=== 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: e3281e024bb35180d3a2a853d323285ecfbece49 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/01 job id: 44493188 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/02 job id: 44493189 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/03 job id: 44493190 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/04 job id: 44493191 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/05 job id: 44493192 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/06 job id: 44493193 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/07 job id: 44493194 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/08 job id: 44493195 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/09 job id: 44493197 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/10 job id: 44493199 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/11 job id: 44493200 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/12 job id: 44493201 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/13 job id: 44493202 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/14 job id: 44493203 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/15 job id: 44493204 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/16 job id: 44493205 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/17 job id: 44493206 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/18 job id: 44493207 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/19 job id: 44493208 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/20 job id: 44493209 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/21 job id: 44493210 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/22 job id: 44493213 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/23 job id: 44493214 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/24 job id: 44493215 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/25 job id: 44493216 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/26 job id: 44493217 --- 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/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 134.281 134.282 farming_run 1 2.0 133.636 133.654 134.250 134.253 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460150E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.024 115.276 115.276 qs_energies 1 2.0 0.000 0.000 115.074 115.076 mp2_main 1 3.0 0.000 0.000 113.204 113.205 mp2_gpw_main 1 4.0 0.023 0.029 112.395 112.397 mp2_ri_gpw_compute_in 1 5.0 0.177 0.182 93.331 93.868 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.416 55.956 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.729 47.228 get_2c_integrals 1 6.0 0.000 0.001 37.184 37.735 integrate_v_rspace 273 8.0 0.437 0.448 25.110 30.226 pw_transfer 6555 10.6 0.374 0.380 27.349 27.874 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.034 26.573 grid_integrate_task_list 273 9.0 20.930 26.510 20.930 26.510 fft_wrap_pw1pw2_100 2178 12.4 1.172 1.264 23.552 24.105 compute_2c_integrals 1 7.0 0.002 0.003 19.214 19.218 rpa_ri_compute_en 1 5.0 0.002 0.003 18.955 19.190 compute_2c_integrals_loop_lm 1 8.0 0.006 0.008 18.836 18.946 mp2_eri_2c_integrate_gpw 1 9.0 2.385 2.430 18.830 18.940 cp_fm_cholesky_decompose 12 8.2 17.927 18.498 17.927 18.498 cholesky_decomp 1 7.0 0.000 0.000 16.821 17.388 fft3d_s 5443 13.4 16.108 16.596 16.130 16.615 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.571 10.850 15.571 calculate_wavefunction 272 8.0 5.421 5.602 12.541 13.198 rpa_num_int 1 6.0 0.000 0.001 10.546 10.546 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.520 10.544 calc_mat_Q 8 8.0 0.000 0.000 9.357 9.476 contract_S_to_Q 8 9.0 0.000 0.000 8.777 8.899 calc_potential_gpw 544 9.5 0.005 0.006 8.275 8.636 parallel_gemm_fm 14 9.1 0.000 0.000 8.367 8.455 parallel_gemm_fm_cosma 14 10.1 8.367 8.455 8.367 8.455 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.195 8.417 potential_pw2rs 545 10.0 0.107 0.109 7.640 8.268 create_integ_mat 1 6.0 0.022 0.027 7.773 7.773 collocate_single_gaussian 272 10.0 0.040 0.042 7.424 7.628 array2fm 1 7.0 0.000 0.000 6.702 7.212 pw_scatter_s 2720 13.7 4.463 4.687 4.463 4.687 pw_gather_s 2722 13.2 3.878 4.231 3.878 4.231 array2fm_buffer_send 1 8.0 3.020 3.142 3.020 3.142 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.394480, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 0.028 0.037 397.502 397.504 farming_run 1 2.0 396.683 396.688 397.466 397.469 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.222902E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 210.060 210.061 qs_energies 1 2.0 0.000 0.000 209.841 209.849 scf_env_do_scf 1 3.0 0.000 0.000 107.131 107.132 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.251 106.259 rebuild_ks_matrix 4 6.0 0.000 0.000 106.249 106.258 qs_ks_build_kohn_sham_matrix 4 7.0 0.058 0.066 106.249 106.258 hfx_ks_matrix 4 8.0 0.001 0.001 105.864 105.867 integrate_four_center 4 9.0 0.144 0.446 105.863 105.866 mp2_main 1 3.0 0.002 0.019 102.420 102.428 mp2_gpw_main 1 4.0 0.042 0.138 101.607 101.618 integrate_four_center_main 4 10.0 0.097 0.519 96.852 99.824 integrate_four_center_bin 265 11.0 96.755 99.718 96.755 99.718 init_scf_loop 1 4.0 0.000 0.000 92.849 92.849 mp2_ri_gpw_compute_in 1 5.0 0.068 0.095 74.893 75.989 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.531 55.624 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 42.219 47.467 integrate_v_rspace 95 8.0 0.398 0.561 28.573 33.643 pw_transfer 2240 10.6 0.144 0.162 29.951 30.400 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.966 29.410 grid_integrate_task_list 95 9.0 23.858 29.127 23.858 29.127 mp2_ri_gpw_compute_en 1 5.0 0.059 0.110 26.484 28.238 fft_wrap_pw1pw2_100 730 12.4 1.274 1.458 26.675 27.100 ao_to_mo_and_store_B_mult_1 91 7.0 10.629 26.993 10.629 26.993 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.898 24.672 24.683 get_2c_integrals 1 6.0 0.002 0.015 20.267 20.301 compute_2c_integrals 1 7.0 0.003 0.005 19.248 19.261 compute_2c_integrals_loop_lm 1 8.0 0.004 0.025 18.876 19.134 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.842 18.872 19.132 fft3d_s 1823 13.4 18.440 18.831 18.453 18.845 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.281 14.281 calculate_wavefunction 91 8.0 2.028 2.069 9.756 10.010 potential_pw2rs 186 10.0 0.034 0.035 8.657 9.155 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.583 8.687 9.019 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.246 8.618 local_gemm 172 8.0 8.130 8.447 8.130 8.447 mp2_ri_gpw_compute_en_comm 22 7.0 0.504 0.535 7.745 8.274 collocate_single_gaussian 91 10.0 0.018 0.051 7.882 8.123 calc_potential_gpw 182 9.5 0.002 0.002 7.904 8.104 mp_sync 38 10.4 3.656 7.082 3.656 7.082 mp2_ri_gpw_compute_en_ener 172 7.0 6.353 6.467 6.353 6.467 mp_sendrecv_dm3 2068 8.0 5.766 6.262 5.766 6.262 pw_gather_s 912 13.2 4.929 5.433 4.929 5.433 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.609574, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.624960E+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 578578. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.029 53.810 53.812 qs_mol_dyn_low 1 2.0 0.003 0.004 53.487 53.496 qs_forces 11 3.9 0.001 0.002 53.427 53.428 qs_energies 11 4.9 0.008 0.056 51.873 51.894 scf_env_do_scf 11 5.9 0.000 0.001 45.643 45.645 scf_env_do_scf_inner_loop 108 6.5 0.004 0.018 43.542 43.544 dbcsr_multiply_generic 2286 12.5 0.100 0.103 34.694 35.225 qs_scf_new_mos 108 7.5 0.000 0.001 33.475 33.751 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.475 33.751 ot_scf_mini 108 9.5 0.005 0.031 31.835 32.022 multiply_cannon 2286 13.5 0.180 0.186 26.729 28.525 multiply_cannon_loop 2286 14.5 1.502 1.579 26.073 27.876 velocity_verlet 10 3.0 0.001 0.001 26.143 26.143 ot_mini 108 10.5 0.001 0.001 20.411 20.657 qs_ot_get_derivative 108 11.5 0.001 0.001 17.287 17.475 mp_waitall_1 245248 16.5 9.222 15.237 9.222 15.237 multiply_cannon_metrocomm3 54864 15.5 0.069 0.074 6.543 13.480 multiply_cannon_multrec 54864 15.5 4.150 6.419 7.620 11.095 rebuild_ks_matrix 119 8.3 0.000 0.000 7.935 8.077 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.013 7.934 8.077 mp_sum_l 7207 12.9 6.069 7.802 6.069 7.802 multiply_cannon_sync_h2d 54864 15.5 5.860 7.462 5.860 7.462 qs_ot_get_p 119 10.4 0.001 0.001 6.767 7.067 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.939 7.056 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.781 6.233 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.959 6.040 init_scf_run 11 5.9 0.000 0.001 4.925 4.925 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.924 4.925 dbcsr_mm_accdrv_process 76910 16.1 1.198 1.849 3.393 4.707 sum_up_and_integrate 119 10.3 0.012 0.015 4.533 4.548 integrate_v_rspace 119 11.3 0.002 0.002 4.521 4.537 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.135 4.245 calculate_rho_elec 119 8.7 0.011 0.017 4.135 4.244 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.730 3.772 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 2.062 3.524 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.919 3.122 apply_single 119 13.6 0.000 0.000 2.919 3.122 calculate_dm_sparse 119 9.5 0.000 0.001 2.961 3.098 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.838 2.839 ot_diis_step 108 11.5 0.006 0.006 2.801 2.802 jit_kernel_multiply 13 15.8 2.133 2.781 2.133 2.781 calculate_first_density_matrix 1 7.0 0.001 0.006 2.769 2.771 rs_pw_transfer 974 11.9 0.011 0.012 2.649 2.737 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.650 2.651 cp_fm_redistribute_end 50 14.0 2.414 2.631 2.419 2.632 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.507 2.596 cp_fm_diag_elpa_base 50 14.0 0.211 2.553 0.212 2.562 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.362 2.364 density_rs2pw 119 9.7 0.004 0.005 2.159 2.289 acc_transpose_blocks 54864 15.5 0.232 0.253 1.786 2.215 grid_integrate_task_list 119 12.3 2.034 2.149 2.034 2.149 mp_sum_d 4129 12.0 1.513 2.120 1.513 2.120 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.042 2.104 wfi_extrapolate 11 7.9 0.001 0.001 2.101 2.101 init_scf_loop 11 6.9 0.002 0.014 2.084 2.087 potential_pw2rs 119 12.3 0.004 0.004 1.880 1.898 pw_transfer 1439 11.6 0.053 0.059 1.803 1.879 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.726 1.807 make_m2s 4572 13.5 0.054 0.056 1.579 1.636 fft3d_ps 1201 14.6 0.369 0.470 1.497 1.571 make_images 4572 14.5 0.132 0.138 1.497 1.553 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.481 1.502 fft_wrap_pw1pw2_140 487 13.2 0.082 0.096 1.345 1.428 mp_waitany 12084 13.8 1.240 1.411 1.240 1.411 mp_alltoall_d11v 2130 13.8 1.228 1.372 1.228 1.372 grid_collocate_task_list 119 9.7 1.291 1.366 1.291 1.366 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.878 1.250 mp_alltoall_i22 627 13.8 0.811 1.086 0.811 1.086 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.812000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.818182, yerr=0.385695 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.570880E+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 1030242. