=== 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: 685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/01 job id: 44733394 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/02 job id: 44733395 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/03 job id: 44733399 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/04 job id: 44733400 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/05 job id: 44733401 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/06 job id: 44733403 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/07 job id: 44733406 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/08 job id: 44733410 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/09 job id: 44733412 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/10 job id: 44733413 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/11 job id: 44733414 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/12 job id: 44733417 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/13 job id: 44733418 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/14 job id: 44733419 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/15 job id: 44733420 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/16 job id: 44733421 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/17 job id: 44733422 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/18 job id: 44733423 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/19 job id: 44733424 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/20 job id: 44733428 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/21 job id: 44733429 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/22 job id: 44733430 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/23 job id: 44733431 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/24 job id: 44733432 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/25 job id: 44733434 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/26 job id: 44733435 --- 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.036 133.765 133.767 farming_run 1 2.0 133.292 133.295 133.734 133.737 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463808E+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 229 1108280. 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 0.012 0.035 115.568 115.580 qs_energies 1 2.0 0.000 0.000 115.370 115.373 mp2_main 1 3.0 0.000 0.000 113.465 113.467 mp2_gpw_main 1 4.0 0.020 0.026 112.647 112.650 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 94.174 94.227 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.382 55.434 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.169 41.691 46.583 get_2c_integrals 1 6.0 0.000 0.002 37.920 38.620 integrate_v_rspace 273 8.0 0.438 0.454 25.054 29.623 pw_transfer 6555 10.6 0.374 0.394 27.475 28.070 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.108 26.517 grid_integrate_task_list 273 9.0 20.868 25.888 20.868 25.888 fft_wrap_pw1pw2_100 2178 12.4 1.231 1.456 23.628 24.056 compute_2c_integrals 1 7.0 0.002 0.003 19.797 19.797 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.004 19.527 mp2_eri_2c_integrate_gpw 1 9.0 2.359 2.420 19.001 19.526 cp_fm_cholesky_decompose 12 8.2 18.055 18.740 18.055 18.740 rpa_ri_compute_en 1 5.0 0.007 0.018 18.368 18.444 cholesky_decomp 1 7.0 0.000 0.000 16.965 17.647 fft3d_s 5443 13.4 16.160 16.354 16.182 16.376 ao_to_mo_and_store_B_mult_1 272 7.0 10.847 15.562 10.847 15.562 calculate_wavefunction 272 8.0 5.443 5.515 12.564 13.189 rpa_num_int 1 6.0 0.000 0.001 10.487 10.496 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.464 10.485 calc_mat_Q 8 8.0 0.000 0.000 9.324 9.423 contract_S_to_Q 8 9.0 0.000 0.000 8.746 8.842 calc_potential_gpw 544 9.5 0.005 0.006 8.284 8.672 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.292 8.605 parallel_gemm_fm 14 9.1 0.000 0.000 8.333 8.410 parallel_gemm_fm_cosma 14 10.1 8.333 8.410 8.333 8.410 potential_pw2rs 545 10.0 0.108 0.110 7.727 8.386 collocate_single_gaussian 272 10.0 0.039 0.042 7.530 7.827 create_integ_mat 1 6.0 0.014 0.026 7.715 7.724 array2fm 1 7.0 0.000 0.000 6.771 7.177 pw_scatter_s 2720 13.7 4.436 4.604 4.436 4.604 pw_gather_s 2722 13.2 3.871 4.250 3.871 4.250 array2fm_buffer_send 1 8.0 3.065 3.177 3.065 3.177 pw_poisson_solve 545 10.5 1.131 1.172 2.205 2.357 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.646409, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.038 396.134 396.136 farming_run 1 2.0 395.271 395.283 396.090 396.094 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.225724E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2390928. 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 0.014 0.040 209.869 209.870 qs_energies 1 2.0 0.000 0.000 209.595 209.611 scf_env_do_scf 1 3.0 0.000 0.000 106.297 106.298 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.422 105.431 rebuild_ks_matrix 4 6.0 0.000 0.000 105.421 105.429 qs_ks_build_kohn_sham_matrix 4 7.0 0.053 0.062 105.421 105.429 hfx_ks_matrix 4 8.0 0.001 0.001 105.039 105.044 integrate_four_center 4 9.0 0.144 0.460 105.038 105.043 mp2_main 1 3.0 0.001 0.009 103.003 103.020 mp2_gpw_main 1 4.0 0.035 0.059 102.148 102.167 integrate_four_center_main 4 10.0 0.080 0.434 96.919 99.673 integrate_four_center_bin 264 11.0 96.840 99.521 96.840 99.521 init_scf_loop 1 4.0 0.000 0.000 91.957 91.957 mp2_ri_gpw_compute_in 1 5.0 0.065 0.072 74.827 75.999 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.543 55.714 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.341 47.474 integrate_v_rspace 95 8.0 0.398 0.575 28.715 33.676 pw_transfer 2240 10.6 0.145 0.165 29.912 30.310 ao_to_mo_and_store_B_mult_1 91 7.0 10.514 29.831 10.514 29.831 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.939 29.338 grid_integrate_task_list 95 9.0 23.999 29.144 23.999 29.144 mp2_ri_gpw_compute_en 1 5.0 0.057 0.076 27.158 28.919 fft_wrap_pw1pw2_100 730 12.4 1.257 1.399 26.645 27.052 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.819 1.895 25.322 25.334 get_2c_integrals 1 6.0 0.000 0.000 20.195 20.221 compute_2c_integrals 1 7.0 0.004 0.019 19.175 19.184 compute_2c_integrals_loop_lm 1 8.0 0.002 0.008 18.817 19.030 mp2_eri_2c_integrate_gpw 1 9.0 1.748 1.868 18.815 19.030 fft3d_s 1823 13.4 18.426 18.744 18.440 18.757 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.339 14.339 calculate_wavefunction 91 8.0 2.015 2.054 9.740 10.026 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.585 8.860 9.538 potential_pw2rs 186 10.0 0.034 0.035 8.649 9.256 local_gemm 172 8.0 8.302 8.965 8.302 8.965 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.520 8.263 8.723 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.208 8.564 calc_potential_gpw 182 9.5 0.002 0.002 7.873 8.110 collocate_single_gaussian 91 10.0 0.017 0.028 7.865 8.080 mp_sendrecv_dm3 2068 8.0 6.304 6.760 6.304 6.760 mp2_ri_gpw_compute_en_ener 172 7.0 6.335 6.429 6.335 6.429 mp_sync 38 10.4 2.787 5.747 2.787 5.747 pw_gather_s 912 13.2 4.938 5.349 4.938 5.349 pw_scatter_s 910 13.7 3.922 4.300 3.922 4.300 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.145209, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1513.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 455.815168E+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 1466051. 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 0.017 0.041 48.628 48.628 qs_mol_dyn_low 1 2.0 0.003 0.007 48.351 48.360 qs_forces 11 3.9 0.003 0.016 48.272 48.278 qs_energies 11 4.9 0.011 0.077 46.760 46.805 scf_env_do_scf 11 5.9 0.001 0.022 40.806 40.807 scf_env_do_scf_inner_loop 108 6.5 0.003 0.015 38.771 38.772 dbcsr_multiply_generic 2286 12.5 0.094 0.099 30.444 30.922 qs_scf_new_mos 108 7.5 0.000 0.001 28.865 29.126 qs_scf_loop_do_ot 108 8.5 0.000 0.001 28.865 29.126 ot_scf_mini 108 9.5 0.002 0.003 27.289 27.466 multiply_cannon 2286 13.5 0.187 0.192 24.683 26.664 multiply_cannon_loop 2286 14.5 1.448 1.543 24.017 25.993 velocity_verlet 10 3.0 0.002 0.003 23.809 23.810 ot_mini 108 10.5 0.001 0.001 16.972 17.222 qs_ot_get_derivative 108 11.5 0.001 0.001 14.144 14.309 mp_waitall_1 245248 16.5 7.092 13.527 7.092 13.527 multiply_cannon_metrocomm3 54864 15.5 0.071 0.075 5.207 12.499 multiply_cannon_multrec 54864 15.5 4.448 6.846 7.567 11.025 rebuild_ks_matrix 119 8.3 0.000 0.000 7.820 7.948 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.025 7.819 7.947 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.906 7.024 multiply_cannon_sync_h2d 54864 15.5 6.139 6.785 6.139 6.785 qs_ot_get_p 119 10.4 0.001 0.001 6.009 6.249 mp_sum_l 7207 12.9 3.914 5.446 3.914 5.446 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.865 5.298 init_scf_run 11 5.9 0.000 0.001 4.668 4.669 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.668 4.668 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.572 4.656 sum_up_and_integrate 119 10.3 0.013 0.017 4.441 4.448 integrate_v_rspace 119 11.3 0.002 0.003 4.428 4.437 dbcsr_mm_accdrv_process 76910 16.1 1.050 1.742 3.041 4.327 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.046 4.169 calculate_rho_elec 119 8.7 0.011 0.017 4.045 4.169 qs_ot_p2m_diag 50 11.0 0.004 0.009 3.556 3.599 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.848 3.054 apply_single 119 13.6 0.000 0.000 2.848 3.054 calculate_dm_sparse 119 9.5 0.000 0.001 2.782 2.913 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.846 2.846 rs_pw_transfer 974 11.9 0.012 0.013 2.655 2.740 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.646 2.646 cp_fm_redistribute_end 50 14.0 2.409 2.626 2.415 2.627 ot_diis_step 108 11.5 0.006 0.008 2.608 2.609 calculate_first_density_matrix 1 7.0 0.000 0.002 2.549 2.553 cp_fm_diag_elpa_base 50 14.0 0.210 2.542 0.211 2.551 jit_kernel_multiply 13 15.8 1.934 2.508 1.934 2.508 multiply_cannon_metrocomm1 54864 15.5 0.054 0.058 1.283 2.417 density_rs2pw 119 9.7 0.004 0.004 2.159 2.289 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.111 2.176 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.150 2.152 grid_integrate_task_list 119 12.3 2.016 2.136 2.016 2.136 acc_transpose_blocks 54864 15.5 0.211 0.238 1.624 2.095 wfi_extrapolate 11 7.9 0.005 0.034 2.058 2.058 init_scf_loop 11 6.9 0.004 0.031 2.019 2.020 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.919 1.965 mp_sum_d 4127 12.0 1.241 1.847 1.241 1.847 potential_pw2rs 119 12.3 0.004 0.004 1.811 1.822 pw_transfer 1439 11.6 0.053 0.059 1.742 1.816 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.666 1.742 make_m2s 4572 13.5 0.054 0.057 1.539 1.593 fft3d_ps 1201 14.6 0.368 0.468 1.435 1.510 make_images 4572 14.5 0.136 0.142 1.455 1.508 mp_waitany 12084 13.8 1.238 1.380 1.238 1.380 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.361 1.378 grid_collocate_task_list 119 9.7 1.291 1.375 1.291 1.375 fft_wrap_pw1pw2_140 487 13.2 0.083 0.098 1.284 1.361 mp_alltoall_d11v 2130 13.8 1.139 1.352 1.139 1.352 parallel_gemm_fm 81 9.0 0.000 0.000 1.033 1.038 parallel_gemm_fm_cosma 81 10.0 1.033 1.038 1.033 1.038 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.647 1.032 acc_transpose_blocks_kernels 54864 16.5 0.234 0.375 0.732 0.993 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.970 0.