.092 15.800 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.522 12.160 multiply_cannon_sync_h2d 444 9.7 8.830 8.871 8.830 8.871 arnoldi_extremal 4 6.8 0.000 0.000 7.351 7.387 arnoldi_normal_ev 4 7.8 0.028 0.036 7.351 7.386 build_subspace 16 8.4 0.026 0.037 6.759 6.766 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.393 5.543 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.309 5.541 dbcsr_matrix_vector_mult_local 304 10.0 4.941 5.233 4.944 5.236 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.832 4.923 dbcsr_mm_accdrv_process 1814 10.4 0.263 0.434 4.497 4.682 dbcsr_mm_accdrv_process_sort 1814 11.4 4.111 4.241 4.111 4.241 mp_allgather_i34 111 8.7 1.207 3.791 1.207 3.791 ls_scf_post 1 4.0 0.000 0.000 3.710 3.723 make_images_sizes 222 9.7 0.000 0.000 1.454 3.495 mp_alltoall_i44 222 10.7 1.453 3.494 1.453 3.494 ls_scf_store_result 1 5.0 0.000 0.000 3.417 3.426 calculate_norms 792 9.8 3.258 3.294 3.258 3.294 dbcsr_finalize 304 7.8 0.082 0.089 3.101 3.157 qs_energies_init_hamiltonians 1 3.0 0.035 0.043 2.914 2.914 dbcsr_complete_redistribute 5 7.6 1.442 1.477 2.782 2.912 dbcsr_merge_all 275 8.9 0.893 0.921 2.869 2.912 dbcsr_data_release 12724 10.6 2.350 2.855 2.350 2.855 matrix_ls_to_qs 2 6.0 0.000 0.000 2.418 2.551 dbcsr_sort_data 325 11.1 2.438 2.505 2.438 2.505 dbcsr_new_transposed 4 7.5 0.252 0.270 2.296 2.315 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.263 2.264 dbcsr_frobenius_norm 74 6.6 2.055 2.135 2.207 2.250 dbcsr_add_d 103 6.2 0.000 0.000 2.154 2.235 dbcsr_add_anytype 103 7.2 0.859 0.894 2.153 2.235 rebuild_ks_matrix 3 7.3 0.000 0.000 2.195 2.197 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.195 2.197 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.559000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6765.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 78096bc50ec3c41c1825630b7cc5ec9938bfc15f Summary: empty Status: OK