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.037 40.806 40.810 qs_mol_dyn_low 1 2.0 0.005 0.016 40.503 40.510 qs_forces 11 3.9 0.003 0.011 40.440 40.442 qs_energies 11 4.9 0.002 0.011 38.769 38.775 scf_env_do_scf 11 5.9 0.001 0.001 32.196 32.197 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 29.789 29.789 dbcsr_multiply_generic 2286 12.5 0.105 0.109 23.550 23.926 qs_scf_new_mos 108 7.5 0.001 0.001 20.672 20.919 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.671 20.919 multiply_cannon 2286 13.5 0.208 0.217 18.087 20.302 ot_scf_mini 108 9.5 0.002 0.003 19.795 19.970 multiply_cannon_loop 2286 14.5 0.906 0.986 17.013 19.288 velocity_verlet 10 3.0 0.001 0.001 17.992 17.993 ot_mini 108 10.5 0.001 0.001 12.615 12.851 mp_waitall_1 200699 16.5 5.959 11.549 5.959 11.549 qs_ot_get_derivative 108 11.5 0.001 0.003 10.161 10.339 multiply_cannon_metrocomm3 27432 15.5 0.069 0.072 4.516 9.962 multiply_cannon_multrec 27432 15.5 1.971 4.543 5.851 8.638 rebuild_ks_matrix 119 8.3 0.000 0.000 7.150 7.290 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.150 7.290 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.321 6.448 dbcsr_mm_accdrv_process 47894 16.0 3.050 5.170 3.811 5.622 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.970 5.402 init_scf_run 11 5.9 0.000 0.001 5.285 5.285 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.285 5.285 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.607 5.124 apply_single 119 13.6 0.000 0.000 3.607 5.123 mp_sum_l 7207 12.9 2.611 5.011 2.611 5.011 qs_ot_get_p 119 10.4 0.001 0.001 4.317 4.545 acc_transpose_blocks 27432 15.5 0.108 0.113 2.685 4.370 sum_up_and_integrate 119 10.3 0.024 0.026 4.152 4.158 integrate_v_rspace 119 11.3 0.002 0.002 4.127 4.134 calculate_first_density_matrix 1 7.0 0.000 0.000 3.794 3.796 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.752 3.789 calculate_rho_elec 119 8.7 0.021 0.024 3.752 3.789 acc_transpose_blocks_kernels 27432 16.5 0.183 0.273 2.151 3.745 calculate_dm_sparse 119 9.5 0.000 0.001 3.458 3.531 jit_kernel_transpose 5 15.5 1.968 3.476 1.968 3.476 rs_pw_transfer 974 11.9 0.010 0.011 2.549 2.981 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.928 2.949 multiply_cannon_sync_h2d 27432 15.5 2.185 2.834 2.185 2.834 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.687 2.695 make_m2s 4572 13.5 0.051 0.053 2.429 2.668 make_images 4572 14.5 0.200 0.238 2.341 2.577 density_rs2pw 119 9.7 0.004 0.004 2.084 2.520 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.469 2.469 ot_diis_step 108 11.5 0.011 0.011 2.404 2.405 init_scf_loop 11 6.9 0.000 0.001 2.388 2.388 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.099 2.191 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.155 2.156 cp_fm_redistribute_end 50 14.0 1.786 2.134 1.790 2.135 cp_fm_diag_elpa_base 50 14.0 0.331 2.028 0.344 2.079 pw_transfer 1439 11.6 0.065 0.072 1.964 2.003 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.918 1.920 grid_integrate_task_list 119 12.3 1.835 1.916 1.835 1.916 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.872 1.914 potential_pw2rs 119 12.3 0.006 0.006 1.886 1.894 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.702 1.744 fft3d_ps 1201 14.6 0.513 0.563 1.572 1.611 jit_kernel_multiply 8 16.3 0.709 1.569 0.709 1.569 prepare_preconditioner 11 7.9 0.000 0.000 1.498 1.526 make_preconditioner 11 8.9 0.000 0.002 1.498 1.526 fft_wrap_pw1pw2_140 487 13.2 0.079 0.085 1.484 1.524 make_images_data 4572 15.5 0.044 0.051 1.124 1.508 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.388 1.443 wfi_extrapolate 11 7.9 0.001 0.001 1.441 1.441 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.974 1.418 grid_collocate_task_list 119 9.7 1.227 1.372 1.227 1.372 mp_allgather_i34 2286 14.5 0.518 1.305 0.518 1.305 mp_alltoall_d11v 2130 13.8 1.145 1.239 1.145 1.239 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.155 1.205 rs_pw_transfer_RS2PW_140 130 11.5 0.143 0.152 0.583 1.016 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.006 1.009 mp_sum_d 4129 12.0 0.557 1.008 0.557 1.008 mp_waitany 5720 13.7 0.553 0.990 0.553 0.990 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.887 0.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.810000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.272727, yerr=1.542778 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 523.333632E+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 2177508. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 32.406 32.407 qs_mol_dyn_low 1 2.0 0.004 0.004 32.108 32.116 qs_forces 11 3.9 0.002 0.003 32.047 32.048 qs_energies 11 4.9 0.001 0.001 30.488 30.491 scf_env_do_scf 11 5.9 0.000 0.001 25.883 25.883 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.314 23.314 dbcsr_multiply_generic 2286 12.5 0.094 0.097 16.715 16.803 velocity_verlet 10 3.0 0.001 0.001 15.227 15.228 qs_scf_new_mos 108 7.5 0.001 0.001 15.149 15.159 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.148 15.159 ot_scf_mini 108 9.5 0.002 0.002 14.431 14.442 multiply_cannon 2286 13.5 0.203 0.220 13.550 14.349 multiply_cannon_loop 2286 14.5 0.638 0.668 12.741 13.536 ot_mini 108 10.5 0.001 0.001 9.110 9.124 qs_ot_get_derivative 108 11.5 0.001 0.001 7.618 7.628 multiply_cannon_multrec 18288 15.5 2.200 3.274 7.324 7.522 rebuild_ks_matrix 119 8.3 0.000 0.000 6.365 6.379 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.365 6.379 dbcsr_mm_accdrv_process 38222 16.0 4.344 6.207 5.041 6.264 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.624 5.636 sum_up_and_integrate 119 10.3 0.029 0.030 3.941 3.946 integrate_v_rspace 119 11.3 0.002 0.002 3.912 3.920 mp_waitall_1 158411 16.6 2.448 3.626 2.448 3.626 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.428 3.440 calculate_rho_elec 119 8.7 0.030 0.031 3.427 3.439 init_scf_run 11 5.9 0.000 0.001 3.426 3.426 scf_env_initial_rho_setup 11 6.9 0.012 0.017 3.426 3.426 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.680 3.315 qs_ot_get_p 119 10.4 0.001 0.001 3.144 3.160 init_scf_loop 11 6.9 0.000 0.000 2.553 2.553 jit_kernel_multiply 10 16.0 0.645 2.527 0.645 2.527 rs_pw_transfer 974 11.9 0.009 0.010 2.244 2.474 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.390 2.444 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.363 apply_single 119 13.6 0.000 0.000 2.035 2.363 calculate_first_density_matrix 1 7.0 0.000 0.000 2.205 2.218 density_rs2pw 119 9.7 0.004 0.004 1.956 2.191 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.091 2.099 pw_transfer 1439 11.6 0.065 0.070 1.926 1.933 grid_integrate_task_list 119 12.3 1.798 1.889 1.798 1.889 make_m2s 4572 13.5 0.044 0.045 1.707 1.872 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.833 1.843 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.830 1.830 calculate_dm_sparse 119 9.5 0.000 0.000 1.820 1.827 make_images 4572 14.5 0.190 0.202 1.623 1.786 prepare_preconditioner 11 7.9 0.000 0.000 1.730 1.751 make_preconditioner 11 8.9 0.000 0.000 1.730 1.751 potential_pw2rs 119 12.3 0.007 0.008 1.720 1.725 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.585 1.665 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.635 1.637 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.616 1.617 cp_fm_redistribute_end 50 14.0 1.206 1.598 1.207 1.598 cp_fm_diag_elpa_base 50 14.0 0.375 1.520 0.389 1.567 multiply_cannon_sync_h2d 18288 15.5 1.422 1.566 1.422 1.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.517 1.523 fft3d_ps 1201 14.6 0.522 0.540 1.511 1.520 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.505 1.510 fft_wrap_pw1pw2_140 487 13.2 0.091 0.094 1.474 1.485 ot_diis_step 108 11.5 0.011 0.012 1.474 1.474 mp_sum_l 7207 12.9 1.067 1.399 1.067 1.399 grid_collocate_task_list 119 9.7 1.215 1.363 1.215 1.363 acc_transpose_blocks 18288 15.5 0.075 0.076 1.247 1.286 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.194 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.163 1.163 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.956 0.958 make_images_data 4572 15.5 0.044 0.048 0.755 0.935 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.872 0.896 hybrid_alltoall_any 4725 16.4 0.054 0.113 0.646 0.857 acc_transpose_blocks_kernels 18288 16.5 0.211 0.219 0.805 0.834 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.811 0.813 mp_waitany 9880 13.7 0.522 0.770 0.522 0.770 rs_pw_transfer_RS2PW_140 130 11.5 0.129 0.136 0.518 0.752 cp_fm_cholesky_invert 11 10.9 0.731 0.734 0.731 0.734 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.652 0.724 mp_alltoall_z22v 1201 16.6 0.596 0.697 0.596 0.697 mp_alltoall_d11v 2130 13.8 0.620 0.688 0.620 0.688 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.407000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.090909, yerr=2.151302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 557.686784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.050 36.541 36.543 qs_mol_dyn_low 1 2.0 0.004 0.005 36.190 36.198 qs_forces 11 3.9 0.004 0.016 36.126 36.127 qs_energies 11 4.9 0.008 0.018 34.385 34.391 scf_env_do_scf 11 5.9 0.002 0.010 28.927 28.929 scf_env_do_scf_inner_loop 108 6.5 0.008 0.027 25.459 25.461 dbcsr_multiply_generic 2286 12.5 0.102 0.114 19.113 19.237 velocity_verlet 10 3.0 0.001 0.003 17.679 17.681 qs_scf_new_mos 108 7.5 0.001 0.001 16.949 16.999 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.949 16.999 ot_scf_mini 108 9.5 0.003 0.003 16.022 16.079 multiply_cannon 2286 13.5 0.222 0.254 15.511 15.955 multiply_cannon_loop 2286 14.5 0.936 0.969 14.569 14.935 ot_mini 108 10.5 0.001 0.001 10.063 10.135 multiply_cannon_multrec 27432 15.5 3.036 3.975 9.528 9.903 qs_ot_get_derivative 108 11.5 0.001 0.001 8.252 8.312 dbcsr_mm_accdrv_process 47916 15.9 5.692 8.129 6.399 8.309 rebuild_ks_matrix 119 8.3 0.000 0.000 6.622 6.669 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.621 6.669 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.872 5.914 init_scf_run 11 5.9 0.000 0.001 3.987 3.988 scf_env_initial_rho_setup 11 6.9 0.008 0.016 3.987 3.987 sum_up_and_integrate 119 10.3 0.036 0.043 3.841 3.850 integrate_v_rspace 119 11.3 0.005 0.015 3.805 3.815 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.583 3.618 calculate_rho_elec 119 8.7 0.040 0.046 3.582 3.618 qs_ot_get_p 119 10.4 0.005 0.035 3.411 3.490 init_scf_loop 11 6.9 0.001 0.003 3.438 3.442 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.677 3.161 calculate_first_density_matrix 1 7.0 0.003 0.013 2.621 2.626 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.116 2.576 apply_single 119 13.6 0.000 0.000 2.116 2.575 prepare_preconditioner 11 7.9 0.000 0.000 2.516 2.525 make_preconditioner 11 8.9 0.000 0.000 2.516 2.524 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.130 2.453 calculate_dm_sparse 119 9.5 0.000 0.000 2.315 2.369 mp_waitall_1 137007 16.6 1.625 2.247 1.625 2.247 make_m2s 4572 13.5 0.054 0.056 2.110 2.232 rs_pw_transfer 974 11.9 0.009 0.009 2.032 2.214 density_rs2pw 119 9.7 0.004 0.004 1.984 2.167 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.116 2.126 pw_transfer 1439 11.6 0.066 0.072 2.090 2.125 make_images 4572 14.5 0.271 0.336 2.003 2.124 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.997 2.037 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.889 1.919 grid_integrate_task_list 119 12.3 1.826 1.915 1.826 1.915 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.821 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.808 1.808 acc_transpose_blocks 27432 15.5 0.111 0.113 1.757 1.782 ot_diis_step 108 11.5 0.012 0.012 1.769 1.769 fft3d_ps 1201 14.6 0.561 0.614 1.670 1.703 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.656 1.699 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.648 1.662 mp_sum_l 7207 12.9 1.069 1.620 1.069 1.620 potential_pw2rs 119 12.3 0.008 0.009 1.592 1.608 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.505 1.514 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.496 1.497 jit_kernel_multiply 8 16.1 0.646 1.478 0.646 1.478 cp_fm_redistribute_end 50 14.0 0.992 1.475 0.993 1.475 cp_fm_diag_elpa_base 50 14.0 0.461 1.412 0.480 1.451 grid_collocate_task_list 119 9.7 1.221 1.350 1.221 1.350 wfi_extrapolate 11 7.9 0.001 0.001 1.310 1.310 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.703 1.216 qs_energies_init_hamiltonians 11 5.9 0.030 0.042 1.167 1.171 acc_transpose_blocks_kernels 27432 16.5 0.269 0.278 1.132 1.147 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.095 1.114 cp_fm_upper_to_full 72 13.5 0.796 1.103 0.796 1.103 multiply_cannon_sync_h2d 27432 15.5 0.991 1.070 0.991 1.070 dbcsr_complete_redistribute 329 12.2 0.122 0.149 0.765 1.035 make_images_data 4572 15.5 0.045 0.048 0.798 0.925 jit_kernel_transpose 5 15.6 0.862 0.874 0.862 0.874 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.791 0.871 mp_alltoall_d11v 2130 13.8 0.724 0.869 0.724 0.869 hybrid_alltoall_any 4725 16.4 0.061 0.151 0.699 0.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.845 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.567 0.834 cp_fm_cholesky_invert 11 10.9 0.743 0.746 0.743 0.746 mp_alltoall_z22v 1201 16.6 0.692 0.737 0.692 0.737 mp_alltoall_i22 627 13.8 0.433 0.736 0.433 0.736 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.543000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.909091, yerr=2.712079 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 602.947584E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.050 28.665 28.671 qs_mol_dyn_low 1 2.0 0.003 0.004 28.441 28.448 qs_forces 11 3.9 0.002 0.002 28.381 28.382 qs_energies 11 4.9 0.001 0.001 26.666 26.669 scf_env_do_scf 11 5.9 0.000 0.001 21.312 21.313 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.816 18.817 velocity_verlet 10 3.0 0.001 0.001 14.275 14.278 dbcsr_multiply_generic 2286 12.5 0.098 0.102 12.595 12.696 qs_scf_new_mos 108 7.5 0.001 0.001 10.826 10.850 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.825 10.850 multiply_cannon 2286 13.5 0.229 0.234 10.161 10.618 ot_scf_mini 108 9.5 0.002 0.002 10.161 10.185 multiply_cannon_loop 2286 14.5 0.331 0.343 9.267 9.469 rebuild_ks_matrix 119 8.3 0.000 0.000 6.009 6.028 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.008 6.028 multiply_cannon_multrec 9144 15.5 1.610 1.891 5.748 5.962 ot_mini 108 10.5 0.001 0.001 5.708 5.738 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.351 5.369 qs_ot_get_derivative 108 11.5 0.001 0.001 4.418 4.442 dbcsr_mm_accdrv_process 12550 15.8 3.264 4.061 4.037 4.126 init_scf_run 11 5.9 0.000 0.001 3.918 3.918 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.917 3.917 sum_up_and_integrate 119 10.3 0.038 0.040 3.658 3.663 integrate_v_rspace 119 11.3 0.003 0.003 3.620 3.625 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.545 3.551 calculate_rho_elec 119 8.7 0.060 0.061 3.544 3.550 calculate_first_density_matrix 1 7.0 0.000 0.000 2.790 2.791 qs_ot_get_p 119 10.4 0.001 0.001 2.703 2.748 init_scf_loop 11 6.9 0.000 0.000 2.477 2.478 