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=48.628000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=434.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.727104E+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 189762. 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 0.024 0.078 38.335 38.336 qs_mol_dyn_low 1 2.0 0.003 0.005 37.954 37.961 qs_forces 11 3.9 0.002 0.005 37.881 37.884 qs_energies 11 4.9 0.001 0.001 36.205 36.209 scf_env_do_scf 11 5.9 0.000 0.001 31.110 31.110 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.656 28.657 dbcsr_multiply_generic 2286 12.5 0.101 0.103 20.985 21.347 qs_scf_new_mos 108 7.5 0.001 0.001 19.556 19.810 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.555 19.809 ot_scf_mini 108 9.5 0.002 0.003 18.684 18.864 velocity_verlet 10 3.0 0.001 0.001 17.976 17.977 multiply_cannon 2286 13.5 0.207 0.218 16.174 17.781 multiply_cannon_loop 2286 14.5 0.906 0.983 15.107 16.456 ot_mini 108 10.5 0.001 0.001 11.567 11.806 mp_waitall_1 200699 16.5 5.495 10.766 5.495 10.766 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.092 9.535 qs_ot_get_derivative 108 11.5 0.001 0.001 9.135 9.317 multiply_cannon_multrec 27432 15.5 1.967 4.408 5.848 8.800 rebuild_ks_matrix 119 8.3 0.000 0.000 7.155 7.300 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 7.154 7.300 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.341 6.476 dbcsr_mm_accdrv_process 47894 16.0 2.974 4.991 3.812 5.642 qs_ot_get_p 119 10.4 0.001 0.001 4.268 4.499 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.524 4.378 sum_up_and_integrate 119 10.3 0.024 0.027 4.147 4.153 integrate_v_rspace 119 11.3 0.002 0.004 4.123 4.132 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.034 4.107 apply_single 119 13.6 0.000 0.000 3.034 4.107 mp_sum_l 7207 12.9 2.021 3.991 2.021 3.991 init_scf_run 11 5.9 0.000 0.001 3.789 3.789 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.788 3.789 qs_rho_update_rho_low 119 7.7 0.001 0.002 3.750 3.782 calculate_rho_elec 119 8.7 0.021 0.024 3.750 3.782 rs_pw_transfer 974 11.9 0.010 0.011 2.574 3.121 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.879 2.897 multiply_cannon_sync_h2d 27432 15.5 2.196 2.839 2.196 2.839 density_rs2pw 119 9.7 0.004 0.004 2.080 2.633 make_m2s 4572 13.5 0.052 0.054 2.375 2.596 make_images 4572 14.5 0.202 0.239 2.286 2.505 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.452 2.453 init_scf_loop 11 6.9 0.000 0.000 2.433 2.434 ot_diis_step 108 11.5 0.011 0.011 2.384 2.385 calculate_first_density_matrix 1 7.0 0.000 0.000 2.297 2.299 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.085 2.179 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.133 2.134 jit_kernel_multiply 11 16.1 0.786 2.116 0.786 2.116 cp_fm_redistribute_end 50 14.0 1.767 2.110 1.770 2.111 calculate_dm_sparse 119 9.5 0.000 0.001 2.004 2.080 cp_fm_diag_elpa_base 50 14.0 0.327 2.003 0.340 2.059 pw_transfer 1439 11.6 0.065 0.071 1.954 1.984 grid_integrate_task_list 119 12.3 1.846 1.942 1.846 1.942 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.862 1.895 potential_pw2rs 119 12.3 0.006 0.006 1.866 1.877 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.861 1.863 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.713 fft3d_ps 1201 14.6 0.510 0.563 1.563 1.591 prepare_preconditioner 11 7.9 0.000 0.000 1.510 1.537 make_preconditioner 11 8.9 0.000 0.000 1.510 1.537 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.485 1.518 make_images_data 4572 15.5 0.045 0.051 1.104 1.488 acc_transpose_blocks 27432 15.5 0.109 0.115 1.191 1.482 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.412 1.468 hybrid_alltoall_any 4725 16.4 0.050 0.111 0.959 1.442 wfi_extrapolate 11 7.9 0.001 0.001 1.441 1.442 grid_collocate_task_list 119 9.7 1.228 1.375 1.228 1.375 mp_alltoall_d11v 2130 13.8 1.147 1.301 1.147 1.301 mp_allgather_i34 2286 14.5 0.512 1.259 0.512 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.246 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.147 1.197 rs_pw_transfer_RS2PW_140 130 11.5 0.144 0.158 0.621 1.169 mp_waitany 5720 13.7 0.596 1.148 0.596 1.148 qs_energies_init_hamiltonians 11 5.9 0.008 0.016 1.049 1.052 mp_sum_d 4127 12.0 0.588 1.042 0.588 1.042 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.889 0.902 acc_transpose_blocks_kernels 27432 16.5 0.184 0.272 0.661 0.872 mp_alltoall_z22v 1201 16.6 0.673 0.776 0.673 0.776 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.336000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.090909, yerr=1.781447 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 522.579968E+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 835869. 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 0.012 0.029 31.861 31.862 qs_mol_dyn_low 1 2.0 0.003 0.003 31.650 31.659 qs_forces 11 3.9 0.002 0.002 31.354 31.355 qs_energies 11 4.9 0.001 0.002 29.775 29.778 scf_env_do_scf 11 5.9 0.000 0.001 25.210 25.210 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 22.707 22.707 dbcsr_multiply_generic 2286 12.5 0.094 0.096 16.098 16.173 velocity_verlet 10 3.0 0.001 0.002 15.304 15.312 qs_scf_new_mos 108 7.5 0.001 0.001 14.526 14.540 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.525 14.539 ot_scf_mini 108 9.5 0.002 0.003 13.810 13.823 multiply_cannon 2286 13.5 0.196 0.205 12.969 13.716 multiply_cannon_loop 2286 14.5 0.638 0.670 12.203 12.969 ot_mini 108 10.5 0.001 0.001 8.554 8.568 qs_ot_get_derivative 108 11.5 0.001 0.001 7.079 7.092 multiply_cannon_multrec 18288 15.5 1.940 2.816 6.733 6.990 rebuild_ks_matrix 119 8.3 0.000 0.000 6.376 6.391 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.376 6.391 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.631 5.644 dbcsr_mm_accdrv_process 38222 16.0 4.040 5.281 4.709 5.588 sum_up_and_integrate 119 10.3 0.031 0.032 3.953 3.959 integrate_v_rspace 119 11.3 0.002 0.003 3.922 3.932 mp_waitall_1 158411 16.6 2.521 3.672 2.521 3.672 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.438 3.443 calculate_rho_elec 119 8.7 0.030 0.031 3.438 3.442 init_scf_run 11 5.9 0.000 0.001 3.417 3.417 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.417 3.417 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.652 3.262 qs_ot_get_p 119 10.4 0.001 0.001 3.117 3.134 init_scf_loop 11 6.9 0.000 0.000 2.487 2.488 rs_pw_transfer 974 11.9 0.009 0.010 2.202 2.451 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 1.404 2.404 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.014 2.336 apply_single 119 13.6 0.000 0.000 2.013 2.336 density_rs2pw 119 9.7 0.004 0.004 1.961 2.220 calculate_first_density_matrix 1 7.0 0.000 0.000 2.219 2.220 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.100 2.105 pw_transfer 1439 11.6 0.065 0.071 1.955 1.962 grid_integrate_task_list 119 12.3 1.803 1.900 1.803 1.900 make_m2s 4572 13.5 0.044 0.045 1.719 1.873 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.861 1.871 jit_kernel_multiply 10 16.1 0.618 1.859 0.618 1.859 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.841 1.841 calculate_dm_sparse 119 9.5 0.000 0.000 1.797 1.809 make_images 4572 14.5 0.192 0.205 1.633 1.785 potential_pw2rs 119 12.3 0.007 0.008 1.712 1.721 prepare_preconditioner 11 7.9 0.000 0.000 1.697 1.699 make_preconditioner 11 8.9 0.000 0.000 1.697 1.699 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.664 1.665 multiply_cannon_sync_h2d 18288 15.5 1.410 1.659 1.410 1.659 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.556 1.635 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.627 1.628 cp_fm_redistribute_end 50 14.0 1.215 1.609 1.215 1.609 cp_fm_diag_elpa_base 50 14.0 0.378 1.527 0.392 1.578 fft3d_ps 1201 14.6 0.523 0.541 1.540 1.550 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.523 1.527 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.517 1.524 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.510 1.521 ot_diis_step 108 11.5 0.011 0.011 1.457 1.457 mp_sum_l 7207 12.9 1.022 1.385 1.022 1.385 grid_collocate_task_list 119 9.7 1.213 1.364 1.213 1.364 acc_transpose_blocks 18288 15.5 0.075 0.077 1.233 1.259 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.172 1.179 wfi_extrapolate 11 7.9 0.001 0.001 1.152 1.152 make_images_data 4572 15.5 0.045 0.048 0.761 0.936 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.928 0.929 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.859 0.881 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.657 0.