calculate_dm_sparse 119 9.5 0.000 0.000 2.267 2.285 pw_transfer 1439 11.6 0.066 0.070 2.157 2.169 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.063 2.077 mp_waitall_1 115863 16.7 1.481 2.050 1.481 2.050 density_rs2pw 119 9.7 0.004 0.004 1.885 2.028 grid_integrate_task_list 119 12.3 1.845 1.915 1.845 1.915 make_m2s 4572 13.5 0.034 0.035 1.684 1.822 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.811 1.814 rs_pw_transfer 974 11.9 0.008 0.008 1.645 1.782 make_images 4572 14.5 0.268 0.302 1.595 1.732 fft3d_ps 1201 14.6 0.564 0.573 1.720 1.732 fft_wrap_pw1pw2_140 487 13.2 0.088 0.094 1.718 1.730 prepare_preconditioner 11 7.9 0.000 0.000 1.705 1.708 make_preconditioner 11 8.9 0.000 0.000 1.705 1.708 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.694 1.699 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.597 1.622 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.620 1.621 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.552 1.554 acc_transpose_blocks 9144 15.5 0.038 0.039 1.483 1.492 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.476 1.490 jit_kernel_multiply 8 16.0 0.735 1.410 0.735 1.410 potential_pw2rs 119 12.3 0.010 0.011 1.399 1.402 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.384 1.391 grid_collocate_task_list 119 9.7 1.270 1.385 1.270 1.385 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.356 1.357 cp_fm_redistribute_end 50 14.0 0.677 1.337 0.678 1.338 cp_fm_diag_elpa_base 50 14.0 0.615 1.269 0.658 1.325 ot_diis_step 108 11.5 0.013 0.013 1.277 1.277 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 1.242 1.246 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.232 1.233 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.162 1.181 apply_single 119 13.6 0.000 0.000 1.162 1.181 jit_kernel_transpose 5 15.6 1.124 1.127 1.124 1.127 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.102 1.114 wfi_extrapolate 11 7.9 0.001 0.001 1.079 1.079 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.777 0.984 make_images_data 4572 15.5 0.038 0.042 0.789 0.954 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.869 0.919 cp_fm_cholesky_invert 11 10.9 0.853 0.855 0.853 0.855 mp_alltoall_d11v 2130 13.8 0.736 0.817 0.736 0.817 multiply_cannon_sync_h2d 9144 15.5 0.714 0.792 0.714 0.792 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.778 0.780 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.684 0.736 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.710 0.719 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 0.455 0.714 mp_alltoall_z22v 1201 16.6 0.602 0.662 0.602 0.662 mp_allgather_i34 2286 14.5 0.211 0.621 0.211 0.621 yz_to_x 606 15.1 0.266 0.277 0.581 0.602 mp_waitany 5200 13.7 0.455 0.579 0.455 0.579 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.671000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.272727, yerr=4.534532 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 762.920960E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.111 41.605 41.609 qs_mol_dyn_low 1 2.0 0.006 0.007 41.303 41.310 qs_forces 11 3.9 0.002 0.003 41.226 41.228 qs_energies 11 4.9 0.001 0.001 39.243 39.246 scf_env_do_scf 11 5.9 0.001 0.001 33.122 33.122 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.384 25.385 velocity_verlet 10 3.0 0.001 0.001 23.129 23.136 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.708 17.872 qs_scf_new_mos 108 7.5 0.001 0.001 15.543 15.639 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.543 15.639 multiply_cannon 2286 13.5 0.305 0.314 13.949 14.883 ot_scf_mini 108 9.5 0.002 0.002 14.472 14.567 multiply_cannon_loop 2286 14.5 0.343 0.349 12.700 13.634 ot_mini 108 10.5 0.001 0.001 8.661 8.773 multiply_cannon_multrec 9144 15.5 3.398 4.829 8.511 8.591 init_scf_loop 11 6.9 0.000 0.000 7.711 7.713 rebuild_ks_matrix 119 8.3 0.000 0.000 7.383 7.524 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.382 7.524 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.680 6.809 qs_ot_get_derivative 108 11.5 0.001 0.001 6.652 6.750 prepare_preconditioner 11 7.9 0.000 0.000 6.726 6.740 make_preconditioner 11 8.9 0.000 0.000 6.726 6.740 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.296 6.619 dbcsr_mm_accdrv_process 12550 15.8 3.877 5.263 4.990 6.326 cp_fm_upper_to_full 72 14.2 3.163 4.536 3.163 4.536 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.250 4.254 calculate_rho_elec 119 8.7 0.118 0.121 4.250 4.253 init_scf_run 11 5.9 0.000 0.001 4.090 4.090 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.090 4.090 sum_up_and_integrate 119 10.3 0.065 0.067 4.010 4.020 integrate_v_rspace 119 11.3 0.003 0.003 3.945 3.957 qs_ot_get_p 119 10.4 0.001 0.001 3.113 3.251 mp_waitall_1 94719 16.7 2.297 3.220 2.297 3.220 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.460 2.874 pw_transfer 1439 11.6 0.069 0.069 2.800 2.804 dbcsr_complete_redistribute 329 12.2 0.286 0.291 1.944 2.751 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.701 2.705 calculate_first_density_matrix 1 7.0 0.000 0.000 2.656 2.658 calculate_dm_sparse 119 9.5 0.000 0.000 2.610 2.624 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.171 2.461 apply_single 119 13.6 0.000 0.000 2.171 2.460 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.648 2.453 make_m2s 4572 13.5 0.038 0.038 2.228 2.403 fft3d_ps 1201 14.6 0.595 0.602 2.324 2.329 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.289 2.296 mp_alltoall_i22 627 13.8 1.466 2.286 1.466 2.286 make_images 4572 14.5 0.361 0.394 2.105 2.270 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.425 2.223 density_rs2pw 119 9.7 0.004 0.004 2.185 2.200 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.274 2.130 grid_integrate_task_list 119 12.3 2.052 2.062 2.052 2.062 ot_diis_step 108 11.5 0.014 0.014 1.986 1.986 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.895 1.896 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.893 1.894 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.773 1.774 jit_kernel_multiply 10 15.7 1.086 1.753 1.086 1.753 mp_sum_l 7207 12.9 1.002 1.727 1.002 1.727 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.654 1.702 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.684 1.689 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.616 1.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.582 1.611 grid_collocate_task_list 119 9.7 1.486 1.495 1.486 1.495 rs_pw_transfer 974 11.9 0.009 0.009 1.411 1.438 potential_pw2rs 119 12.3 0.014 0.014 1.414 1.417 cp_fm_cholesky_invert 11 10.9 1.399 1.402 1.399 1.402 acc_transpose_blocks 9144 15.5 0.038 0.038 1.376 1.395 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.142 1.370 wfi_extrapolate 11 7.9 0.001 0.001 1.368 1.368 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.333 1.333 cp_fm_diag_elpa_base 50 14.0 1.190 1.240 1.331 1.331 make_images_data 4572 15.5 0.042 0.045 1.071 1.260 acc_transpose_blocks_kernels 9144 16.5 0.118 0.119 1.126 1.144 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.120 1.141 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.108 1.123 mp_alltoall_d11v 2130 13.8 1.059 1.091 1.059 1.091 multiply_cannon_sync_h2d 9144 15.5 1.044 1.047 1.044 1.047 jit_kernel_transpose 5 15.6 1.009 1.026 1.009 1.026 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.981 1.023 qs_create_task_list 11 7.9 0.009 0.009 0.946 0.960 generate_qs_task_list 11 8.9 0.370 0.388 0.937 0.951 yz_to_x 606 15.1 0.460 0.471 0.929 0.938 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.914 0.927 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.609000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.090909, yerr=15.658019 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 501.870592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 91.666 91.667 qs_mol_dyn_low 1 2.0 0.004 0.005 91.331 91.340 qs_forces 11 3.9 0.002 0.003 91.256 91.258 qs_energies 11 4.9 0.003 0.010 88.094 88.112 scf_env_do_scf 11 5.9 0.001 0.001 78.511 78.513 scf_env_do_scf_inner_loop 99 6.5 0.006 0.041 71.923 71.924 dbcsr_multiply_generic 2055 12.4 0.108 0.113 54.046 54.480 qs_scf_new_mos 99 7.5 0.000 0.001 51.406 51.534 qs_scf_loop_do_ot 99 8.5 0.001 0.001 51.406 51.534 ot_scf_mini 99 9.5 0.002 0.003 48.775 48.902 velocity_verlet 10 3.0 0.001 0.003 46.657 46.658 multiply_cannon 2055 13.4 0.175 0.182 43.072 43.819 multiply_cannon_loop 2055 14.4 1.557 1.604 42.002 42.731 ot_mini 99 10.5 0.001 0.001 29.039 29.177 qs_ot_get_derivative 99 11.5 0.001 0.001 22.152 22.281 multiply_cannon_multrec 49320 15.4 11.960 12.644 17.049 17.723 rebuild_ks_matrix 110 8.3 0.000 0.001 16.179 16.408 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.027 16.178 16.407 mp_waitall_1 220248 16.4 13.243 14.595 13.243 14.595 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.146 14.350 qs_ot_get_p 110 10.4 0.001 0.005 11.047 11.209 multiply_cannon_sync_h2d 49320 15.4 9.771 10.317 9.771 10.317 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.470 8.596 calculate_rho_elec 110 8.6 0.021 0.026 8.469 8.595 multiply_cannon_metrocomm3 49320 15.4 0.078 0.083 7.165 8.480 sum_up_and_integrate 110 10.3 0.036 0.043 8.427 8.441 integrate_v_rspace 110 11.3 0.002 0.003 8.390 8.413 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.802 8.387 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.597 apply_single 110 13.6 0.000 0.001 7.171 7.597 init_scf_run 11 5.9 0.000 0.001 7.328 7.328 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.328 7.328 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 7.011 7.082 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.883 6.913 mp_sum_l 6514 12.8 5.795 6.624 5.795 6.624 ot_diis_step 99 11.5 0.006 0.006 6.574 6.575 init_scf_loop 11 6.9 0.003 0.025 6.561 6.563 rs_pw_transfer 902 11.9 0.011 0.013 5.286 5.786 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.780 5.783 density_rs2pw 110 9.6 0.004 0.005 4.987 5.511 dbcsr_mm_accdrv_process 87628 16.1 2.127 2.226 4.969 5.299 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.005 5.007 cp_fm_redistribute_end 48 14.0 4.355 4.960 4.360 4.962 make_m2s 4110 13.4 0.061 0.065 4.719 4.851 cp_fm_diag_elpa_base 48 14.0 0.594 4.769 0.597 4.801 make_images 4110 14.4 0.177 0.192 4.623 4.760 pw_transfer 1331 11.6 0.055 0.066 4.499 4.611 wfi_extrapolate 11 7.9 0.001 0.001 4.568 4.569 multiply_cannon_metrocomm1 49320 15.4 0.059 0.064 3.169 4.531 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.411 4.527 prepare_preconditioner 11 7.9 0.000 0.004 4.279 4.294 make_preconditioner 11 8.9 0.001 0.005 4.279 4.294 calculate_dm_sparse 110 9.5 0.000 0.001 4.033 4.164 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.042 4.090 fft3d_ps 1111 14.6 0.780 0.887 3.940 4.041 potential_pw2rs 110 12.3 0.006 0.007 3.702 3.768 fft_wrap_pw1pw2_140 451 13.1 0.169 0.190 3.647 3.768 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.742 3.747 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.398 3.464 grid_integrate_task_list 110 12.3 3.231 3.378 3.231 3.378 mp_alltoall_d11v 2046 13.8 2.843 3.333 2.843 3.333 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.247 3.290 mp_alltoall_z22v 1111 16.6 2.397 2.699 2.397 2.699 mp_waitany 14300 13.8 2.106 2.662 2.106 2.662 calculate_first_density_matrix 1 7.0 0.000 0.002 2.652 2.657 jit_kernel_multiply 13 15.9 2.557 2.607 2.557 2.607 cp_fm_cholesky_invert 11 10.9 2.563 2.567 2.563 2.567 make_images_data 4110 15.4 0.042 0.045 2.168 2.453 acc_transpose_blocks 49320 15.4 0.227 0.236 2.219 2.281 grid_collocate_task_list 110 9.6 2.087 2.277 2.087 2.277 mp_sum_d 3883 11.9 1.742 2.262 1.742 2.262 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.182 2.205 hybrid_alltoall_any 4261 16.3 0.081 0.483 1.843 2.167 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=91.667000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.727273, yerr=2.299838 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 587.669504E+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 3329312. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.045 75.789 75.793 qs_mol_dyn_low 1 2.0 0.041 0.051 75.442 75.451 qs_forces 11 3.9 0.010 0.018 75.336 75.344 qs_energies 11 4.9 0.025 0.083 71.999 72.002 scf_env_do_scf 11 5.9 0.000 0.001 62.660 62.664 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 54.242 54.244 dbcsr_multiply_generic 2055 12.4 0.115 0.120 40.129 40.410 velocity_verlet 10 3.0 0.001 0.001 39.350 39.352 qs_scf_new_mos 99 7.5 0.001 0.001 36.591 36.755 qs_scf_loop_do_ot 99 8.5 0.001 0.001 36.590 36.754 ot_scf_mini 99 9.5 0.003 0.003 34.863 35.010 multiply_cannon 2055 13.4 0.223 0.245 32.551 33.885 multiply_cannon_loop 2055 14.4 0.938 0.957 31.082 32.129 ot_mini 99 10.5 0.001 0.001 20.627 20.792 multiply_cannon_multrec 24660 15.4 7.577 9.221 14.831 16.523 qs_ot_get_derivative 99 11.5 0.001 0.001 14.684 14.837 rebuild_ks_matrix 110 8.3 0.000 0.001 14.690 14.787 qs_ks_build_kohn_sham_matrix 110 9.3 0.018 0.037 14.689 14.787 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.045 13.132 mp_waitall_1 176588 16.5 8.529 11.646 8.529 11.646 multiply_cannon_metrocomm3 24660 15.4 0.071 0.074 5.484 8.427 init_scf_loop 11 6.9 0.004 0.030 8.381 8.382 multiply_cannon_sync_h2d 24660 15.4 7.019 7.908 7.019 7.908 qs_ot_get_p 110 10.4 0.001 0.001 7.432 7.604 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.591 7.361 apply_single 110 13.6 0.000 0.001 6.591 7.361 dbcsr_mm_accdrv_process 52282 16.1 4.931 6.389 7.096 7.356 sum_up_and_integrate 110 10.3 0.051 0.057 7.339 7.349 integrate_v_rspace 110 11.3 0.002 0.002 7.287 7.299 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.934 6.952 calculate_rho_elec 110 8.6 0.039 0.047 6.934 6.951 init_scf_run 11 5.9 0.000 0.001 6.658 6.659 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.658 6.659 prepare_preconditioner 11 7.9 0.000 0.003 6.130 6.157 make_preconditioner 11 8.9 0.000 0.000 6.129 6.157 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.710 5.876 ot_diis_step 99 11.5 0.010 0.010 5.862 5.862 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.143 5.826 qs_ot_p2m_diag 48 11.0 0.028 0.044 5.204 5.224 make_m2s 4110 13.4 0.057 0.060 4.610 5.142 make_images 4110 14.4 0.398 0.442 4.500 5.028 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.720 4.721 density_rs2pw 110 9.6 0.004 0.004 4.024 4.619 pw_transfer 1331 11.6 0.066 0.072 4.337 4.516 rs_pw_transfer 902 11.9 0.012 0.013 3.763 4.420 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.230 4.411 wfi_extrapolate 11 7.9 0.001 0.001 4.024 4.024 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.009 4.013 cp_fm_redistribute_end 48 14.0 2.988 3.963 2.991 3.965 cp_fm_diag_elpa_base 48 14.0 0.935 3.792 0.968 3.882 fft3d_ps 1111 14.6 1.100 1.324 3.583 3.755 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.503 3.585 fft_wrap_pw1pw2_140 451 13.1 0.199 0.216 3.396 3.578 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.385 3.386 grid_integrate_task_list 110 12.3 3.134 3.332 3.134 3.332 mp_sum_l 6514 12.8 2.369 3.194 2.369 3.194 potential_pw2rs 110 12.3 0.008 0.009 3.043 3.104 calculate_dm_sparse 110 9.5 0.001 0.001 3.047 3.084 make_images_data 4110 15.4 0.046 0.050 2.529 3.073 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.936 2.990 hybrid_alltoall_any 4261 16.3 0.101 0.443 2.192 2.989 jit_kernel_multiply 9 16.4 1.815 2.844 1.815 2.844 cp_fm_cholesky_invert 11 10.9 2.808 2.816 2.808 2.816 calculate_first_density_matrix 1 7.0 0.001 0.008 2.537 2.539 mp_alltoall_d11v 2046 13.8 2.094 2.502 2.094 2.502 grid_collocate_task_list 110 9.6 2.044 2.475 2.044 2.475 qs_ot_get_orbitals 99 10.5 0.000 0.001 2.082 2.109 qs_energies_init_hamiltonians 11 5.9 0.008 0.011 2.081 2.083 mp_waitany 10164 13.8 1.336 2.038 1.336 2.038 mp_alltoall_z22v 1111 16.6 1.762 1.896 1.762 1.896 mp_allgather_i34 2055 14.4 0.796 1.870 0.796 1.870 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.814 1.826 multiply_cannon_metrocomm4 22605 15.4 0.072 0.076 0.770 1.777 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.218 1.022 1.663 cp_fm_cholesky_decompose 22 10.9 1.638 1.644 1.638 1.644 dbcsr_complete_redistribute 325 12.2 0.239 0.302 1.350 1.614 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.552 1.565 mp_irecv_dv 57340 16.2 0.649 1.563 0.649 1.563 acc_transpose_blocks 24660 15.4 0.111 0.114 1.537 1.560 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=75.793000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.818182, yerr=7.030042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 661.118976E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.034 64.647 64.648 qs_mol_dyn_low 1 2.0 0.003 0.005 64.348 64.358 qs_forces 11 3.9 0.002 0.003 64.277 64.278 qs_energies 11 4.9 0.001 0.001 60.960 60.963 scf_env_do_scf 11 5.9 0.000 0.001 52.843 52.844 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.194 43.194 velocity_verlet 10 3.0 0.001 0.001 34.957 34.959 dbcsr_multiply_generic 2055 12.4 0.109 0.112 30.202 30.540 qs_scf_new_mos 99 7.5 0.001 0.001 27.477 27.570 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.477 27.570 ot_scf_mini 99 9.5 0.002 0.003 26.203 26.309 multiply_cannon 2055 13.4 0.211 0.223 23.410 24.994 multiply_cannon_loop 2055 14.4 0.618 0.633 22.170 23.420 ot_mini 99 10.5 0.001 0.001 15.168 15.286 rebuild_ks_matrix 110 8.3 0.000 0.001 12.899 13.041 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.015 12.898 13.040 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.347 11.475 multiply_cannon_multrec 16440 15.4 3.758 4.736 10.454 11.287 mp_waitall_1 139946 16.5 7.841 11.192 7.841 11.192 qs_ot_get_derivative 99 11.5 0.001 0.001 10.559 10.663 init_scf_loop 11 6.9 0.000 0.000 9.612 9.613 multiply_cannon_metrocomm3 16440 15.4 0.044 0.045 4.781 7.982 prepare_preconditioner 11 7.9 0.000 0.000 7.764 7.782 make_preconditioner 11 8.9 0.000 0.000 7.764 7.781 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.070 7.431 sum_up_and_integrate 110 10.3 0.059 0.060 6.919 6.934 integrate_v_rspace 110 11.3 0.002 0.003 6.860 6.875 dbcsr_mm_accdrv_process 34862 16.1 4.663 6.271 6.552 6.690 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.278 6.288 calculate_rho_elec 110 8.6 0.058 0.059 6.277 6.288 qs_ot_get_p 110 10.4 0.001 0.001 6.126 6.251 init_scf_run 11 5.9 0.000 0.001 5.685 5.686 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.685 5.686 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.115 5.578 apply_single 110 13.6 0.000 0.000 5.115 5.578 make_m2s 4110 13.4 0.050 0.051 4.457 4.834 make_images 4110 14.4 0.396 0.518 4.342 4.720 density_rs2pw 110 9.6 0.004 0.004 3.350 4.624 ot_diis_step 99 11.5 0.010 0.011 4.568 4.568 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.291 4.297 multiply_cannon_sync_h2d 16440 15.4 3.634 4.252 3.634 4.252 rs_pw_transfer 902 11.9 0.010 0.011 2.994 4.235 pw_transfer 1331 11.6 0.066 0.073 3.963 3.974 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.302 3.972 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.909 3.909 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.856 3.868 grid_integrate_task_list 110 12.3 3.162 3.402 3.162 3.402 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.312 3.314 cp_fm_redistribute_end 48 14.0 2.065 3.282 2.068 3.282 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.240 3.255 cp_fm_diag_elpa_base 48 14.0 1.151 3.129 1.210 3.237 fft3d_ps 1111 14.6 1.090 1.099 3.153 3.168 wfi_extrapolate 11 7.9 0.001 0.001 3.129 3.129 make_images_data 4110 15.4 0.043 0.047 2.613 3.087 cp_fm_cholesky_invert 11 10.9 3.022 3.031 3.022 3.031 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.335 3.013 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.785 2.786 jit_kernel_multiply 9 16.5 1.502 2.689 1.502 2.689 potential_pw2rs 110 12.3 0.011 0.011 2.608 2.630 calculate_dm_sparse 110 9.5 0.001 0.001 2.566 2.595 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.486 2.548 mp_waitany 17072 13.8 1.248 2.542 1.248 2.542 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.913 2.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.494 calculate_first_density_matrix 1 7.0 0.000 0.000 2.457 2.458 grid_collocate_task_list 110 9.6 2.079 2.457 2.079 2.457 mp_sum_l 6514 12.8 1.717 2.399 1.717 2.399 mp_irecv_dv 48980 15.7 0.844 2.383 0.844 2.383 mp_alltoall_d11v 2046 13.8 1.998 2.318 1.998 2.318 rs_pw_transfer_RS2PW_140 121 11.5 0.182 0.188 0.919 2.166 cp_fm_cholesky_decompose 22 10.9 2.090 2.115 2.090 2.115 dbcsr_complete_redistribute 325 12.2 0.325 0.370 1.595 2.063 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.992 1.993 cp_fm_upper_to_full 70 13.6 1.412 1.875 1.412 1.875 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.751 1.767 mp_allgather_i34 2055 14.4 0.558 1.732 0.558 1.732 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.046 1.519 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.475 1.487 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.360 1.482 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.330 1.338 mp_alltoall_z22v 1111 16.6 1.296 1.321 1.296 1.321 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=64.648000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=7.997934 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 737.734656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.076 71.793 71.794 qs_mol_dyn_low 1 2.0 0.003 0.004 71.201 71.211 qs_forces 11 3.9 0.002 0.006 71.132 71.133 qs_energies 11 4.9 0.001 0.004 67.415 67.418 scf_env_do_scf 11 5.9 0.001 0.001 58.171 58.174 scf_env_do_scf_inner_loop 99 6.5 0.004 0.017 45.897 45.897 velocity_verlet 10 3.0 0.001 0.001 40.611 40.614 dbcsr_multiply_generic 2055 12.4 0.116 0.118 32.187 32.567 qs_scf_new_mos 99 7.5 0.001 0.001 29.538 29.712 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.538 29.711 ot_scf_mini 99 9.5 0.003 0.003 27.796 27.970 multiply_cannon 2055 13.4 0.244 0.262 23.636 24.694 multiply_cannon_loop 2055 14.4 0.884 0.905 21.867 23.026 ot_mini 99 10.5 0.001 0.001 16.249 16.418 multiply_cannon_multrec 24660 15.4 4.182 6.718 12.599 13.844 rebuild_ks_matrix 110 8.3 0.000 0.000 13.445 13.759 qs_ks_build_kohn_sham_matrix 110 9.3 0.022 0.038 13.445 13.758 init_scf_loop 11 6.9 0.003 0.022 12.233 12.234 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.768 12.053 qs_ot_get_derivative 99 11.5 0.001 0.001 11.695 11.864 prepare_preconditioner 11 7.9 0.001 0.004 10.144 10.169 make_preconditioner 11 8.9 0.002 0.014 10.143 10.169 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.143 9.638 dbcsr_mm_accdrv_process 52304 16.0 6.972 8.390 8.272 9.217 mp_waitall_1 121746 16.5 6.027 9.076 6.027 9.076 sum_up_and_integrate 110 10.3 0.066 0.069 7.289 7.303 integrate_v_rspace 110 11.3 0.003 0.004 7.222 7.235 make_m2s 4110 13.4 0.059 0.062 6.193 6.807 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.785 6.806 calculate_rho_elec 110 8.6 0.077 0.081 6.784 6.806 make_images 4110 14.4 0.576 0.697 6.053 6.668 qs_ot_get_p 110 10.4 0.001 0.001 6.294 6.452 init_scf_run 11 5.9 0.000 0.001 6.373 6.374 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.373 6.373 cp_fm_upper_to_full 70 13.8 3.336 4.760 3.336 4.760 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 2.011 4.704 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.343 4.614 apply_single 110 13.6 0.000 0.000 4.342 4.614 ot_diis_step 99 11.5 0.011 0.011 4.493 4.494 pw_transfer 1331 11.6 0.065 0.074 4.389 4.434 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.283 4.332 dbcsr_complete_redistribute 325 12.2 0.414 0.455 3.146 4.325 qs_ot_p2m_diag 48 11.0 0.054 0.063 4.078 4.093 make_images_data 4110 15.4 0.046 0.050 3.301 4.041 density_rs2pw 110 9.6 0.004 0.005 3.523 4.031 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.697 3.793 hybrid_alltoall_any 4261 16.3 0.119 0.458 2.730 3.717 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.618 3.691 wfi_extrapolate 11 7.9 0.001 0.001 3.660 3.661 fft_wrap_pw1pw2_140 451 13.1 0.201 0.210 3.598 3.654 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.474 3.646 fft3d_ps 1111 14.6 1.092 1.130 3.573 3.605 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.572 3.573 grid_integrate_task_list 110 12.3 3.256 3.495 3.256 3.495 rs_pw_transfer 902 11.9 0.010 0.011 2.945 3.483 multiply_cannon_sync_h2d 24660 15.4 3.163 3.410 3.163 3.410 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.989 3.146 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.136 3.140 mp_alltoall_i22 605 13.7 1.860 3.127 1.860 3.127 calculate_dm_sparse 110 9.5 0.001 0.001 3.061 3.083 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.924 2.927 cp_fm_redistribute_end 48 14.0 1.461 2.882 1.463 2.884 cp_fm_diag_elpa_base 48 14.0 1.339 2.731 1.417 2.842 mp_alltoall_d11v 2046 13.8 2.519 2.791 2.519 2.791 potential_pw2rs 110 12.3 0.012 0.013 2.747 2.779 cp_fm_cholesky_invert 11 10.9 2.685 2.693 2.685 2.693 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.548 2.690 calculate_first_density_matrix 1 7.0 0.000 0.002 2.477 2.480 mp_sum_l 6514 12.8 1.706 2.476 1.706 2.476 grid_collocate_task_list 110 9.6 2.180 2.421 2.180 2.421 qs_energies_init_hamiltonians 11 5.9 0.003 0.013 2.379 2.380 mp_allgather_i34 2055 14.4 0.871 2.100 0.871 2.100 jit_kernel_multiply 11 15.8 0.970 1.993 0.970 1.993 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.925 1.952 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.823 1.853 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 0.614 1.762 mp_waitany 13376 13.8 1.250 1.751 1.250 1.751 mp_alltoall_z22v 1111 16.6 1.622 1.729 1.622 1.729 cp_fm_cholesky_decompose 22 10.9 1.679 1.729 1.679 1.729 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.597 1.698 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.554 1.568 acc_transpose_blocks 24660 15.4 0.106 0.109 1.516 1.538 mp_sum_d 3881 11.9 0.995 1.479 0.995 1.479 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.846 1.451 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=71.794000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.909091, yerr=11.555829 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 838.455296E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.060 