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.815 0.816 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.798 0.815 mp_waitany 9880 13.7 0.493 0.763 0.493 0.763 mp_alltoall_d11v 2130 13.8 0.643 0.754 0.643 0.754 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.658 0.747 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.489 0.741 mp_alltoall_z22v 1201 16.6 0.627 0.733 0.627 0.733 cp_fm_cholesky_invert 11 10.9 0.702 0.705 0.702 0.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.862000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=2.425757 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 553.693184E+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 0.015 0.033 35.497 35.498 qs_mol_dyn_low 1 2.0 0.003 0.004 35.228 35.236 qs_forces 11 3.9 0.026 0.056 34.719 34.720 qs_energies 11 4.9 0.002 0.006 32.982 33.012 scf_env_do_scf 11 5.9 0.001 0.001 27.605 27.606 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.268 24.268 dbcsr_multiply_generic 2286 12.5 0.100 0.103 18.225 18.339 velocity_verlet 10 3.0 0.002 0.002 17.982 18.002 qs_scf_new_mos 108 7.5 0.001 0.001 15.814 15.866 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.814 15.865 multiply_cannon 2286 13.5 0.228 0.261 14.726 15.210 ot_scf_mini 108 9.5 0.002 0.003 14.888 14.939 multiply_cannon_loop 2286 14.5 0.940 0.973 13.815 14.154 multiply_cannon_multrec 27432 15.5 2.341 3.022 8.901 9.367 ot_mini 108 10.5 0.001 0.001 9.093 9.155 dbcsr_mm_accdrv_process 47916 15.9 5.592 7.167 6.467 7.753 qs_ot_get_derivative 108 11.5 0.001 0.001 7.303 7.353 rebuild_ks_matrix 119 8.3 0.000 0.000 6.585 6.641 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 6.585 6.641 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.851 5.900 init_scf_run 11 5.9 0.000 0.001 3.986 3.986 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.986 3.986 sum_up_and_integrate 119 10.3 0.036 0.039 3.795 3.803 integrate_v_rspace 119 11.3 0.002 0.003 3.759 3.768 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.531 3.561 calculate_rho_elec 119 8.7 0.040 0.046 3.531 3.560 init_scf_loop 11 6.9 0.001 0.004 3.318 3.319 qs_ot_get_p 119 10.4 0.001 0.001 3.206 3.277 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.725 3.174 calculate_first_density_matrix 1 7.0 0.000 0.002 2.637 2.638 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.075 2.522 apply_single 119 13.6 0.000 0.000 2.075 2.521 prepare_preconditioner 11 7.9 0.000 0.001 2.466 2.473 make_preconditioner 11 8.9 0.000 0.002 2.466 2.473 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.081 2.402 mp_waitall_1 137007 16.6 1.706 2.359 1.706 2.359 make_m2s 4572 13.5 0.054 0.056 2.113 2.230 rs_pw_transfer 974 11.9 0.009 0.010 1.984 2.198 calculate_dm_sparse 119 9.5 0.000 0.000 2.116 2.162 density_rs2pw 119 9.7 0.004 0.004 1.956 2.136 pw_transfer 1439 11.6 0.065 0.071 2.094 2.130 make_images 4572 14.5 0.273 0.333 2.004 2.120 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.090 2.091 jit_kernel_multiply 10 16.0 0.814 2.089 0.814 2.089 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.001 2.043 qs_ot_p2m_diag 50 11.0 0.015 0.023 1.982 1.991 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.902 1.914 grid_integrate_task_list 119 12.3 1.815 1.898 1.815 1.898 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.865 1.891 ot_diis_step 108 11.5 0.012 0.012 1.750 1.751 fft3d_ps 1201 14.6 0.559 0.612 1.676 1.713 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.668 1.712 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.675 1.675 potential_pw2rs 119 12.3 0.008 0.010 1.552 1.555 acc_transpose_blocks 27432 15.5 0.111 0.114 1.526 1.549 mp_sum_l 7207 12.9 0.963 1.518 0.963 1.518 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.455 1.456 cp_fm_redistribute_end 50 14.0 0.965 1.435 0.966 1.435 cp_fm_diag_elpa_base 50 14.0 0.448 1.373 0.467 1.413 grid_collocate_task_list 119 9.7 1.220 1.356 1.220 1.356 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.307 1.315 wfi_extrapolate 11 7.9 0.001 0.001 1.300 1.300 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.728 1.242 qs_energies_init_hamiltonians 11 5.9 0.003 0.016 1.128 1.153 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.082 1.102 multiply_cannon_sync_h2d 27432 15.5 1.002 1.099 1.002 1.099 cp_fm_upper_to_full 72 13.5 0.792 1.096 0.792 1.096 dbcsr_complete_redistribute 329 12.2 0.134 0.174 0.761 1.033 make_images_data 4572 15.5 0.045 0.050 0.824 0.944 acc_transpose_blocks_kernels 27432 16.5 0.271 0.277 0.907 0.921 mp_alltoall_d11v 2130 13.8 0.700 0.899 0.700 0.899 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.703 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.572 0.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.824 0.830 mp_alltoall_z22v 1201 16.6 0.703 0.742 0.703 0.742 rs_pw_transfer_RS2PW_140 130 11.5 0.108 0.112 0.516 0.725 cp_fm_cholesky_invert 11 10.9 0.715 0.718 0.715 0.718 mp_waitany 7280 13.7 0.506 0.714 0.506 0.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.498000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.727273, yerr=2.733327 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 599.638016E+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 0.016 0.029 27.829 27.830 qs_mol_dyn_low 1 2.0 0.003 0.003 27.647 27.654 qs_forces 11 3.9 0.002 0.002 27.526 27.529 qs_energies 11 4.9 0.001 0.001 25.810 25.814 scf_env_do_scf 11 5.9 0.000 0.001 21.227 21.227 scf_env_do_scf_inner_loop 108 6.5 0.004 0.008 18.747 18.748 velocity_verlet 10 3.0 0.002 0.002 14.286 14.291 dbcsr_multiply_generic 2286 12.5 0.092 0.095 11.984 12.064 qs_scf_new_mos 108 7.5 0.001 0.001 10.749 10.772 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.749 10.771 ot_scf_mini 108 9.5 0.002 0.002 10.084 10.111 multiply_cannon 2286 13.5 0.228 0.238 9.568 10.025 multiply_cannon_loop 2286 14.5 0.331 0.342 8.673 8.848 rebuild_ks_matrix 119 8.3 0.000 0.000 6.019 6.041 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.019 6.041 multiply_cannon_multrec 9144 15.5 1.600 1.818 5.745 5.932 ot_mini 108 10.5 0.001 0.001 5.693 5.724 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.367 5.387 qs_ot_get_derivative 108 11.5 0.001 0.001 4.408 4.435 dbcsr_mm_accdrv_process 12550 15.8 3.124 4.049 4.044 4.120 sum_up_and_integrate 119 10.3 0.038 0.041 3.664 3.670 integrate_v_rspace 119 11.3 0.003 0.003 3.625 3.632 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.536 3.542 calculate_rho_elec 119 8.7 0.059 0.061 3.535 3.542 init_scf_run 11 5.9 0.000 0.001 3.163 3.163 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.163 3.163 qs_ot_get_p 119 10.4 0.001 0.001 2.646 2.684 init_scf_loop 11 6.9 0.000 0.000 2.460 2.461 pw_transfer 1439 11.6 0.066 0.069 2.145 2.156 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.051 2.064 calculate_first_density_matrix 1 7.0 0.000 0.000 2.038 2.039 density_rs2pw 119 9.7 0.004 0.004 1.871 2.010 mp_waitall_1 115863 16.7 1.470 2.007 1.470 2.007 grid_integrate_task_list 119 12.3 1.840 1.917 1.840 1.917 make_m2s 4572 13.5 0.035 0.036 1.677 1.837 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.764 1.766 rs_pw_transfer 974 11.9 0.008 0.008 1.636 1.757 make_images 4572 14.5 0.270 0.305 1.587 1.746 fft3d_ps 1201 14.6 0.562 0.573 1.711 1.722 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.706 1.717 calculate_dm_sparse 119 9.5 0.000 0.000 1.683 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.684 1.688 make_preconditioner 11 8.9 0.000 0.000 1.684 1.688 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.574 1.601 jit_kernel_multiply 9 15.5 0.883 1.582 0.883 1.582 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.569 1.569 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.496 1.497 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.470 1.483 potential_pw2rs 119 12.3 0.010 0.010 1.402 1.405 grid_collocate_task_list 119 9.7 1.272 1.386 1.272 1.386 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.346 1.356 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.309 1.309 cp_fm_redistribute_end 50 14.0 0.653 1.290 0.654 1.291 cp_fm_diag_elpa_base 50 14.0 0.592 1.222 0.635 1.279 ot_diis_step 108 11.5 0.013 0.013 1.273 1.273 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.214 1.217 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.151 1.173 apply_single 119 13.6 0.000 0.000 1.151 1.172 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.103 1.117 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.109 1.115 wfi_extrapolate 11 7.9 0.001 0.001 1.077 1.077 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.735 0.983 make_images_data 4572 15.5 0.039 0.042 0.779 0.971 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.923 acc_transpose_blocks 9144 15.5 0.038 0.038 0.878 0.892 cp_fm_cholesky_invert 11 10.9 0.841 0.844 0.841 0.844 mp_alltoall_d11v 2130 13.8 0.733 0.830 0.733 0.830 multiply_cannon_sync_h2d 9144 15.5 0.711 0.796 0.711 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.776 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.669 0.717 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.704 0.712 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.422 0.654 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 0.641 0.653 mp_alltoall_z22v 1201 16.6 0.597 0.650 0.597 0.650 mp_allgather_i34 2286 14.5 0.229 0.628 0.229 0.628 yz_to_x 606 15.1 0.266 0.277 0.576 0.592 multiply_cannon_metrocomm1 9144 15.5 0.021 0.022 0.299 0.584 x_to_yz 595 16.2 0.276 0.287 0.562 0.569 mp_waitany 5200 13.7 0.452 0.568 0.452 0.568 qs_create_task_list 11 7.9 0.000 0.001 0.541 0.565 generate_qs_task_list 11 8.9 0.189 0.211 0.541 0.