63.988 63.991 qs_mol_dyn_low 1 2.0 0.003 0.004 63.506 63.521 qs_forces 11 3.9 0.002 0.002 63.438 63.439 qs_energies 11 4.9 0.001 0.001 59.570 59.573 scf_env_do_scf 11 5.9 0.000 0.001 50.305 50.305 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 40.850 40.851 velocity_verlet 10 3.0 0.001 0.001 35.368 35.371 dbcsr_multiply_generic 2055 12.4 0.106 0.107 25.293 25.626 qs_scf_new_mos 99 7.5 0.001 0.001 24.016 24.114 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.015 24.114 ot_scf_mini 99 9.5 0.002 0.002 22.670 22.743 multiply_cannon 2055 13.4 0.246 0.257 18.596 20.514 multiply_cannon_loop 2055 14.4 0.323 0.334 17.096 17.904 rebuild_ks_matrix 110 8.3 0.000 0.000 13.115 13.307 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.114 13.307 ot_mini 99 10.5 0.001 0.001 12.243 12.315 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.698 11.865 init_scf_loop 11 6.9 0.000 0.000 9.394 9.395 mp_waitall_1 103326 16.6 7.392 9.090 7.392 9.090 multiply_cannon_multrec 8220 15.4 3.192 4.562 7.413 8.425 qs_ot_get_derivative 99 11.5 0.001 0.001 8.143 8.211 prepare_preconditioner 11 7.9 0.000 0.000 7.561 7.567 make_preconditioner 11 8.9 0.000 0.000 7.561 7.567 sum_up_and_integrate 110 10.3 0.079 0.080 7.332 7.346 integrate_v_rspace 110 11.3 0.003 0.003 7.253 7.265 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.142 7.163 calculate_rho_elec 110 8.6 0.115 0.116 7.141 7.162 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.041 7.139 qs_ot_get_p 110 10.4 0.001 0.001 6.315 6.384 init_scf_run 11 5.9 0.000 0.001 6.068 6.068 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.067 6.068 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.442 5.375 dbcsr_mm_accdrv_process 17442 15.9 2.828 3.612 4.091 4.990 make_m2s 4110 13.4 0.038 0.040 4.721 4.985 make_images 4110 14.4 0.641 0.695 4.591 4.857 pw_transfer 1331 11.6 0.066 0.072 4.736 4.756 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.628 4.652 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.549 4.553 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.227 4.227 density_rs2pw 110 9.6 0.004 0.004 3.756 4.131 ot_diis_step 99 11.5 0.012 0.012 4.028 4.029 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.845 3.983 apply_single 110 13.6 0.000 0.000 3.845 3.983 fft_wrap_pw1pw2_140 451 13.1 0.214 0.217 3.919 3.946 fft3d_ps 1111 14.6 1.146 1.181 3.872 3.890 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.625 3.627 cp_fm_redistribute_end 48 14.0 0.918 3.565 0.924 3.567 cp_fm_cholesky_invert 11 10.9 3.554 3.559 3.554 3.559 grid_integrate_task_list 110 12.3 3.348 3.533 3.348 3.533 cp_fm_diag_elpa_base 48 14.0 2.447 3.345 2.629 3.511 rs_pw_transfer 902 11.9 0.010 0.011 2.886 3.259 wfi_extrapolate 11 7.9 0.001 0.001 3.054 3.054 multiply_cannon_sync_h2d 8220 15.4 2.896 3.010 2.896 3.010 make_images_data 4110 15.4 0.038 0.043 2.598 2.993 calculate_dm_sparse 110 9.5 0.001 0.001 2.942 2.984 calculate_first_density_matrix 1 7.0 0.000 0.000 2.861 2.868 hybrid_alltoall_any 4261 16.3 0.198 0.856 2.481 2.846 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.688 2.688 potential_pw2rs 110 12.3 0.015 0.016 2.672 2.687 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.640 2.641 grid_collocate_task_list 110 9.6 2.278 2.543 2.278 2.543 mp_alltoall_d11v 2046 13.8 2.234 2.530 2.234 2.530 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.406 2.439 cp_fm_cholesky_decompose 22 10.9 2.416 2.433 2.416 2.433 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.302 2.333 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.188 2.222 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.242 2.006 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.987 dbcsr_complete_redistribute 325 12.2 0.557 0.584 1.782 1.872 mp_allgather_i34 2055 14.4 0.682 1.863 0.682 1.863 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.821 1.836 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.757 1.783 mp_waitany 9240 13.8 1.298 1.713 1.298 1.713 mp_sum_l 6514 12.8 1.405 1.704 1.405 1.704 mp_alltoall_z22v 1111 16.6 1.615 1.647 1.615 1.647 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.526 1.646 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.381 1.418 x_to_yz 550 16.1 0.573 0.582 1.377 1.395 rs_gather_matrices 110 12.3 0.327 0.371 1.143 1.381 jit_kernel_multiply 8 15.8 0.956 1.373 0.956 1.373 yz_to_x 561 15.1 0.524 0.537 1.335 1.354 qs_create_task_list 11 7.9 0.003 0.006 1.230 1.335 generate_qs_task_list 11 8.9 0.379 0.447 1.228 1.332 rs_pw_transfer_RS2PW_140 121 11.5 0.165 0.170 0.892 1.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=63.991000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.000000, yerr=15.159755 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 1.348891E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.060 90.532 90.534 qs_mol_dyn_low 1 2.0 0.004 0.005 90.153 90.191 qs_forces 11 3.9 0.002 0.002 90.081 90.082 qs_energies 11 4.9 0.012 0.014 85.873 85.875 scf_env_do_scf 11 5.9 0.001 0.001 75.252 75.252 velocity_verlet 10 3.0 0.001 0.001 57.270 57.276 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 46.172 46.173 dbcsr_multiply_generic 2055 12.4 0.130 0.132 28.972 29.067 init_scf_loop 11 6.9 0.000 0.000 29.007 29.007 qs_scf_new_mos 99 7.5 0.001 0.001 27.402 27.436 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.401 27.435 prepare_preconditioner 11 7.9 0.000 0.000 26.927 26.933 make_preconditioner 11 8.9 0.000 0.000 26.927 26.933 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.128 26.391 ot_scf_mini 99 9.5 0.002 0.002 25.636 25.671 multiply_cannon 2055 13.4 0.356 0.390 21.692 22.498 multiply_cannon_loop 2055 14.4 0.342 0.346 19.939 20.475 cp_fm_upper_to_full 70 14.2 12.652 18.155 12.652 18.155 rebuild_ks_matrix 110 8.3 0.001 0.001 14.505 14.540 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.505 14.539 ot_mini 99 10.5 0.001 0.001 13.909 13.950 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.139 13.168 dbcsr_complete_redistribute 325 12.2 1.042 1.064 7.499 10.672 multiply_cannon_multrec 8220 15.4 4.368 4.561 9.543 9.608 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.395 9.572 qs_ot_get_derivative 99 11.5 0.001 0.001 9.338 9.386 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.784 8.933 mp_alltoall_i22 605 13.7 5.431 8.614 5.431 8.614 mp_waitall_1 84994 16.7 7.802 8.548 7.802 8.548 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.971 8.012 calculate_rho_elec 110 8.6 0.227 0.228 7.971 8.011 sum_up_and_integrate 110 10.3 0.150 0.152 7.622 7.635 integrate_v_rspace 110 11.3 0.004 0.004 7.472 7.486 qs_ot_get_p 110 10.4 0.001 0.001 6.545 6.599 init_scf_run 11 5.9 0.000 0.001 6.302 6.302 scf_env_initial_rho_setup 11 6.9 0.004 0.005 6.302 6.302 make_m2s 4110 13.4 0.043 0.044 5.330 5.709 cp_fm_cholesky_invert 11 10.9 5.629 5.633 5.629 5.633 make_images 4110 14.4 0.876 0.912 5.142 5.521 pw_transfer 1331 11.6 0.075 0.076 5.509 5.519 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.876 5.439 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.392 5.402 dbcsr_mm_accdrv_process 11614 15.7 3.296 3.615 5.032 5.237 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.633 5.101 apply_single 110 13.6 0.000 0.000 4.633 5.100 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.836 4.842 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.662 4.671 fft3d_ps 1111 14.6 1.293 1.300 4.585 4.596 ot_diis_step 99 11.5 0.015 0.016 4.524 4.524 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.377 4.377 density_rs2pw 110 9.6 0.004 0.004 3.959 3.985 multiply_cannon_sync_h2d 8220 15.4 3.950 3.954 3.950 3.954 grid_integrate_task_list 110 12.3 3.675 3.746 3.675 3.746 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.721 3.722 cp_fm_diag_elpa_base 48 14.0 3.193 3.373 3.719 3.719 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.683 3.684 wfi_extrapolate 11 7.9 0.001 0.001 3.637 3.637 hybrid_alltoall_any 4261 16.3 0.257 0.559 2.957 3.475 make_images_data 4110 15.4 0.041 0.044 2.922 3.444 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.928 3.377 calculate_dm_sparse 110 9.5 0.001 0.001 3.306 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.033 3.036 potential_pw2rs 110 12.3 0.021 0.021 2.677 2.684 grid_collocate_task_list 110 9.6 2.643 2.664 2.643 2.664 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.568 2.586 calculate_first_density_matrix 1 7.0 0.000 0.000 2.523 2.525 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.434 2.458 rs_pw_transfer 902 11.9 0.011 0.011 2.428 2.457 cp_fm_cholesky_decompose 22 10.9 2.328 2.347 2.328 2.347 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.214 2.274 mp_alltoall_d11v 2046 13.8 2.089 2.185 2.089 2.185 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.086 2.157 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.973 1.978 qs_create_task_list 11 7.9 0.000 0.001 1.904 1.952 generate_qs_task_list 11 8.9 0.736 0.789 1.903 1.951 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.861 1.901 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.834 1.846 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.534000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1216.636364, yerr=55.097059 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/15/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 1.086553E+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 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.736576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4089 56701. MP_Allreduce 11263 785. MP_Sync 168 MP_Alltoall 2210 2520169. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_comm_split 82 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.084 205.512 205.539 qs_mol_dyn_low 1 2.0 0.007 0.011 204.895 204.922 qs_forces 11 3.9 0.004 0.009 204.803 204.833 qs_energies 11 4.9 0.001 0.002 199.209 199.239 scf_env_do_scf 11 5.9 0.001 0.001 182.627 182.654 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 161.934 161.957 dbcsr_multiply_generic 2485 12.5 0.209 0.217 124.795 125.456 qs_scf_new_mos 116 7.6 0.001 0.001 121.828 122.062 qs_scf_loop_do_ot 116 8.6 0.001 0.001 121.827 122.062 velocity_verlet 10 3.0 0.001 0.001 121.684 121.701 ot_scf_mini 116 9.6 0.003 0.003 115.340 115.531 multiply_cannon 2485 13.5 0.236 0.247 101.752 103.764 multiply_cannon_loop 2485 14.5 2.099 2.167 99.607 101.852 ot_mini 116 10.6 0.001 0.001 67.064 67.282 multiply_cannon_multrec 59640 15.5 32.813 35.407 42.839 44.897 qs_ot_get_derivative 116 11.6 0.001 0.001 42.432 42.624 rebuild_ks_matrix 127 8.3 0.001 0.001 33.747 34.134 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.025 33.746 34.134 mp_waitall_1 264810 16.5 28.126 32.029 28.126 32.029 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.300 30.634 multiply_cannon_sync_h2d 59640 15.5 27.006 28.849 27.006 28.849 qs_ot_get_p 127 10.4 0.001 0.001 26.718 26.907 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.052 24.918 apply_single 127 13.6 0.001 0.001 24.052 24.918 ot_diis_step 116 11.6 0.007 0.008 24.252 24.257 init_scf_loop 11 6.9 0.000 0.000 20.615 20.619 qs_ot_p2m_diag 82 11.4 0.076 0.090 20.140 20.228 multiply_cannon_metrocomm3 59640 15.5 0.113 0.118 15.554 19.186 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 18.728 18.922 cp_dbcsr_syevd 82 12.4 0.004 0.005 17.541 17.543 prepare_preconditioner 11 7.9 0.000 0.000 15.892 15.931 make_preconditioner 11 8.9 0.000 0.000 15.892 15.931 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.129 15.289 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.749 14.754 cp_fm_redistribute_end 82 14.4 11.659 14.679 11.672 14.683 cp_fm_diag_elpa_base 82 14.4 2.964 14.326 2.997 14.436 sum_up_and_integrate 127 10.3 0.089 0.105 14.347 14.361 integrate_v_rspace 127 11.3 0.003 0.004 14.258 14.277 make_m2s 4970 13.5 0.103 0.113 13.640 14.075 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.866 13.974 calculate_rho_elec 127 8.7 0.045 0.064 13.865 13.973 make_images 4970 14.5 0.396 0.414 13.460 13.908 init_scf_run 11 5.9 0.000 0.001 12.398 12.400 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.398 12.400 density_rs2pw 127 9.7 0.006 0.007 7.250 10.397 dbcsr_mm_accdrv_process 123452 16.2 3.300 3.453 9.595 10.112 mp_sum_l 7804 13.0 8.116 9.350 8.116 9.350 cp_fm_cholesky_invert 11 10.9 9.134 9.143 9.134 9.143 wfi_extrapolate 11 7.9 0.001 0.001 9.077 9.079 rs_pw_transfer 1038 11.9 0.016 0.018 5.857 8.994 calculate_dm_sparse 127 9.5 0.001 0.001 8.304 8.402 pw_transfer 1535 11.6 0.076 0.103 8.063 8.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.156 8.298 multiply_cannon_metrocomm1 59640 15.5 0.087 0.092 6.055 8.152 fft_wrap_pw1pw2 1281 12.7 0.010 0.013 7.859 8.117 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.681 7.782 make_images_data 4970 15.5 0.065 0.070 6.701 7.661 grid_integrate_task_list 127 12.3 7.010 7.486 7.010 7.486 hybrid_alltoall_any 5155 16.4 0.287 2.240 5.888 7.248 fft_wrap_pw1pw2_140 519 13.2 0.441 0.506 6.841 7.154 fft3d_ps 1281 14.7 2.086 2.857 6.658 6.828 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.662 6.670 mp_waitany 15900 13.9 2.693 5.952 2.693 5.952 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.810 5.869 grid_collocate_task_list 127 9.7 4.524 5.735 4.524 5.735 mp_alltoall_d11v 2401 14.1 4.238 5.344 4.238 5.344 rs_pw_transfer_RS2PW_140 138 11.5 0.281 0.298 2.119 5.304 potential_pw2rs 127 12.3 0.009 0.011 4.774 4.798 cp_fm_cholesky_decompose 22 10.9 4.588 4.602 4.588 4.602 mp_sum_d 4448 12.1 3.641 4.462 3.641 4.462 jit_kernel_multiply 13 15.9 4.246 4.263 4.246 4.263 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.539000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.636986 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/16/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 2.183246E+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 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 832.077824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+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 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4093 57094. MP_Allreduce 11272 945. MP_Sync 170 MP_Alltoall 1969 5622776. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.042 191.404 191.405 qs_mol_dyn_low 1 2.0 0.003 0.004 191.023 191.037 qs_forces 11 3.9 0.003 0.003 190.939 190.940 qs_energies 11 4.9 0.006 0.007 184.149 184.159 scf_env_do_scf 11 5.9 0.001 0.001 167.821 167.832 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.846 134.849 velocity_verlet 10 3.0 0.001 0.001 120.022 120.024 dbcsr_multiply_generic 2507 12.6 0.185 0.191 97.855 99.120 qs_scf_new_mos 117 7.6 0.001 0.001 94.451 95.072 qs_scf_loop_do_ot 117 8.6 0.001 0.001 94.450 95.071 ot_scf_mini 117 9.6 0.004 0.004 89.642 90.294 multiply_cannon 2507 13.6 0.478 0.528 77.811 82.671 multiply_cannon_loop 2507 14.6 1.266 1.308 74.557 77.291 ot_mini 117 10.6 0.001 0.001 50.517 51.154 mp_waitall_1 214728 16.6 25.525 40.002 25.525 40.002 multiply_cannon_multrec 30084 15.6 21.643 25.454 31.235 35.371 rebuild_ks_matrix 128 8.3 0.001 0.001 33.460 34.067 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 33.460 34.067 init_scf_loop 11 6.9 0.000 0.000 32.887 32.888 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.131 30.673 multiply_cannon_metrocomm3 30084 15.6 0.095 0.102 16.719 29.592 qs_ot_get_derivative 117 11.6 0.001 0.002 28.442 29.068 prepare_preconditioner 11 7.9 0.000 0.000 28.460 28.519 make_preconditioner 11 8.9 0.000 0.000 28.460 28.519 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.152 27.671 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.148 23.339 apply_single 128 13.6 0.001 0.001 22.148 23.338 ot_diis_step 117 11.6 0.014 0.015 21.899 21.901 qs_ot_get_p 128 10.4 0.001 0.001 20.997 21.576 multiply_cannon_sync_h2d 30084 15.6 18.874 21.095 18.874 21.095 cp_fm_cholesky_invert 11 10.9 16.675 16.688 16.675 16.688 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.202 16.239 make_m2s 5014 13.6 0.090 0.095 14.273 15.794 make_images 5014 14.6 1.156 1.341 14.062 15.582 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.057 15.059 sum_up_and_integrate 128 10.3 0.114 0.133 14.887 14.916 integrate_v_rspace 128 11.3 0.003 0.004 14.773 14.807 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.281 14.311 calculate_rho_elec 128 8.7 0.088 0.105 14.281 14.310 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.987 11.999 cp_fm_redistribute_end 83 14.4 7.002 11.930 7.013 11.933 cp_fm_diag_elpa_base 83 14.4 4.681 11.423 4.897 11.795 init_scf_run 11 5.9 0.000 0.001 11.528 11.529 scf_env_initial_rho_setup 11 6.9 0.003 0.006 11.527 11.529 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.939 11.364 multiply_cannon_metrocomm4 27577 15.6 0.094 0.108 3.859 10.635 make_images_data 5014 15.6 0.063 0.072 8.724 10.590 mp_irecv_dv 69486 16.3 3.668 10.250 3.668 10.250 density_rs2pw 128 9.7 0.006 0.007 7.635 10.160 hybrid_alltoall_any 5200 16.5 0.339 1.493 7.508 9.764 dbcsr_mm_accdrv_process 62242 16.2 4.543 5.477 9.056 9.642 pw_transfer 1547 11.6 0.088 0.110 9.110 9.206 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 8.883 8.984 wfi_extrapolate 11 7.9 0.001 0.001 8.377 8.378 rs_pw_transfer 1046 11.9 0.014 0.016 5.790 8.332 fft_wrap_pw1pw2_140 523 13.2 0.472 0.530 7.833 7.955 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.901 7.664 grid_integrate_task_list 128 12.3 7.187 7.598 7.187 7.598 fft3d_ps 1291 14.7 2.752 2.934 7.245 7.287 cp_fm_cholesky_decompose 22 10.9 6.901 6.982 6.901 6.982 calculate_dm_sparse 128 9.5 0.001 0.001 6.486 6.642 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.069 6.079 mp_sum_l 7870 13.0 3.990 5.989 3.990 5.989 grid_collocate_task_list 128 9.7 4.704 5.849 4.704 5.849 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.337 5.410 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.221 5.395 potential_pw2rs 128 12.3 0.015 0.017 5.143 5.162 mp_allgather_i34 2507 14.6 1.824 5.118 1.824 5.118 mp_waitany 11748 13.9 2.465 5.009 2.465 5.009 mp_alltoall_d11v 2415 14.1 4.109 4.845 4.109 4.845 rs_pw_transfer_RS2PW_140 139 11.5 0.353 0.379 2.064 4.593 mp_sum_d 4468 12.1 2.640 3.992 2.640 3.992 dbcsr_complete_redistribute 395 12.7 0.765 0.857 3.077 3.925 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.405000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.454545, yerr=3.340213 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 944.574464E+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 4093 57087. MP_Allreduce 11271 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 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 0.039 0.084 177.909 177.914 qs_mol_dyn_low 1 2.0 0.004 0.005 177.234 177.254 qs_forces 11 3.9 0.003 0.006 177.145 177.148 qs_energies 11 4.9 0.001 0.002 170.509 170.518 scf_env_do_scf 11 5.9 0.001 0.001 154.991 154.992 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.939 119.940 velocity_verlet 10 3.0 0.001 0.001 113.418 113.420 dbcsr_multiply_generic 2507 12.6 0.181 0.184 82.064 83.213 qs_scf_new_mos 117 7.6 0.001 0.001 81.325 81.632 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.324 81.632 ot_scf_mini 117 9.6 0.003 0.004 77.192 77.534 multiply_cannon 2507 13.6 0.502 0.525 62.128 66.007 multiply_cannon_loop 2507 14.6 0.859 0.891 58.868 61.893 ot_mini 117 10.6 0.001 0.001 43.102 43.445 init_scf_loop 11 6.9 0.000 0.000 34.950 34.951 mp_waitall_1 170520 16.6 25.355 34.710 25.355 34.710 rebuild_ks_matrix 128 8.3 0.001 0.001 31.072 31.578 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.071 31.578 prepare_preconditioner 11 7.9 0.000 0.000 30.904 30.955 make_preconditioner 11 8.9 0.000 0.000 30.904 30.955 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.559 29.977 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.005 28.464 multiply_cannon_metrocomm3 20056 15.6 0.060 0.063 15.663 25.273 multiply_cannon_multrec 20056 15.6 13.345 16.089 21.972 24.668 qs_ot_get_derivative 117 11.6 0.001 0.002 23.143 23.486 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.465 21.822 apply_single 128 13.6 0.001 0.001 20.465 21.822 qs_ot_get_p 128 10.4 0.001 0.001 19.413 19.906 ot_diis_step 117 11.6 0.018 0.018 19.857 19.857 multiply_cannon_sync_h2d 20056 15.6 14.172 16.052 14.172 16.052 make_m2s 5014 13.6 0.081 0.087 14.633 15.767 make_images 5014 14.6 1.177 1.262 14.400 15.535 qs_ot_p2m_diag 83 11.4 0.265 0.272 15.262 15.271 sum_up_and_integrate 128 10.3 0.131 0.142 14.829 14.863 integrate_v_rspace 128 11.3 0.004 0.004 14.698 14.727 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.438 14.477 calculate_rho_elec 128 8.7 0.132 0.147 14.437 14.477 cp_fm_cholesky_invert 11 10.9 14.423 14.432 14.423 14.432 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.242 14.243 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.079 11.080 cp_fm_redistribute_end 83 14.4 4.187 11.012 4.203 11.016 cp_fm_diag_elpa_base 83 14.4 6.386 10.421 6.791 10.891 make_images_data 5014 15.6 0.060 0.068 9.016 10.604 init_scf_run 11 5.9 0.000 0.001 10.488 10.489 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.487 10.489 hybrid_alltoall_any 5200 16.5 0.433 1.989 7.848 9.596 multiply_cannon_metrocomm4 17549 15.6 0.060 0.069 3.504 9.545 density_rs2pw 128 9.7 0.006 0.006 7.357 9.418 pw_transfer 1547 11.6 0.087 0.108 9.215 9.323 mp_irecv_dv 50230 16.2 3.383 9.302 3.383 9.302 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.855 9.115 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.989 9.104 dbcsr_mm_accdrv_process 41502 16.2 4.428 4.933 8.080 8.195 fft_wrap_pw1pw2_140 523 13.2 0.478 0.522 7.957 8.087 grid_integrate_task_list 128 12.3 7.263 7.758 7.263 7.758 cp_fm_upper_to_full 105 14.5 5.783 7.503 5.783 7.503 wfi_extrapolate 11 7.9 0.001 0.001 7.426 7.426 fft3d_ps 1291 14.7 2.664 2.884 7.251 7.344 cp_fm_cholesky_decompose 22 10.9 7.311 7.343 7.311 7.343 rs_pw_transfer 1046 11.9 0.013 0.015 5.249 7.306 dbcsr_complete_redistribute 395 12.7 1.164 1.193 4.574 6.361 calculate_dm_sparse 128 9.5 0.001 0.001 5.744 5.841 grid_collocate_task_list 128 9.7 4.899 5.800 4.899 5.800 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.702 5.748 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.390 5.395 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.394 5.177 mp_sum_l 7870 13.0 3.541 5.135 3.541 5.135 mp_alltoall_d11v 2415 14.1 4.331 4.972 4.331 4.972 potential_pw2rs 128 12.3 0.020 0.022 4.924 4.944 mp_allgather_i34 2507 14.6 1.780 4.740 1.780 4.740 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.498 4.674 mp_waitany 11748 13.9 2.381 4.507 2.381 4.507 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.325 4.072 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.042 4.071 rs_pw_transfer_RS2PW_140 139 11.5 0.331 0.353 1.940 3.996 mp_alltoall_i22 716 14.1 1.913 3.892 1.913 3.892 qs_energies_init_hamiltonians 11 5.9 0.003 0.005 3.793 3.795 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.580 3.627 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=177.914000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.909091, yerr=11.942701 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.143063E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.083 188.992 189.008 qs_mol_dyn_low 1 2.0 0.004 0.005 188.474 188.489 qs_forces 11 3.9 0.005 0.015 188.363 188.379 qs_energies 11 4.9 0.035 0.050 181.172 181.197 scf_env_do_scf 11 5.9 0.001 0.001 163.959 163.969 velocity_verlet 10 3.0 0.001 0.001 124.806 124.818 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.756 117.769 dbcsr_multiply_generic 2507 12.6 0.191 0.199 79.813 80.510 qs_scf_new_mos 117 7.6 0.001 0.001 79.701 79.975 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.700 79.974 ot_scf_mini 117 9.6 0.003 0.004 75.002 75.319 multiply_cannon 2507 13.6 0.559 0.597 55.077 58.256 multiply_cannon_loop 2507 14.6 1.182 1.216 51.105 52.789 init_scf_loop 11 6.9 0.002 0.014 46.082 46.091 ot_mini 117 10.6 0.001 0.001 42.302 42.602 prepare_preconditioner 11 7.9 0.000 0.001 41.945 41.968 make_preconditioner 11 8.9 0.000 0.001 41.945 41.968 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.562 40.556 multiply_cannon_multrec 30084 15.6 14.210 19.192 26.092 30.940 rebuild_ks_matrix 128 8.3 0.001 0.001 29.977 30.208 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.027 29.976 30.208 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.022 27.233 mp_waitall_1 147882 16.7 17.018 26.103 17.018 26.103 qs_ot_get_derivative 117 11.6 0.001 0.002 22.696 23.021 make_m2s 5014 13.6 0.094 0.098 20.449 21.585 make_images 5014 14.6 1.962 2.321 20.144 21.281 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.008 19.537 apply_single 128 13.6 0.001 0.001 19.008 19.537 ot_diis_step 117 11.6 0.017 0.018 19.476 19.477 qs_ot_get_p 128 10.4 0.001 0.001 18.923 19.235 cp_fm_upper_to_full 105 14.7 11.025 16.288 11.025 16.288 cp_fm_cholesky_invert 11 10.9 15.878 15.887 15.878 15.887 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.128 15.178 calculate_rho_elec 128 8.7 0.175 0.190 15.128 15.178 sum_up_and_integrate 128 10.3 0.140 0.153 14.780 14.800 integrate_v_rspace 128 11.3 0.004 0.004 14.639 14.665 qs_ot_p2m_diag 83 11.4 0.343 0.392 14.610 14.662 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.299 14.635 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.345 13.347 multiply_cannon_sync_h2d 30084 15.6 11.752 12.783 11.752 12.783 make_images_data 5014 15.6 0.064 0.069 10.926 12.732 dbcsr_complete_redistribute 395 12.7 1.531 1.641 8.893 12.653 dbcsr_mm_accdrv_process 62264 16.2 7.337 8.108 11.452 11.915 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.552 11.282 hybrid_alltoall_any 5200 16.5 0.518 2.079 9.780 11.184 init_scf_run 11 5.9 0.000 0.001 11.116 11.117 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.116 11.117 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.274 10.277 cp_fm_redistribute_end 83 14.4 1.777 10.203 1.792 10.208 cp_fm_diag_elpa_base 83 14.4 7.804 9.590 8.383 10.074 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.359 10.011 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.505 