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.830000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=568.818182, yerr=3.712832 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 737.538048E+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 0.019 0.044 41.316 41.317 qs_mol_dyn_low 1 2.0 0.003 0.003 41.093 41.101 qs_forces 11 3.9 0.002 0.002 40.971 40.972 qs_energies 11 4.9 0.001 0.002 38.971 38.974 scf_env_do_scf 11 5.9 0.001 0.001 33.315 33.315 scf_env_do_scf_inner_loop 108 6.5 0.004 0.007 25.444 25.446 velocity_verlet 10 3.0 0.002 0.002 23.418 23.429 dbcsr_multiply_generic 2286 12.5 0.101 0.101 17.205 17.381 qs_scf_new_mos 108 7.5 0.001 0.001 15.597 15.691 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.596 15.690 ot_scf_mini 108 9.5 0.002 0.002 14.527 14.626 multiply_cannon 2286 13.5 0.304 0.316 13.447 14.345 multiply_cannon_loop 2286 14.5 0.346 0.353 12.222 13.154 ot_mini 108 10.5 0.001 0.001 8.674 8.789 multiply_cannon_multrec 9144 15.5 3.367 4.648 8.517 8.618 init_scf_loop 11 6.9 0.000 0.000 7.839 7.842 rebuild_ks_matrix 119 8.3 0.000 0.000 7.357 7.495 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.356 7.495 prepare_preconditioner 11 7.9 0.000 0.000 6.868 6.882 make_preconditioner 11 8.9 0.000 0.000 6.868 6.882 qs_ot_get_derivative 108 11.5 0.001 0.001 6.681 6.778 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.645 6.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.396 6.759 dbcsr_mm_accdrv_process 12550 15.8 4.041 5.287 5.028 6.348 cp_fm_upper_to_full 72 14.2 3.244 4.659 3.244 4.659 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.275 4.282 calculate_rho_elec 119 8.7 0.118 0.121 4.274 4.281 sum_up_and_integrate 119 10.3 0.066 0.067 4.058 4.065 integrate_v_rspace 119 11.3 0.003 0.003 3.992 3.999 init_scf_run 11 5.9 0.000 0.001 3.625 3.625 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.625 3.625 qs_ot_get_p 119 10.4 0.001 0.001 3.143 3.277 mp_waitall_1 94719 16.7 2.248 3.186 2.248 3.186 dbcsr_complete_redistribute 329 12.2 0.285 0.292 1.991 2.865 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.440 2.854 pw_transfer 1439 11.6 0.068 0.069 2.831 2.835 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.732 2.736 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.692 2.551 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.153 2.440 apply_single 119 13.6 0.000 0.000 2.153 2.440 mp_alltoall_i22 627 13.8 1.509 2.395 1.509 2.395 fft3d_ps 1201 14.6 0.597 0.606 2.356 2.361 make_m2s 4572 13.5 0.038 0.039 2.215 2.360 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.467 2.318 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.300 2.305 density_rs2pw 119 9.7 0.004 0.004 2.221 2.241 make_images 4572 14.5 0.358 0.388 2.092 2.237 calculate_first_density_matrix 1 7.0 0.000 0.000 2.187 2.189 calculate_dm_sparse 119 9.5 0.000 0.000 2.134 2.154 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.273 2.121 grid_integrate_task_list 119 12.3 2.047 2.066 2.047 2.066 ot_diis_step 108 11.5 0.014 0.015 1.969 1.969 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.910 1.911 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.907 1.908 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.763 1.764 mp_sum_l 7207 12.9 1.004 1.740 1.004 1.740 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.669 1.719 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.639 1.639 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.594 1.622 grid_collocate_task_list 119 9.7 1.479 1.493 1.479 1.493 rs_pw_transfer 974 11.9 0.009 0.010 1.465 1.483 potential_pw2rs 119 12.3 0.014 0.015 1.465 1.467 jit_kernel_multiply 8 15.7 0.960 1.414 0.960 1.414 cp_fm_cholesky_invert 11 10.9 1.409 1.412 1.409 1.412 wfi_extrapolate 11 7.9 0.001 0.001 1.374 1.374 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.352 1.352 cp_fm_diag_elpa_base 50 14.0 1.206 1.260 1.350 1.350 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.062 1.267 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.211 1.220 make_images_data 4572 15.5 0.043 0.046 1.000 1.187 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.139 1.157 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.094 1.110 mp_alltoall_d11v 2130 13.8 1.040 1.062 1.040 1.062 multiply_cannon_sync_h2d 9144 15.5 1.044 1.047 1.044 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.992 1.026 yz_to_x 606 15.1 0.459 0.471 0.944 0.955 qs_create_task_list 11 7.9 0.000 0.000 0.935 0.948 generate_qs_task_list 11 8.9 0.370 0.390 0.935 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.923 0.936 acc_transpose_blocks 9144 15.5 0.038 0.039 0.888 0.897 mp_alltoall_z22v 1201 16.6 0.810 0.855 0.810 0.855 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.317000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=694.454545, yerr=13.103529 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 503.439360E+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 0.011 0.027 83.136 83.137 qs_mol_dyn_low 1 2.0 0.003 0.003 82.901 82.910 qs_forces 11 3.9 0.002 0.003 82.807 82.808 qs_energies 11 4.9 0.001 0.002 79.924 79.938 scf_env_do_scf 11 5.9 0.000 0.001 71.188 71.191 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.606 65.606 dbcsr_multiply_generic 2055 12.4 0.107 0.111 51.894 52.164 qs_scf_new_mos 99 7.5 0.001 0.001 48.003 48.128 qs_scf_loop_do_ot 99 8.5 0.001 0.001 48.002 48.127 ot_scf_mini 99 9.5 0.002 0.002 45.570 45.661 multiply_cannon 2055 13.4 0.179 0.185 42.407 43.265 multiply_cannon_loop 2055 14.4 1.556 1.597 41.473 42.279 velocity_verlet 10 3.0 0.001 0.002 42.187 42.188 ot_mini 99 10.5 0.001 0.001 27.509 27.615 qs_ot_get_derivative 99 11.5 0.001 0.001 20.672 20.770 multiply_cannon_multrec 49320 15.4 12.147 12.845 17.170 18.040 rebuild_ks_matrix 110 8.3 0.000 0.001 14.697 14.822 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.696 14.822 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.878 12.989 mp_waitall_1 220248 16.4 11.373 12.155 11.373 12.155 multiply_cannon_sync_h2d 49320 15.4 9.956 10.511 9.956 10.511 qs_ot_get_p 110 10.4 0.001 0.001 9.529 9.659 multiply_cannon_metrocomm3 49320 15.4 0.079 0.083 6.889 7.869 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.159 7.709 apply_single 110 13.6 0.000 0.001 7.159 7.708 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.007 7.527 sum_up_and_integrate 110 10.3 0.037 0.043 7.167 7.181 integrate_v_rspace 110 11.3 0.003 0.003 7.130 7.151 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.738 6.890 calculate_rho_elec 110 8.6 0.021 0.026 6.737 6.889 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.670 6.732 init_scf_run 11 5.9 0.000 0.001 6.713 6.713 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.713 6.713 ot_diis_step 99 11.5 0.006 0.006 6.538 6.538 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.339 6.385 mp_sum_l 6514 12.8 5.095 5.871 5.095 5.871 init_scf_loop 11 6.9 0.000 0.000 5.553 5.554 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.341 5.341 dbcsr_mm_accdrv_process 87628 16.1 2.078 2.182 4.901 5.212 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.869 4.871 cp_fm_redistribute_end 48 14.0 4.252 4.848 4.255 4.848 cp_fm_diag_elpa_base 48 14.0 0.586 4.715 0.590 4.737 rs_pw_transfer 902 11.9 0.012 0.013 3.735 4.283 make_m2s 4110 13.4 0.061 0.065 3.929 4.035 density_rs2pw 110 9.6 0.004 0.005 3.537 4.032 wfi_extrapolate 11 7.9 0.001 0.001 4.013 4.013 make_images 4110 14.4 0.180 0.195 3.833 3.944 calculate_dm_sparse 110 9.5 0.000 0.001 3.794 3.882 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.628 3.686 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.540 3.544 pw_transfer 1331 11.6 0.054 0.067 3.338 3.403 prepare_preconditioner 11 7.9 0.000 0.000 3.379 3.399 make_preconditioner 11 8.9 0.000 0.000 3.379 3.399 grid_integrate_task_list 110 12.3 3.240 3.397 3.240 3.397 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.284 3.348 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.250 3.318 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.211 3.254 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.134 3.178 fft3d_ps 1111 14.6 0.778 0.879 2.777 2.838 fft_wrap_pw1pw2_140 451 13.1 0.172 0.193 2.749 2.820 potential_pw2rs 110 12.3 0.006 0.007 2.683 2.701 calculate_first_density_matrix 1 7.0 0.000 0.000 2.608 2.616 jit_kernel_multiply 13 15.9 2.540 2.560 2.540 2.560 mp_alltoall_d11v 2046 13.8 2.122 2.441 2.122 2.441 mp_waitany 14300 13.8 1.812 2.378 1.812 2.378 acc_transpose_blocks 49320 15.4 0.225 0.237 2.196 2.277 grid_collocate_task_list 110 9.6 2.084 2.271 2.084 2.271 mp_sum_d 3881 11.9 1.458 2.050 1.458 2.050 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.959 1.976 make_images_data 4110 15.4 0.042 0.045 1.771 1.898 hybrid_alltoall_any 4261 16.3 0.081 0.478 1.529 1.795 cp_fm_cholesky_invert 11 10.9 1.789 1.793 1.789 1.793 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.137000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.454545, yerr=3.367320 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.234752E+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 3284021. 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 0.012 0.032 68.801 68.802 qs_mol_dyn_low 1 2.0 0.003 0.003 68.592 68.601 qs_forces 11 3.9 0.002 0.002 68.507 68.509 qs_energies 11 4.9 0.001 0.002 65.160 65.164 scf_env_do_scf 11 5.9 0.000 0.001 56.662 56.666 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 49.043 49.044 dbcsr_multiply_generic 2055 12.4 0.114 0.118 37.618 37.784 velocity_verlet 10 3.0 0.001 0.002 36.335 36.337 qs_scf_new_mos 99 7.5 0.001 0.001 32.533 32.673 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.532 32.673 multiply_cannon 2055 13.4 0.221 0.245 31.004 32.215 ot_scf_mini 99 9.5 0.003 0.003 30.875 31.022 multiply_cannon_loop 2055 14.4 0.933 0.959 29.696 30.790 ot_mini 99 10.5 0.001 0.001 18.125 18.274 multiply_cannon_multrec 24660 15.4 7.619 9.978 13.850 16.130 rebuild_ks_matrix 110 8.3 0.000 0.001 13.930 14.061 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.930 14.060 qs_ot_get_derivative 99 11.5 0.001 0.001 12.350 12.501 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.271 12.378 mp_waitall_1 176588 16.5 7.582 10.511 7.582 10.511 multiply_cannon_sync_h2d 24660 15.4 7.049 8.063 7.049 8.063 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.117 7.969 init_scf_loop 11 6.9 0.000 0.000 7.583 7.583 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.507 7.256 apply_single 110 13.6 0.000 0.001 6.507 7.255 sum_up_and_integrate 110 10.3 0.052 0.059 6.708 6.721 integrate_v_rspace 110 11.3 0.002 0.003 6.656 6.670 dbcsr_mm_accdrv_process 52282 16.1 4.554 5.304 6.069 6.437 qs_ot_get_p 110 10.4 0.001 0.001 6.117 6.292 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.253 6.265 calculate_rho_elec 110 8.6 0.039 0.047 6.253 6.265 init_scf_run 11 5.9 0.000 0.001 6.145 6.145 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.144 6.145 ot_diis_step 99 11.5 0.010 0.010 5.730 5.730 prepare_preconditioner 11 7.9 0.000 0.000 5.528 5.560 make_preconditioner 11 8.9 0.000 0.000 5.528 5.560 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.658 5.408 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.121 5.288 make_m2s 4110 13.4 0.057 0.060 4.188 4.685 make_images 4110 14.4 0.399 0.444 4.079 4.572 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.217 4.237 pw_transfer 1331 11.6 0.066 0.072 3.791 3.937 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.684 3.832 density_rs2pw 110 9.6 0.004 0.005 3.367 3.807 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.795 3.795 wfi_extrapolate 11 7.9 0.001 0.001 3.542 3.542 rs_pw_transfer 902 11.9 0.012 0.014 2.995 3.466 grid_integrate_task_list 110 12.3 3.143 3.318 3.143 3.318 fft_wrap_pw1pw2_140 451 13.1 0.203 0.221 3.143 3.295 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.287 3.289 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.265 3.268 cp_fm_redistribute_end 48 14.0 2.443 3.241 2.445 3.242 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.161 3.227 cp_fm_diag_elpa_base 48 14.0 0.762 3.099 0.793 3.184 fft3d_ps 1111 14.6 1.103 1.327 3.032 3.168 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.946 3.005 calculate_dm_sparse 110 9.5 0.001 0.001 2.958 2.984 make_images_data 4110 15.4 0.046 0.050 2.347 2.828 hybrid_alltoall_any 4261 16.3 0.101 0.442 2.068 2.813 cp_fm_cholesky_invert 11 10.9 2.560 2.567 2.560 2.567 calculate_first_density_matrix 1 7.0 0.000 0.000 2.513 2.516 potential_pw2rs 110 12.3 0.008 0.009 2.489 2.507 grid_collocate_task_list 110 9.6 2.052 2.479 2.052 2.479 mp_sum_l 6514 12.8 1.798 2.471 1.798 2.471 jit_kernel_multiply 11 16.3 1.166 2.158 1.166 2.158 mp_alltoall_d11v 2046 13.8 1.775 2.137 1.775 2.137 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.939 1.962 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.846 1.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.822 1.842 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.783 1.719 mp_waitany 10164 13.8 1.203 1.690 1.203 1.690 mp_allgather_i34 2055 14.4 0.640 1.667 0.640 1.667 mp_irecv_dv 57340 16.2 0.659 1.601 0.659 1.601 acc_transpose_blocks 24660 15.4 0.111 0.115 1.538 1.563 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.533 1.543 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.475 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.217 0.945 1.421 dbcsr_complete_redistribute 325 12.2 0.239 0.293 1.116 1.379 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.802000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.000000, yerr=6.266796 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 660.119552E+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 0.014 0.028 60.136 60.136 qs_mol_dyn_low 1 2.0 0.003 0.003 59.875 59.885 qs_forces 11 3.9 0.006 0.010 59.744 59.752 qs_energies 11 4.9 0.001 0.002 56.513 56.522 scf_env_do_scf 11 5.9 0.000 0.001 48.591 48.592 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.998 39.999 velocity_verlet 10 3.0 0.002 0.002 32.811 32.815 dbcsr_multiply_generic 2055 12.4 0.109 0.110 28.202 28.417 qs_scf_new_mos 99 7.5 0.001 0.001 24.735 24.829 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.734 24.829 ot_scf_mini 99 9.5 0.002 0.003 23.506 23.613 multiply_cannon 2055 13.4 0.209 0.218 21.990 23.057 multiply_cannon_loop 2055 14.4 0.620 0.637 20.799 21.939 ot_mini 99 10.5 0.001 0.001 13.482 13.588 rebuild_ks_matrix 110 8.3 0.000 0.000 12.473 12.587 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.473 12.586 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.982 11.086 multiply_cannon_multrec 16440 15.4 3.928 4.806 9.675 10.504 mp_waitall_1 139946 16.5 6.826 10.060 6.826 10.060 qs_ot_get_derivative 99 11.5 0.001 0.001 9.042 9.147 init_scf_loop 11 6.9 0.000 0.000 8.554 8.554 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.233 7.323 prepare_preconditioner 11 7.9 0.000 0.000 6.791 6.808 make_preconditioner 11 8.9 0.000 0.000 6.791 6.808 sum_up_and_integrate 110 10.3 0.059 0.061 6.657 6.670 integrate_v_rspace 110 11.3 0.003 0.003 6.597 6.610 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.141 6.494 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.118 6.128 calculate_rho_elec 110 8.6 0.058 0.058 6.117 6.127 dbcsr_mm_accdrv_process 34862 16.1 4.735 5.298 5.601 5.788 init_scf_run 11 5.9 0.000 0.001 5.515 5.515 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.514 5.514 qs_ot_get_p 110 10.4 0.001 0.001 5.264 5.393 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.940 5.348 apply_single 110 13.6 0.000 0.000 4.940 5.348 make_m2s 4110 13.4 0.050 0.052 4.084 4.427 ot_diis_step 99 11.5 0.011 0.011 4.408 4.408 density_rs2pw 110 9.6 0.004 0.005 3.213 4.368 make_images 4110 14.4 0.393 0.509 3.969 4.311 multiply_cannon_sync_h2d 16440 15.4 3.691 4.206 3.691 4.206 rs_pw_transfer 902 11.9 0.010 0.011 2.782 3.950 pw_transfer 1331 11.6 0.065 0.072 3.777 3.788 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.077 3.718 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.670 3.685 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.642 3.646 grid_integrate_task_list 110 12.3 3.162 3.425 3.162 3.425 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.298 3.299 fft_wrap_pw1pw2_140 451 13.1 0.213 0.216 3.153 3.166 fft3d_ps 1111 14.6 1.086 1.097 2.969 2.982 wfi_extrapolate 11 7.9 0.001 0.001 2.972 2.972 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.817 2.818 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.778 2.779 cp_fm_redistribute_end 48 14.0 1.733 2.752 1.736 2.752 make_images_data 4110 15.4 0.043 0.047 2.314 2.741 cp_fm_diag_elpa_base 48 14.0 0.955 2.599 1.012 2.709 hybrid_alltoall_any 4261 16.3 0.105 0.373 2.098 2.676 cp_fm_cholesky_invert 11 10.9 2.651 2.657 2.651 2.657 calculate_dm_sparse 110 9.5 0.001 0.001 2.529 2.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.430 2.473 calculate_first_density_matrix 1 7.0 0.000 0.000 2.456 2.457 grid_collocate_task_list 110 9.6 2.085 2.434 2.085 2.434 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.347 2.410 potential_pw2rs 110 12.3 0.010 0.011 2.379 2.396 mp_waitany 17072 13.8 1.182 2.386 1.182 2.386 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.871 2.334 mp_alltoall_d11v 2046 13.8 1.789 2.254 1.789 2.254 mp_irecv_dv 48980 15.7 0.801 2.217 0.801 2.217 mp_sum_l 6514 12.8 1.510 2.121 1.510 2.121 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.181 0.888 2.033 qs_energies_init_hamiltonians 11 5.9 0.009 0.016 1.979 1.988 dbcsr_complete_redistribute 325 12.2 0.386 0.418 1.487 1.939 cp_fm_upper_to_full 70 13.6 1.390 1.866 1.390 1.866 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.690 1.702 cp_fm_cholesky_decompose 22 10.9 1.617 1.641 1.617 1.641 mp_allgather_i34 2055 14.4 0.520 1.496 0.520 1.496 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.482 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.478 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.998 1.461 rs_gather_matrices 110 12.3 0.232 0.258 0.975 1.380 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.264 1.270 acc_transpose_blocks 16440 15.4 0.072 0.074 1.207 1.226 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.136000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.909091, yerr=9.548484 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 735.997952E+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 0.044 0.112 65.666 65.711 qs_mol_dyn_low 1 2.0 0.003 0.004 65.314 65.324 qs_forces 11 3.9 0.003 0.010 65.247 65.248 qs_energies 11 4.9 0.006 0.021 61.812 61.817 scf_env_do_scf 11 5.9 0.001 0.003 53.441 53.444 scf_env_do_scf_inner_loop 99 6.5 0.006 0.011 41.614 41.615 velocity_verlet 10 3.0 0.002 0.002 37.305 37.307 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.393 29.626 qs_scf_new_mos 99 7.5 0.001 0.001 26.338 26.444 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.338 26.443 ot_scf_mini 99 9.5 0.003 0.004 24.718 24.840 multiply_cannon 2055 13.4 0.242 0.258 22.451 23.381 multiply_cannon_loop 2055 14.4 0.887 0.911 21.008 21.620 ot_mini 99 10.5 0.001 0.001 14.069 14.190 multiply_cannon_multrec 24660 15.4 4.217 6.833 12.670 13.948 rebuild_ks_matrix 110 8.3 0.000 0.001 12.351 12.454 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.018 12.351 12.454 init_scf_loop 11 6.9 0.001 0.004 11.785 11.786 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.905 10.999 qs_ot_get_derivative 99 11.5 0.001 0.001 9.919 10.037 prepare_preconditioner 11 7.9 0.000 0.000 10.007 10.023 make_preconditioner 11 8.9 0.002 0.005 10.007 10.023 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.213 9.707 