9.743 pw_transfer 1547 11.6 0.088 0.107 9.373 9.465 mp_alltoall_i22 716 14.1 5.576 9.292 5.576 9.292 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.144 9.243 density_rs2pw 128 9.7 0.006 0.006 7.088 8.393 fft_wrap_pw1pw2_140 523 13.2 0.482 0.489 8.130 8.251 grid_integrate_task_list 128 12.3 7.518 7.934 7.518 7.934 cp_fm_cholesky_decompose 22 10.9 7.547 7.656 7.547 7.656 wfi_extrapolate 11 7.9 0.001 0.001 7.580 7.581 fft3d_ps 1291 14.7 2.776 2.900 7.368 7.442 multiply_cannon_metrocomm4 25070 15.6 0.075 0.084 2.750 7.105 mp_irecv_dv 76098 16.2 2.607 6.845 2.607 6.845 calculate_dm_sparse 128 9.5 0.001 0.001 6.644 6.736 mp_alltoall_d11v 2415 14.1 5.354 6.226 5.354 6.226 rs_pw_transfer 1046 11.9 0.013 0.015 4.689 6.067 grid_collocate_task_list 128 9.7 5.081 5.818 5.081 5.818 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.504 5.532 qs_energies_init_hamiltonians 11 5.9 0.004 0.007 4.642 4.645 potential_pw2rs 128 12.3 0.023 0.023 4.619 4.634 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.416 4.504 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.389 4.451 mp_allgather_i34 2507 14.6 2.062 4.379 2.062 4.379 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.282 4.343 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.008000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1084.090909, yerr=15.825495 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.496359E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.055 179.274 179.283 qs_mol_dyn_low 1 2.0 0.009 0.013 178.839 178.850 qs_forces 11 3.9 0.007 0.010 178.730 178.744 qs_energies 11 4.9 0.002 0.002 171.170 171.188 scf_env_do_scf 11 5.9 0.001 0.001 152.860 152.873 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.733 116.740 velocity_verlet 10 3.0 0.001 0.001 116.328 116.337 qs_scf_new_mos 118 7.6 0.001 0.001 77.150 77.306 qs_scf_loop_do_ot 118 8.6 0.001 0.001 77.149 77.305 dbcsr_multiply_generic 2527 12.6 0.196 0.202 75.962 76.641 ot_scf_mini 118 9.6 0.004 0.004 72.726 72.919 multiply_cannon 2527 13.6 0.590 0.631 55.889 59.556 multiply_cannon_loop 2527 14.6 0.451 0.465 51.216 52.649 ot_mini 118 10.6 0.001 0.001 40.367 40.570 init_scf_loop 11 6.9 0.000 0.000 35.972 35.975 mp_waitall_1 126780 16.7 26.614 34.638 26.614 34.638 prepare_preconditioner 11 7.9 0.000 0.000 31.986 32.010 make_preconditioner 11 8.9 0.000 0.000 31.986 32.010 rebuild_ks_matrix 129 8.3 0.001 0.001 30.506 30.682 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.506 30.682 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.869 30.113 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.759 27.921 multiply_cannon_multrec 10108 15.6 10.414 15.032 17.708 22.755 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 12.746 20.510 qs_ot_get_derivative 118 11.6 0.001 0.002 20.080 20.273 ot_diis_step 118 11.6 0.020 0.021 20.217 20.218 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.649 20.110 apply_single 129 13.6 0.001 0.001 19.648 20.110 qs_ot_get_p 129 10.4 0.001 0.001 19.313 19.521 make_m2s 5054 13.6 0.066 0.068 16.261 18.509 cp_fm_cholesky_invert 11 10.9 18.308 18.314 18.308 18.314 make_images 5054 14.6 2.289 2.748 15.953 18.202 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.781 15.830 calculate_rho_elec 129 8.7 0.261 0.271 15.780 15.829 sum_up_and_integrate 129 10.3 0.181 0.191 15.379 15.426 qs_ot_p2m_diag 83 11.4 0.495 0.501 15.383 15.399 integrate_v_rspace 129 11.3 0.004 0.004 15.198 15.253 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.230 14.232 multiply_cannon_sync_h2d 10108 15.6 11.570 12.550 11.570 12.550 make_images_data 5054 15.6 0.052 0.061 9.921 12.291 hybrid_alltoall_any 5240 16.5 0.833 3.706 9.708 12.195 init_scf_run 11 5.9 0.000 0.001 11.461 11.462 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.461 11.462 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.108 11.120 cp_fm_diag_elpa_base 83 14.4 10.852 10.939 11.097 11.108 pw_transfer 1559 11.6 0.086 0.093 10.060 10.099 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.835 9.879 fft_wrap_pw1pw2_140 527 13.2 0.499 0.521 8.653 8.699 multiply_cannon_metrocomm1 10108 15.6 0.029 0.030 4.652 8.534 grid_integrate_task_list 129 12.3 7.810 8.231 7.810 8.231 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.089 8.218 cp_fm_cholesky_decompose 22 10.9 8.022 8.148 8.022 8.148 density_rs2pw 129 9.7 0.005 0.006 7.235 8.120 fft3d_ps 1301 14.7 2.727 2.805 7.983 8.004 dbcsr_mm_accdrv_process 20926 16.1 2.729 3.462 6.923 7.542 wfi_extrapolate 11 7.9 0.001 0.001 7.468 7.468 mp_allgather_i34 2527 14.6 2.840 7.129 2.840 7.129 calculate_dm_sparse 129 9.5 0.001 0.001 6.969 7.087 mp_alltoall_d11v 2423 14.1 5.273 6.149 5.273 6.149 grid_collocate_task_list 129 9.7 5.435 6.102 5.435 6.102 dbcsr_complete_redistribute 395 12.7 2.173 2.297 5.394 5.768 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.887 5.754 mp_irecv_dv 29086 15.9 1.850 5.659 1.850 5.659 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.311 5.313 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.305 5.312 rs_pw_transfer 1054 12.0 0.012 0.013 4.241 5.176 potential_pw2rs 129 12.3 0.026 0.027 4.699 4.716 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.090 4.122 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.560 3.856 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.556 3.830 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.753 3.784 calculate_first_density_matrix 1 7.0 0.000 0.000 3.778 3.781 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.623 3.745 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.662 3.676 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.569 3.631 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=179.283000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1395.000000, yerr=41.759049 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 11.606412E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.014730E+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 716108490000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11005 1515. MP_Sync 86 MP_Alltoall 1700 36954382. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.103 301.134 301.136 qs_mol_dyn_low 1 2.0 0.003 0.005 300.535 300.546 qs_forces 11 3.9 0.003 0.003 299.966 299.969 qs_energies 11 4.9 0.001 0.002 290.562 290.569 scf_env_do_scf 11 5.9 0.001 0.001 268.213 268.222 velocity_verlet 10 3.0 0.001 0.001 215.105 215.172 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 142.615 142.617 init_scf_loop 11 6.9 0.000 0.000 125.338 125.341 prepare_preconditioner 11 7.9 0.000 0.000 120.210 120.239 make_preconditioner 11 8.9 0.000 0.000 120.210 120.239 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.189 117.353 qs_scf_new_mos 116 7.6 0.001 0.001 92.404 92.485 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.403 92.484 ot_scf_mini 116 9.6 0.004 0.004 87.665 87.736 dbcsr_multiply_generic 2485 12.5 0.212 0.217 83.344 83.936 cp_fm_upper_to_full 104 14.8 52.920 75.880 52.920 75.880 multiply_cannon 2485 13.5 0.707 0.755 58.238 59.406 multiply_cannon_loop 2485 14.5 0.469 0.482 54.171 55.730 ot_mini 116 10.6 0.001 0.001 44.800 44.865 dbcsr_complete_redistribute 393 12.7 3.972 4.016 29.729 42.259 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.299 38.843 rebuild_ks_matrix 127 8.3 0.001 0.001 37.902 37.985 qs_ks_build_kohn_sham_matrix 127 9.3 0.025 0.026 37.901 37.985 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.975 36.365 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.894 34.972 mp_alltoall_i22 712 14.1 21.771 34.584 21.771 34.584 cp_fm_cholesky_invert 11 10.9 33.992 33.998 33.992 33.998 mp_waitall_1 102768 16.8 28.254 32.483 28.254 32.483 qs_ot_get_p 127 10.4 0.001 0.001 27.410 27.448 qs_ot_get_derivative 116 11.6 0.002 0.002 24.257 24.328 qs_ot_p2m_diag 82 11.4 0.867 0.872 23.214 23.242 make_m2s 4970 13.5 0.075 0.077 20.358 21.727 cp_dbcsr_syevd 82 12.4 0.005 0.006 21.479 21.481 make_images 4970 14.5 3.725 3.843 19.882 21.254 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.507 20.516 calculate_rho_elec 127 8.7 0.479 0.480 20.507 20.515 ot_diis_step 116 11.6 0.021 0.022 20.509 20.509 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.814 20.028 apply_single 127 13.6 0.001 0.001 19.813 20.028 multiply_cannon_metrocomm3 9940 15.5 0.024 0.024 18.360 19.738 sum_up_and_integrate 127 10.3 0.318 0.320 19.253 19.337 integrate_v_rspace 127 11.3 0.004 0.004 18.935 19.019 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.212 18.212 cp_fm_diag_elpa_base 82 14.4 13.943 15.511 18.208 18.208 multiply_cannon_multrec 9940 15.5 10.398 12.142 17.750 17.804 multiply_cannon_sync_h2d 9940 15.5 15.529 15.538 15.529 15.538 hybrid_alltoall_any 5155 16.4 1.290 3.011 11.265 13.476 make_images_data 4970 15.5 0.059 0.064 11.108 13.312 init_scf_run 11 5.9 0.000 0.001 12.295 12.295 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.294 12.295 pw_transfer 1535 11.6 0.092 0.093 12.245 12.249 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 12.012 12.017 fft_wrap_pw1pw2_140 519 13.2 0.539 0.544 10.650 10.660 fft3d_ps 1281 14.7 2.718 2.729 10.060 10.066 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.501 9.556 wfi_extrapolate 11 7.9 0.001 0.001 9.169 9.169 cp_fm_cholesky_decompose 22 10.9 8.957 8.983 8.957 8.983 dbcsr_mm_accdrv_process 20590 16.0 3.745 5.345 7.115 8.904 grid_integrate_task_list 127 12.3 8.510 8.678 8.510 8.678 mp_alltoall_d11v 2401 14.1 7.787 8.577 7.787 8.577 density_rs2pw 127 9.7 0.005 0.005 8.266 8.394 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.065 8.065 calculate_dm_sparse 127 9.5 0.001 0.001 6.585 6.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.350 6.419 grid_collocate_task_list 127 9.7 6.309 6.349 6.309 6.349 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.094 6.153 rs_scatter_matrices 138 9.7 3.566 4.397 5.902 6.128 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.136000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2673.454545, yerr=176.300551 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 1.262256E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.034 85.026 85.035 qs_energies 1 2.0 0.000 0.000 84.578 84.588 ls_scf 1 3.0 0.000 0.001 83.653 83.663 dbcsr_multiply_generic 111 6.7 0.016 0.020 72.604 72.779 multiply_cannon 111 7.7 0.018 0.020 55.798 56.894 multiply_cannon_loop 111 8.7 0.211 0.225 52.406 53.694 ls_scf_main 1 4.0 0.000 0.000 52.442 52.448 density_matrix_trs4 2 5.0 0.002 0.003 46.901 46.981 ls_scf_init_scf 1 4.0 0.001 0.005 28.168 28.171 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.049 27.105 mp_waitall_1 11031 10.9 22.420 26.062 22.420 26.062 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.941 24.964 multiply_cannon_multrec 2664 9.7 8.173 8.925 15.489 17.291 multiply_cannon_sync_h2d 2664 9.7 13.725 15.073 13.725 15.073 make_m2s 222 7.7 0.008 0.011 13.180 13.645 make_images 222 8.7 0.099 0.109 13.158 13.626 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.506 12.007 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.477 9.079 make_images_data 222 9.7 0.004 0.005 7.767 8.354 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.639 6.936 7.933 hybrid_alltoall_any 227 10.6 0.215 1.839 6.643 7.881 dbcsr_mm_accdrv_process_sort 4760 11.4 6.223 7.131 6.223 7.131 calculate_norms 4752 9.8 5.551 6.267 5.551 6.267 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.021 5.205 mp_sum_l 807 5.4 3.089 5.141 3.089 5.141 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.280 4.017 make_images_sizes 222 9.7 0.000 0.000 0.741 3.587 mp_alltoall_i44 222 10.7 0.740 3.587 0.740 3.587 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.066 3.453 mp_irecv_dv 6231 10.9 2.048 3.426 2.048 3.426 arnoldi_extremal 4 6.8 0.000 0.000 3.261 3.288 arnoldi_normal_ev 4 7.8 0.001 0.003 3.261 3.288 build_subspace 16 8.4 0.009 0.012 3.162 3.164 ls_scf_post 1 4.0 0.000 0.001 3.042 3.049 ls_scf_store_result 1 5.0 0.000 0.000 2.862 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.318 2.738 dbcsr_merge_single_wm 555 10.7 0.457 0.586 2.311 2.730 make_images_pack 222 9.7 2.208 2.620 2.210 2.622 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.341 2.582 dbcsr_sort_data 658 11.4 2.110 2.473 2.110 2.473 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.465 2.073 2.467 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.308 2.390 buffer_matrices_ensure_size 222 8.7 1.750 2.057 1.750 2.057 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.800 1.801 rebuild_ks_matrix 3 7.3 0.000 0.000 1.791 1.792 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.791 1.