dbcsr_mm_accdrv_process 52304 16.0 7.049 8.610 8.307 9.237 sum_up_and_integrate 110 10.3 0.068 0.071 6.599 6.611 integrate_v_rspace 110 11.3 0.003 0.003 6.531 6.540 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.233 6.242 calculate_rho_elec 110 8.6 0.077 0.081 6.233 6.242 mp_waitall_1 121746 16.5 4.243 6.062 4.243 6.062 qs_ot_get_p 110 10.4 0.001 0.001 5.647 5.805 make_m2s 4110 13.4 0.060 0.062 5.350 5.756 make_images 4110 14.4 0.576 0.700 5.208 5.611 init_scf_run 11 5.9 0.005 0.024 5.555 5.556 scf_env_initial_rho_setup 11 6.9 0.005 0.023 5.550 5.555 cp_fm_upper_to_full 70 13.8 3.337 4.755 3.337 4.755 ot_diis_step 99 11.5 0.011 0.012 4.111 4.112 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.007 4.111 apply_single 110 13.6 0.000 0.000 4.007 4.110 pw_transfer 1331 11.6 0.065 0.075 3.924 3.955 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.884 3.897 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.819 3.853 dbcsr_complete_redistribute 325 12.2 0.419 0.461 2.682 3.840 density_rs2pw 110 9.6 0.004 0.004 3.143 3.652 grid_integrate_task_list 110 12.3 3.267 3.449 3.267 3.449 cp_dbcsr_syevd 48 12.0 0.004 0.006 3.395 3.395 multiply_cannon_sync_h2d 24660 15.4 3.196 3.381 3.196 3.381 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.296 3.358 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 3.264 3.299 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.143 3.282 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.378 3.239 fft3d_ps 1111 14.6 1.094 1.128 3.110 3.139 hybrid_alltoall_any 4261 16.3 0.120 0.456 2.312 3.131 make_images_data 4110 15.4 0.047 0.050 2.704 3.097 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.004 3.069 wfi_extrapolate 11 7.9 0.001 0.001 2.999 2.999 rs_pw_transfer 902 11.9 0.011 0.011 2.449 2.969 calculate_dm_sparse 110 9.5 0.001 0.001 2.920 2.951 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.783 2.910 mp_alltoall_i22 605 13.7 1.681 2.856 1.681 2.856 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.824 2.827 cp_fm_redistribute_end 48 14.0 1.413 2.797 1.415 2.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.778 2.780 cp_fm_diag_elpa_base 48 14.0 1.301 2.659 1.380 2.773 cp_fm_cholesky_invert 11 10.9 2.738 2.746 2.738 2.746 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.416 2.465 calculate_first_density_matrix 1 7.0 0.001 0.003 2.454 2.464 grid_collocate_task_list 110 9.6 2.181 2.441 2.181 2.441 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.298 2.300 potential_pw2rs 110 12.3 0.013 0.013 2.234 2.240 jit_kernel_multiply 12 15.7 0.929 2.210 0.929 2.210 mp_alltoall_d11v 2046 13.8 1.788 2.098 1.788 2.098 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.696 1.728 cp_fm_cholesky_decompose 22 10.9 1.668 1.720 1.668 1.720 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.691 1.702 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.699 mp_waitany 13376 13.8 1.092 1.591 1.092 1.591 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.551 acc_transpose_blocks 24660 15.4 0.105 0.107 1.514 1.533 mp_sum_l 6514 12.8 0.919 1.527 0.919 1.527 multiply_cannon_metrocomm4 20550 15.4 0.058 0.061 0.843 1.485 mp_allgather_i34 2055 14.4 0.543 1.451 0.543 1.451 mp_irecv_dv 62702 16.1 0.743 1.404 0.743 1.404 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.185 0.836 1.366 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.248 1.352 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.711000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.454545, yerr=10.646072 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 842.502144E+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 0.014 0.030 56.187 56.188 qs_mol_dyn_low 1 2.0 0.003 0.003 55.944 55.953 qs_forces 11 3.9 0.002 0.002 55.777 55.778 qs_energies 11 4.9 0.001 0.002 52.063 52.067 scf_env_do_scf 11 5.9 0.001 0.001 43.821 43.822 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 35.956 35.958 velocity_verlet 10 3.0 0.001 0.002 31.647 31.656 dbcsr_multiply_generic 2055 12.4 0.105 0.107 22.848 23.018 qs_scf_new_mos 99 7.5 0.001 0.001 20.624 20.686 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.623 20.685 ot_scf_mini 99 9.5 0.002 0.002 19.381 19.426 multiply_cannon 2055 13.4 0.247 0.255 17.397 18.762 multiply_cannon_loop 2055 14.4 0.321 0.331 16.069 16.386 rebuild_ks_matrix 110 8.3 0.000 0.000 12.123 12.171 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.122 12.171 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.782 10.822 ot_mini 99 10.5 0.001 0.001 10.490 10.535 multiply_cannon_multrec 8220 15.4 3.231 4.945 7.445 8.745 init_scf_loop 11 6.9 0.000 0.000 7.812 7.814 mp_waitall_1 103326 16.6 6.153 7.670 6.153 7.670 qs_ot_get_derivative 99 11.5 0.001 0.001 6.630 6.676 sum_up_and_integrate 110 10.3 0.079 0.080 6.657 6.671 integrate_v_rspace 110 11.3 0.003 0.003 6.578 6.590 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.422 6.439 calculate_rho_elec 110 8.6 0.115 0.116 6.422 6.439 prepare_preconditioner 11 7.9 0.000 0.000 6.101 6.107 make_preconditioner 11 8.9 0.000 0.000 6.101 6.107 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.670 5.746 init_scf_run 11 5.9 0.000 0.001 5.101 5.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.100 5.101 qs_ot_get_p 110 10.4 0.001 0.001 4.968 5.016 dbcsr_mm_accdrv_process 17442 15.9 2.834 3.788 4.083 4.973 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.132 4.524 make_m2s 4110 13.4 0.039 0.040 4.146 4.379 pw_transfer 1331 11.6 0.066 0.071 4.237 4.250 make_images 4110 14.4 0.636 0.689 4.015 4.248 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.130 4.147 ot_diis_step 99 11.5 0.012 0.014 3.838 3.838 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.743 3.794 apply_single 110 13.6 0.000 0.000 3.743 3.794 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.596 3.599 fft_wrap_pw1pw2_140 451 13.1 0.216 0.221 3.560 3.582 density_rs2pw 110 9.6 0.004 0.004 3.165 3.542 grid_integrate_task_list 110 12.3 3.353 3.528 3.353 3.528 fft3d_ps 1111 14.6 1.148 1.188 3.373 3.387 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.294 3.295 multiply_cannon_sync_h2d 8220 15.4 2.898 3.027 2.898 3.027 cp_fm_cholesky_invert 11 10.9 2.929 2.933 2.929 2.933 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.778 2.779 wfi_extrapolate 11 7.9 0.001 0.001 2.765 2.765 cp_fm_redistribute_end 48 14.0 0.707 2.749 0.711 2.749 make_images_data 4110 15.4 0.038 0.043 2.341 2.727 cp_fm_diag_elpa_base 48 14.0 1.857 2.555 2.031 2.717 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.687 2.688 rs_pw_transfer 902 11.9 0.010 0.010 2.205 2.610 hybrid_alltoall_any 4261 16.3 0.199 0.863 2.277 2.608 grid_collocate_task_list 110 9.6 2.279 2.540 2.279 2.540 calculate_dm_sparse 110 9.5 0.001 0.001 2.462 2.505 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.493 2.495 calculate_first_density_matrix 1 7.0 0.000 0.000 2.237 2.238 potential_pw2rs 110 12.3 0.015 0.016 2.192 2.198 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.132 2.153 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.780 2.006 mp_alltoall_d11v 2046 13.8 1.669 1.935 1.669 1.935 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.902 1.927 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.798 1.823 cp_fm_cholesky_decompose 22 10.9 1.705 1.718 1.705 1.718 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.676 1.684 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.520 1.629 jit_kernel_multiply 9 15.8 0.943 1.582 0.943 1.582 mp_allgather_i34 2055 14.4 0.515 1.572 0.515 1.572 dbcsr_complete_redistribute 325 12.2 0.561 0.580 1.470 1.568 mp_waitany 9240 13.8 1.071 1.495 1.071 1.495 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.485 1.437 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.422 1.433 mp_irecv_dv 24056 15.7 0.461 1.396 0.461 1.396 qs_create_task_list 11 7.9 0.000 0.001 1.223 1.331 generate_qs_task_list 11 8.9 0.380 0.448 1.222 1.331 multiply_cannon_metrocomm1 8220 15.4 0.021 0.021 0.827 1.245 rs_gather_matrices 110 12.3 0.324 0.364 0.945 1.202 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.794 1.196 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.158 1.181 mp_alltoall_z22v 1111 16.6 1.115 1.142 1.115 1.142 x_to_yz 550 16.1 0.572 0.580 1.117 1.125 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.188000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=787.636364, yerr=14.599332 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 1.321005E+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 0.033 0.069 88.394 88.396 qs_mol_dyn_low 1 2.0 0.003 0.003 88.057 88.067 qs_forces 11 3.9 0.002 0.002 87.982 87.984 qs_energies 11 4.9 0.001 0.001 83.758 83.762 scf_env_do_scf 11 5.9 0.001 0.001 73.765 73.765 velocity_verlet 10 3.0 0.002 0.003 56.496 56.502 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 44.900 44.902 init_scf_loop 11 6.9 0.000 0.000 28.791 28.793 dbcsr_multiply_generic 2055 12.4 0.119 0.121 28.607 28.704 prepare_preconditioner 11 7.9 0.000 0.000 26.746 26.755 make_preconditioner 11 8.9 0.000 0.000 26.746 26.755 qs_scf_new_mos 99 7.5 0.001 0.001 26.335 26.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.334 26.392 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.830 26.221 ot_scf_mini 99 9.5 0.002 0.002 24.576 24.621 multiply_cannon 2055 13.4 0.348 0.369 21.549 22.159 multiply_cannon_loop 2055 14.4 0.341 0.344 19.774 20.123 cp_fm_upper_to_full 70 14.2 12.925 18.568 12.925 18.568 rebuild_ks_matrix 110 8.3 0.001 0.001 14.344 14.398 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.343 14.397 ot_mini 99 10.5 0.001 0.001 13.675 13.713 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.998 13.047 dbcsr_complete_redistribute 325 12.2 1.018 1.043 7.490 10.805 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.463 9.773 multiply_cannon_multrec 8220 15.4 4.442 4.624 9.633 9.719 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.900 9.174 qs_ot_get_derivative 99 11.5 0.001 0.001 9.058 9.106 mp_alltoall_i22 605 13.7 5.534 8.872 5.534 8.872 mp_waitall_1 84994 16.7 7.609 8.498 7.609 8.498 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.850 7.887 calculate_rho_elec 110 8.6 0.227 0.227 7.850 7.887 sum_up_and_integrate 110 10.3 0.151 0.152 7.460 7.472 integrate_v_rspace 110 11.3 0.004 0.004 7.309 7.321 make_m2s 4110 13.4 0.044 0.044 5.294 5.793 qs_ot_get_p 110 10.4 0.001 0.001 5.713 5.760 init_scf_run 11 5.9 0.000 0.001 5.713 5.713 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.713 5.713 make_images 4110 14.4 0.881 0.922 5.105 5.602 cp_fm_cholesky_invert 11 10.9 5.385 5.390 5.385 5.390 pw_transfer 1331 11.6 0.075 0.075 5.384 5.387 dbcsr_mm_accdrv_process 11614 15.7 3.463 4.133 5.048 5.298 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.268 5.270 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.718 5.212 apply_single 110 13.6 0.000 0.000 4.717 5.211 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.788 5.146 ot_diis_step 99 11.5 0.015 0.016 4.593 4.593 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 4.579 4.583 fft3d_ps 1111 14.6 1.297 1.304 4.461 4.465 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.152 4.159 multiply_cannon_sync_h2d 8220 15.4 3.953 3.959 3.953 3.959 density_rs2pw 110 9.6 0.004 0.004 3.871 3.911 grid_integrate_task_list 110 12.3 3.677 3.724 3.677 3.724 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.708 3.708 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.689 3.691 hybrid_alltoall_any 4261 16.3 0.257 0.557 2.935 3.603 make_images_data 4110 15.4 0.042 0.045 2.922 3.548 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.864 3.330 wfi_extrapolate 11 7.9 0.001 0.001 3.317 3.318 calculate_dm_sparse 110 9.5 0.001 0.001 3.117 3.143 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.108 3.108 cp_fm_diag_elpa_base 48 14.0 2.571 2.772 3.106 3.106 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.864 2.867 grid_collocate_task_list 110 9.6 2.641 2.679 2.641 2.679 potential_pw2rs 110 12.3 0.021 0.022 2.550 2.555 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.388 2.413 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.367 2.391 rs_pw_transfer 902 11.9 0.011 0.011 2.300 2.344 calculate_first_density_matrix 1 7.0 0.000 0.000 2.269 2.270 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.205 2.260 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.100 2.207 cp_fm_cholesky_decompose 22 10.9 1.997 2.016 1.997 2.016 mp_alltoall_d11v 2046 13.8 1.940 2.014 1.940 2.014 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.952 1.957 qs_create_task_list 11 7.9 0.001 0.001 1.895 1.942 generate_qs_task_list 11 8.9 0.735 0.788 1.895 1.941 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.756 1.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=88.396000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1183.000000, yerr=49.979996 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 628.682752E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.036 203.657 203.658 qs_mol_dyn_low 1 2.0 0.003 0.003 203.263 203.277 qs_forces 11 3.9 0.003 0.004 202.401 202.402 qs_energies 11 4.9 0.002 0.002 196.802 196.816 scf_env_do_scf 11 5.9 0.001 0.001 180.413 180.417 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 159.820 159.821 dbcsr_multiply_generic 2507 12.6 0.178 0.182 122.479 123.432 velocity_verlet 10 3.0 0.002 0.003 123.158 123.167 qs_scf_new_mos 117 7.6 0.001 0.001 119.814 120.053 qs_scf_loop_do_ot 117 8.6 0.001 0.001 119.813 120.053 ot_scf_mini 117 9.6 0.003 0.003 113.267 113.490 multiply_cannon 2507 13.6 0.237 0.250 100.386 102.832 multiply_cannon_loop 2507 14.6 2.108 2.194 98.213 100.489 ot_mini 117 10.6 0.001 0.001 64.618 64.841 multiply_cannon_multrec 60168 15.6 33.468 36.186 41.712 44.022 qs_ot_get_derivative 117 11.6 0.001 0.001 39.708 39.919 rebuild_ks_matrix 128 8.3 0.001 0.001 33.652 34.017 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.024 33.651 34.016 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.249 30.574 multiply_cannon_sync_h2d 60168 15.6 27.671 30.327 27.671 30.327 mp_waitall_1 267128 16.5 27.006 30.196 27.006 30.196 qs_ot_get_p 128 10.4 0.001 0.001 26.955 27.303 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.349 25.505 apply_single 128 13.6 0.001 0.001 24.348 25.505 ot_diis_step 117 11.6 0.007 0.008 24.581 24.582 init_scf_loop 11 6.9 0.000 0.001 20.522 20.524 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.249 20.291 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.242 18.396 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.843 17.843 multiply_cannon_metrocomm3 60168 15.6 0.118 0.124 15.368 17.592 prepare_preconditioner 11 7.9 0.000 0.000 15.949 15.992 make_preconditioner 11 8.9 0.000 0.000 15.949 15.992 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.242 15.397 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.897 14.902 cp_fm_redistribute_end 83 14.4 11.775 14.821 11.788 14.825 cp_fm_diag_elpa_base 83 14.4 2.992 14.448 3.024 14.556 sum_up_and_integrate 128 10.3 0.090 0.108 14.277 14.293 integrate_v_rspace 128 11.3 0.003 0.004 14.186 14.205 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.711 13.829 calculate_rho_elec 128 8.7 0.045 0.064 13.710 13.828 make_m2s 5014 13.6 0.106 0.113 13.370 13.770 make_images 5014 14.6 0.404 0.423 13.186 13.595 init_scf_run 11 5.9 0.000 0.001 12.255 12.255 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.254 12.255 density_rs2pw 128 9.7 0.007 0.007 7.192 10.345 cp_fm_cholesky_invert 11 10.9 9.186 9.194 9.186 9.194 rs_pw_transfer 1046 11.9 0.017 0.019 5.895 9.072 wfi_extrapolate 11 7.9 0.001 0.001 9.064 9.064 calculate_dm_sparse 128 9.5 0.001 0.001 8.356 8.436 dbcsr_mm_accdrv_process 124484 16.2 3.207 3.460 7.799 8.422 mp_sum_l 7870 13.0 7.472 8.362 7.472 8.362 pw_transfer 1547 11.6 0.076 0.097 7.811 8.051 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.864 7.994 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.605 7.851 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.703 7.770 grid_integrate_task_list 128 12.3 7.038 7.502 7.038 7.502 make_images_data 5014 15.6 0.066 0.072 6.345 7.177 multiply_cannon_metrocomm1 60168 15.6 0.088 0.094 5.468 6.845 fft_wrap_pw1pw2_140 523 13.2 0.445 0.515 6.539 6.837 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.586 6.593 fft3d_ps 1291 14.7 2.106 2.903 6.394 6.562 hybrid_alltoall_any 5200 16.5 0.291 2.269 5.554 6.541 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.772 5.872 mp_waitany 16020 13.9 2.626 5.777 2.626 5.777 grid_collocate_task_list 128 9.7 4.555 5.776 4.555 5.776 rs_pw_transfer_RS2PW_140 139 11.5 0.282 0.297 2.095 5.277 mp_alltoall_d11v 2415 14.1 3.861 5.236 3.861 5.236 potential_pw2rs 128 12.3 0.009 0.011 4.739 4.763 cp_fm_cholesky_decompose 22 10.9 4.675 4.690 4.675 4.690 mp_sum_d 4457 12.1 3.619 4.522 3.619 4.522 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=203.658000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.272727, yerr=7.556946 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 836.648960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132588E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702167E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828901424 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4129 56608. MP_Allreduce 11362 942. MP_Sync 172 MP_Alltoall 1983 5621096. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_comm_split 84 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.099 191.489 191.490 qs_mol_dyn_low 1 2.0 0.004 0.014 190.975 190.991 qs_forces 11 3.9 0.006 0.010 190.775 190.781 qs_energies 11 4.9 0.003 0.016 183.959 183.970 scf_env_do_scf 11 5.9 0.001 0.001 167.502 167.513 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 134.578 134.581 velocity_verlet 10 3.0 0.010 0.048 120.883 120.885 dbcsr_multiply_generic 2529 12.6 0.188 0.194 97.413 98.521 qs_scf_new_mos 118 7.6 0.001 0.001 94.075 94.502 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.075 94.501 ot_scf_mini 118 9.6 0.004 0.005 89.243 89.649 multiply_cannon 2529 13.6 0.483 0.536 77.773 82.408 multiply_cannon_loop 2529 14.6 1.272 1.311 74.600 76.937 ot_mini 118 10.6 0.001 0.001 49.831 50.275 multiply_cannon_multrec 30348 15.6 22.188 26.902 31.870 36.789 mp_waitall_1 216598 16.6 24.076 36.703 24.076 36.703 rebuild_ks_matrix 129 8.3 0.001 0.001 33.602 34.033 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.023 33.601 34.032 init_scf_loop 11 6.9 0.001 0.010 32.838 32.839 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.276 30.681 multiply_cannon_metrocomm3 30348 15.6 0.098 0.104 15.715 28.523 prepare_preconditioner 11 7.9 0.000 0.001 28.412 28.482 make_preconditioner 11 8.9 0.000 0.002 28.412 28.482 qs_ot_get_derivative 118 11.6 0.001 0.002 27.793 28.219 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.123 27.650 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.077 23.224 apply_single 129 13.6 0.001 0.001 22.076 23.224 multiply_cannon_sync_h2d 30348 15.6 19.455 22.129 19.455 22.129 ot_diis_step 118 11.6 0.014 0.015 21.861 21.863 qs_ot_get_p 129 10.4 0.001 0.001 21.162 21.641 cp_fm_cholesky_invert 11 10.9 16.625 16.638 16.625 16.638 qs_ot_p2m_diag 84 11.4 0.189 0.218 16.365 16.398 make_m2s 5058 13.6 0.091 0.096 13.960 15.419 cp_dbcsr_syevd 84 12.4 0.005 0.006 15.207 15.208 make_images 5058 14.6 1.159 1.351 13.747 15.207 sum_up_and_integrate 129 10.3 0.117 0.133 14.883 14.915 integrate_v_rspace 129 11.3 0.003 0.003 14.765 14.801 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.246 14.283 calculate_rho_elec 129 8.7 0.089 