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.035000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 2.105754E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.045 91.379 91.387 qs_energies 1 2.0 0.000 0.000 90.693 90.702 ls_scf 1 3.0 0.001 0.003 89.335 89.349 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.290 75.682 multiply_cannon 111 7.7 0.030 0.043 53.310 56.838 ls_scf_main 1 4.0 0.000 0.000 55.073 55.077 multiply_cannon_loop 111 8.7 0.116 0.122 50.042 53.135 density_matrix_trs4 2 5.0 0.002 0.003 49.279 49.466 mp_waitall_1 9105 10.9 21.463 31.007 21.463 31.007 ls_scf_init_scf 1 4.0 0.001 0.006 30.786 30.797 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.607 29.675 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.224 27.245 multiply_cannon_multrec 1332 9.7 13.067 16.534 22.272 26.900 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 12.012 21.490 make_m2s 222 7.7 0.006 0.007 15.383 16.004 make_images 222 8.7 1.577 1.978 15.353 15.975 dbcsr_mm_accdrv_process 4041 10.4 0.297 0.452 8.809 10.377 dbcsr_mm_accdrv_process_sort 4041 11.4 8.400 9.925 8.400 9.925 make_images_data 222 9.7 0.004 0.004 8.847 9.781 hybrid_alltoall_any 227 10.6 0.520 2.489 8.292 9.160 mp_sum_l 807 5.4 5.299 8.397 5.299 8.397 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.247 7.724 mp_irecv_dv 3311 11.0 3.229 7.672 3.229 7.672 calculate_norms 2376 9.8 5.983 6.839 5.983 6.839 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.130 6.799 multiply_cannon_sync_h2d 1332 9.7 4.721 5.956 4.721 5.956 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.057 5.239 arnoldi_extremal 4 6.8 0.000 0.000 4.668 4.688 arnoldi_normal_ev 4 7.8 0.001 0.005 4.667 4.688 build_subspace 16 8.4 0.014 0.021 4.411 4.414 ls_scf_post 1 4.0 0.000 0.001 3.475 3.479 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.126 3.372 ls_scf_store_result 1 5.0 0.000 0.000 3.223 3.300 dbcsr_matrix_vector_mult_local 304 10.0 2.729 3.213 2.731 3.214 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.230 2.983 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.645 2.758 make_images_pack 222 9.7 2.019 2.419 2.021 2.421 mp_allgather_i34 111 8.7 0.967 2.313 0.967 2.313 dbcsr_sort_data 436 11.2 1.837 2.101 1.837 2.101 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.902 1.905 rebuild_ks_matrix 3 7.3 0.000 0.000 1.889 1.892 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.006 1.889 1.892 dbcsr_data_new 4174 10.1 1.614 1.839 1.614 1.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.387000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1777.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 2.704421E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.101 94.183 94.190 qs_energies 1 2.0 0.000 0.000 93.456 93.461 ls_scf 1 3.0 0.000 0.000 92.037 92.040 dbcsr_multiply_generic 111 6.7 0.015 0.016 76.615 76.929 ls_scf_main 1 4.0 0.000 0.000 57.604 57.609 multiply_cannon 111 7.7 0.037 0.076 52.863 56.710 multiply_cannon_loop 111 8.7 0.099 0.104 49.158 53.178 density_matrix_trs4 2 5.0 0.002 0.003 51.525 51.720 mp_waitall_1 7281 11.0 23.518 33.894 23.518 33.894 ls_scf_init_scf 1 4.0 0.001 0.010 30.791 30.793 ls_scf_init_matrix_S 1 5.0 0.004 0.037 29.580 29.659 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.173 27.190 multiply_cannon_multrec 888 9.7 12.683 15.081 21.217 24.200 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.706 23.167 make_m2s 222 7.7 0.006 0.007 16.887 18.259 make_images 222 8.7 1.979 2.307 16.849 18.219 make_images_data 222 9.7 0.003 0.004 9.538 10.746 hybrid_alltoall_any 227 10.6 0.620 2.854 9.072 10.569 mp_sum_l 807 5.4 5.374 9.312 5.374 9.312 dbcsr_mm_accdrv_process 3754 10.4 0.253 0.417 8.054 9.250 dbcsr_mm_accdrv_process_sort 3754 11.4 7.683 8.833 7.683 8.833 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.078 7.408 multiply_cannon_sync_h2d 888 9.7 6.018 7.330 6.018 7.330 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.896 6.608 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.419 6.559 mp_irecv_dv 2335 11.1 2.404 6.516 2.404 6.516 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.004 5.262 arnoldi_extremal 4 6.8 0.000 0.000 5.152 5.163 arnoldi_normal_ev 4 7.8 0.003 0.008 5.152 5.163 build_subspace 16 8.4 0.014 0.020 4.839 4.846 calculate_norms 1584 9.8 4.292 4.637 4.292 4.637 mp_allgather_i34 111 8.7 1.541 3.860 1.541 3.860 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.435 3.768 ls_scf_post 1 4.0 0.000 0.001 3.642 3.647 dbcsr_matrix_vector_mult_local 304 10.0 2.999 3.572 3.001 3.574 ls_scf_store_result 1 5.0 0.000 0.000 3.358 3.470 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.888 3.001 make_images_sizes 222 9.7 0.000 0.000 1.045 2.274 mp_alltoall_i44 222 10.7 1.045 2.273 1.045 2.273 dbcsr_sort_data 325 11.1 1.866 2.203 1.866 2.203 make_images_pack 222 9.7 1.813 2.143 1.816 2.146 dbcsr_data_release 9322 10.9 1.314 2.051 1.314 2.051 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.944 1.946 rebuild_ks_matrix 3 7.3 0.000 0.000 1.926 1.928 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.926 1.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.190000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2190.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 3.371471E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.070 97.614 97.618 qs_energies 1 2.0 0.000 0.000 96.901 96.905 ls_scf 1 3.0 0.002 0.014 95.242 95.249 dbcsr_multiply_generic 111 6.7 0.016 0.018 78.919 79.142 ls_scf_main 1 4.0 0.000 0.000 59.284 59.286 multiply_cannon 111 7.7 0.063 0.157 51.693 56.387 density_matrix_trs4 2 5.0 0.002 0.003 53.078 53.278 multiply_cannon_loop 111 8.7 0.115 0.127 46.644 50.140 ls_scf_init_scf 1 4.0 0.000 0.005 32.727 32.731 ls_scf_init_matrix_S 1 5.0 0.000 0.002 31.454 31.516 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.984 28.996 mp_waitall_1 6369 11.0 22.997 28.677 22.997 28.677 multiply_cannon_multrec 1332 9.7 14.098 17.141 21.936 25.245 make_m2s 222 7.7 0.006 0.008 21.267 22.677 make_images 222 8.7 3.139 3.618 21.217 22.630 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.316 17.573 make_images_data 222 9.7 0.004 0.004 11.853 13.424 hybrid_alltoall_any 227 10.6 0.797 3.790 11.255 13.084 dbcsr_mm_accdrv_process 3641 10.4 0.219 0.400 7.483 8.996 dbcsr_mm_accdrv_process_sort 3641 11.4 7.110 8.580 7.110 8.580 mp_sum_l 807 5.4 4.380 8.500 4.380 8.500 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.353 6.719 multiply_cannon_sync_h2d 1332 9.7 5.502 6.095 5.502 6.095 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.091 5.987 mp_irecv_dv 3229 10.9 2.068 5.908 2.068 5.908 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.656 5.588 arnoldi_extremal 4 6.8 0.000 0.000 5.206 5.219 arnoldi_normal_ev 4 7.8 0.001 0.004 5.206 5.219 build_subspace 16 8.4 0.014 0.021 4.875 4.883 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.576 4.787 mp_allgather_i34 111 8.7 2.172 4.641 2.172 4.641 calculate_norms 2376 9.8 4.208 4.565 4.208 4.565 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.586 3.867 dbcsr_matrix_vector_mult_local 304 10.0 3.170 3.665 3.172 3.667 dbcsr_sort_data 658 11.4 3.111 3.434 3.111 3.434 dbcsr_special_finalize 555 9.7 0.006 0.007 2.856 3.241 ls_scf_post 1 4.0 0.000 0.001 3.228 3.235 dbcsr_merge_single_wm 555 10.7 0.539 0.670 2.847 3.233 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.072 3.133 ls_scf_store_result 1 5.0 0.000 0.000 2.972 3.028 dbcsr_data_release 10477 10.7 1.569 2.416 1.569 2.416 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.020 2.022 rebuild_ks_matrix 3 7.3 0.000 0.000 1.996 1.999 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.012 1.996 1.999 dbcsr_finalize 304 7.8 0.049 0.061 1.803 1.992 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.618000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 4.630340E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.058 100.031 100.032 qs_energies 1 2.0 0.000 0.000 99.189 99.196 ls_scf 1 3.0 0.000 0.000 97.229 97.241 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.461 78.706 ls_scf_main 1 4.0 0.000 0.000 62.642 62.642 multiply_cannon 111 7.7 0.088 0.175 55.627 60.402 density_matrix_trs4 2 5.0 0.002 0.003 55.316 55.399 multiply_cannon_loop 111 8.7 0.069 0.078 51.195 53.362 mp_waitall_1 5436 11.0 27.315 32.834 27.315 32.834 ls_scf_init_scf 1 4.0 0.000 0.000 31.025 31.031 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.641 29.684 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.465 27.477 multiply_cannon_multrec 444 9.7 13.933 15.932 20.964 23.724 make_m2s 222 7.7 0.004 0.005 17.886 20.311 make_images 222 8.7 3.726 4.401 17.824 20.251 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.824 17.052 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.056 15.087 make_images_data 222 9.7 0.003 0.004 10.235 12.467 hybrid_alltoall_any 227 10.6 0.788 3.760 10.003 12.369 multiply_cannon_sync_h2d 444 9.7 6.579 8.143 6.579 8.143 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.340 6.731 7.869 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.523 6.417 7.523 mp_allgather_i34 111 8.7 2.661 7.022 2.661 7.022 arnoldi_extremal 4 6.8 0.000 0.000 5.807 5.817 arnoldi_normal_ev 4 7.8 0.006 0.019 5.807 5.817 build_subspace 16 8.4 0.015 0.020 5.402 5.411 mp_sum_l 807 5.4 2.999 4.787 2.999 4.787 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.608 4.782 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.147 4.349 dbcsr_matrix_vector_mult_local 304 10.0 3.651 4.122 3.653 4.124 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.601 3.987 mp_irecv_dv 1241 11.2 1.581 3.939 1.581 3.939 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.678 3.763 calculate_norms 792 9.8 3.529 3.696 3.529 3.696 make_images_sizes 222 9.7 0.000 0.000 0.994 3.610 mp_alltoall_i44 222 10.7 0.994 3.610 0.994 3.610 ls_scf_post 1 4.0 0.000 0.000 3.562 3.568 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.066 3.539 ls_scf_store_result 1 5.0 0.000 0.000 3.355 3.386 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.266 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.224 2.225 rebuild_ks_matrix 3 7.3 0.000 0.000 2.192 2.193 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.192 2.193 dbcsr_merge_all 275 8.9 0.474 0.522 2.051 2.098 dbcsr_data_release 10123 10.8 1.325 2.002 1.325 2.002 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.032000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3633.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e3281e024bb35180d3a2a853d323285ecfbece49_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 8.769176E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.084 0.103 108.556 108.566 qs_energies 1 2.0 0.000 0.000 107.130 107.146 ls_scf 1 3.0 0.000 0.000 104.175 104.190 dbcsr_multiply_generic 111 6.7 0.024 0.025 77.562 77.669 ls_scf_main 1 4.0 0.000 0.000 66.261 66.267 density_matrix_trs4 2 5.0 0.002 0.003 56.920 56.971 multiply_cannon 111 7.7 0.147 0.273 50.336 52.335 multiply_cannon_loop 111 8.7 0.068 0.071 46.918 47.722 ls_scf_init_scf 1 4.0 0.000 0.000 34.262 34.266 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.567 32.576 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.747 29.756 mp_waitall_1 4527 11.1 22.622 25.761 22.622 25.761 make_m2s 222 7.7 0.005 0.005 23.714 24.659 make_images 222 8.7 4.593 4.975 23.608 24.551 multiply_cannon_multrec 444 9.7 17.794 18.621 22.398 23.001 hybrid_alltoall_any 227 10.6 1.662 3.641 12.785 15.455 make_images_data 222 9.7 0.003 0.003 12.970 15.355 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.040 11.552 multiply_cannon_sync_h2d 444 9.7 8.873 9.061 8.873 9.061 arnoldi_extremal 4 6.8 0.000 0.000 7.382 7.392 arnoldi_normal_ev 4 7.8 0.005 0.012 7.382 7.392 build_subspace 16 8.4 0.026 0.036 6.834 6.849 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.431 5.560 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.248 5.336 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.061 5.321 dbcsr_matrix_vector_mult_local 304 10.0 4.928 5.201 4.931 5.203 dbcsr_mm_accdrv_process 1814 10.4 0.207 0.308 4.425 4.549 dbcsr_mm_accdrv_process_sort 1814 11.4 4.127 4.254 4.127 4.254 ls_scf_post 1 4.0 0.000 0.000 3.652 3.666 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.415 make_images_sizes 222 9.7 0.000 0.000 1.421 3.414 mp_alltoall_i44 222 10.7 1.421 3.414 1.421 3.414 mp_allgather_i34 111 8.7 1.072 3.377 1.072 3.377 calculate_norms 792 9.8 3.216 3.274 3.216 3.274 dbcsr_finalize 304 7.8 0.082 0.089 3.108 3.200 dbcsr_merge_all 275 8.9 0.892 0.923 2.875 2.948 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.925 2.925 dbcsr_complete_redistribute 5 7.6 1.433 1.471 2.780 2.879 dbcsr_data_release 12724 10.6 2.343 2.862 2.343 2.862 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.607 2.608 rebuild_ks_matrix 3 7.3 0.000 0.000 2.541 2.542 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.002 2.541 2.542 matrix_ls_to_qs 2 6.0 0.000 0.000 2.410 2.521 dbcsr_sort_data 325 11.1 2.443 2.497 2.443 2.497 dbcsr_new_transposed 4 7.5 0.243 0.252 2.269 2.287 dbcsr_add_d 103 6.2 0.000 0.000 2.159 2.238 dbcsr_add_anytype 103 7.2 0.860 0.892 2.158 2.238 dbcsr_frobenius_norm 74 6.6 2.056 2.133 2.189 2.225 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.566000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6843.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e3281e024bb35180d3a2a853d323285ecfbece49 Summary: empty Status: OK