0.107 14.246 14.283 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.178 12.187 cp_fm_redistribute_end 84 14.4 7.112 12.115 7.122 12.115 cp_fm_diag_elpa_base 84 14.4 4.756 11.605 4.975 11.967 init_scf_run 11 5.9 0.000 0.001 11.586 11.588 scf_env_initial_rho_setup 11 6.9 0.001 0.005 11.586 11.588 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.985 11.317 multiply_cannon_metrocomm4 27819 15.6 0.097 0.110 3.808 10.176 make_images_data 5058 15.6 0.064 0.073 8.300 10.030 density_rs2pw 129 9.7 0.006 0.007 7.579 9.922 mp_irecv_dv 70084 16.3 3.612 9.802 3.612 9.802 dbcsr_mm_accdrv_process 62758 16.2 4.526 5.232 9.135 9.722 hybrid_alltoall_any 5245 16.5 0.346 1.523 7.210 9.601 pw_transfer 1559 11.6 0.087 0.111 9.071 9.146 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.843 8.925 wfi_extrapolate 11 7.9 0.001 0.001 8.378 8.379 rs_pw_transfer 1054 12.0 0.015 0.017 5.803 8.169 fft_wrap_pw1pw2_140 527 13.2 0.476 0.532 7.798 7.906 grid_integrate_task_list 129 12.3 7.225 7.736 7.225 7.736 fft3d_ps 1301 14.7 2.767 2.943 7.196 7.243 cp_fm_cholesky_decompose 22 10.9 6.905 7.000 6.905 7.000 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.151 6.935 calculate_dm_sparse 129 9.5 0.001 0.001 6.481 6.676 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.081 6.091 grid_collocate_task_list 129 9.7 4.735 5.911 4.735 5.911 mp_sum_l 7936 13.1 3.925 5.652 3.925 5.652 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.353 5.418 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.245 5.389 potential_pw2rs 129 12.3 0.015 0.018 5.117 5.131 mp_waitany 11836 13.9 2.524 4.914 2.524 4.914 mp_allgather_i34 2529 14.6 1.741 4.680 1.741 4.680 mp_alltoall_d11v 2429 14.1 3.943 4.635 3.943 4.635 rs_pw_transfer_RS2PW_140 140 11.5 0.357 0.390 2.132 4.460 mp_sum_d 4502 12.2 2.711 4.045 2.711 4.045 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.490000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=794.727273, yerr=5.136163 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 931.778560E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57197. MP_Allreduce 11251 985. MP_Sync 170 MP_Alltoall 1712 9388896. MP_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 0.069 0.096 176.006 176.014 qs_mol_dyn_low 1 2.0 0.003 0.003 175.542 175.555 qs_forces 11 3.9 0.003 0.003 175.419 175.430 qs_energies 11 4.9 0.006 0.026 168.772 168.789 scf_env_do_scf 11 5.9 0.001 0.004 153.265 153.266 scf_env_do_scf_inner_loop 117 6.6 0.005 0.012 118.376 118.377 velocity_verlet 10 3.0 0.005 0.011 112.874 112.876 dbcsr_multiply_generic 2507 12.6 0.182 0.186 81.080 82.041 qs_scf_new_mos 117 7.6 0.001 0.001 80.013 80.309 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.012 80.308 ot_scf_mini 117 9.6 0.004 0.007 75.898 76.264 multiply_cannon 2507 13.6 0.497 0.525 61.199 65.192 multiply_cannon_loop 2507 14.6 0.860 0.894 58.104 60.613 ot_mini 117 10.6 0.001 0.001 42.211 42.567 init_scf_loop 11 6.9 0.001 0.006 34.789 34.790 mp_waitall_1 170520 16.6 24.567 34.649 24.567 34.649 rebuild_ks_matrix 128 8.3 0.001 0.001 30.911 31.321 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.910 31.321 prepare_preconditioner 11 7.9 0.000 0.001 30.647 30.700 make_preconditioner 11 8.9 0.000 0.003 30.647 30.700 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.323 29.684 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.838 28.229 multiply_cannon_multrec 20056 15.6 13.705 16.985 22.361 25.530 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 14.672 24.131 qs_ot_get_derivative 117 11.6 0.001 0.002 22.471 22.832 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.812 20.770 apply_single 128 13.6 0.001 0.001 19.812 20.770 ot_diis_step 117 11.6 0.018 0.018 19.630 19.630 qs_ot_get_p 128 10.4 0.001 0.001 19.103 19.554 make_m2s 5014 13.6 0.082 0.087 14.991 15.919 multiply_cannon_sync_h2d 20056 15.6 14.282 15.840 14.282 15.840 make_images 5014 14.6 1.180 1.283 14.756 15.681 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.891 14.899 sum_up_and_integrate 128 10.3 0.134 0.145 14.771 14.797 integrate_v_rspace 128 11.3 0.003 0.004 14.637 14.663 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.373 14.416 calculate_rho_elec 128 8.7 0.132 0.147 14.372 14.415 cp_fm_cholesky_invert 11 10.9 14.282 14.291 14.282 14.291 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.889 13.891 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.853 10.854 cp_fm_redistribute_end 83 14.4 4.102 10.791 4.116 10.793 make_images_data 5014 15.6 0.061 0.069 9.352 10.752 cp_fm_diag_elpa_base 83 14.4 6.257 10.217 6.658 10.673 init_scf_run 11 5.9 0.000 0.001 10.436 10.436 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.436 10.436 hybrid_alltoall_any 5200 16.5 0.435 2.005 7.991 9.785 density_rs2pw 128 9.7 0.006 0.007 7.285 9.390 pw_transfer 1547 11.6 0.086 0.106 9.146 9.258 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.817 9.087 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.921 9.039 multiply_cannon_metrocomm4 17549 15.6 0.061 0.070 3.394 8.946 mp_irecv_dv 50230 16.2 3.273 8.700 3.273 8.700 dbcsr_mm_accdrv_process 41502 16.2 4.366 5.173 8.105 8.218 fft_wrap_pw1pw2_140 523 13.2 0.479 0.525 7.885 8.006 grid_integrate_task_list 128 12.3 7.299 7.693 7.299 7.693 cp_fm_upper_to_full 105 14.5 5.744 7.479 5.744 7.479 wfi_extrapolate 11 7.9 0.001 0.001 7.445 7.445 cp_fm_cholesky_decompose 22 10.9 7.246 7.287 7.246 7.287 fft3d_ps 1291 14.7 2.673 2.917 7.181 7.276 rs_pw_transfer 1046 11.9 0.014 0.015 5.135 7.203 dbcsr_complete_redistribute 395 12.7 1.168 1.196 4.428 6.195 calculate_dm_sparse 128 9.5 0.001 0.001 5.718 5.805 grid_collocate_task_list 128 9.7 4.903 5.697 4.903 5.697 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.412 5.421 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.509 5.184 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.237 5.002 potential_pw2rs 128 12.3 0.020 0.023 4.860 4.876 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.540 4.675 mp_allgather_i34 2507 14.6 1.614 4.671 1.614 4.671 mp_alltoall_d11v 2415 14.1 4.175 4.638 4.175 4.638 mp_waitany 11748 13.9 2.325 4.496 2.325 4.496 mp_sum_l 7870 13.0 3.094 4.437 3.094 4.437 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.052 4.084 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.304 4.031 rs_pw_transfer_RS2PW_140 139 11.5 0.335 0.358 1.905 3.979 mp_alltoall_i22 716 14.1 1.871 3.765 1.871 3.765 qs_energies_init_hamiltonians 11 5.9 0.003 0.015 3.744 3.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.586 3.628 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.014000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.636364, yerr=10.516418 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 4.320341E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.133924E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4033 57915. MP_Allreduce 11121 1072. MP_Sync 168 MP_Alltoall 1700 12496378. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_comm_split 82 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.103 186.287 186.288 qs_mol_dyn_low 1 2.0 0.003 0.004 185.768 185.781 qs_forces 11 3.9 0.004 0.009 185.639 185.643 qs_energies 11 4.9 0.002 0.009 178.433 178.442 scf_env_do_scf 11 5.9 0.001 0.002 161.660 161.670 velocity_verlet 10 3.0 0.002 0.002 123.039 123.043 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 115.746 115.747 dbcsr_multiply_generic 2485 12.5 0.189 0.195 78.497 79.272 qs_scf_new_mos 116 7.6 0.001 0.001 78.300 78.624 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.299 78.623 ot_scf_mini 116 9.6 0.003 0.004 73.852 74.134 multiply_cannon 2485 13.5 0.551 0.595 54.052 57.021 multiply_cannon_loop 2485 14.5 1.173 1.202 50.567 52.451 init_scf_loop 11 6.9 0.001 0.005 45.795 45.796 ot_mini 116 10.6 0.001 0.001 41.812 42.106 prepare_preconditioner 11 7.9 0.000 0.000 41.635 41.667 make_preconditioner 11 8.9 0.001 0.004 41.635 41.667 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.353 40.250 multiply_cannon_multrec 29820 15.5 13.946 19.343 25.775 30.641 rebuild_ks_matrix 127 8.3 0.001 0.001 29.913 30.167 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.021 29.912 30.166 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.915 27.138 mp_waitall_1 146592 16.7 17.233 26.570 17.233 26.570 qs_ot_get_derivative 116 11.6 0.001 0.002 22.426 22.718 make_m2s 4970 13.5 0.095 0.099 20.121 21.031 make_images 4970 14.5 1.937 2.244 19.815 20.724 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.860 19.348 apply_single 127 13.6 0.001 0.001 18.860 19.348 ot_diis_step 116 11.6 0.017 0.018 19.251 19.252 qs_ot_get_p 127 10.4 0.001 0.001 18.456 18.752 cp_fm_upper_to_full 104 14.7 10.942 16.119 10.942 16.119 cp_fm_cholesky_invert 11 10.9 15.835 15.844 15.835 15.844 multiply_cannon_metrocomm3 29820 15.5 0.046 0.049 6.426 15.755 sum_up_and_integrate 127 10.3 0.142 0.158 14.777 14.801 integrate_v_rspace 127 11.3 0.005 0.016 14.635 14.665 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.615 14.659 calculate_rho_elec 127 8.7 0.174 0.189 14.615 14.659 qs_ot_p2m_diag 82 11.4 0.338 0.384 14.395 14.447 cp_dbcsr_syevd 82 12.4 0.005 0.008 13.143 13.144 multiply_cannon_sync_h2d 29820 15.5 11.620 12.686 11.620 12.686 dbcsr_complete_redistribute 393 12.7 1.512 1.624 8.798 12.509 make_images_data 4970 15.5 0.063 0.068 10.872 12.498 dbcsr_mm_accdrv_process 61748 16.2 7.313 7.995 11.412 11.914 hybrid_alltoall_any 5155 16.4 0.523 2.195 9.731 11.274 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.460 11.168 init_scf_run 11 5.9 0.000 0.001 10.788 10.789 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.788 10.789 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.127 10.129 cp_fm_redistribute_end 82 14.4 1.753 10.056 1.768 10.062 cp_fm_diag_elpa_base 82 14.4 7.697 9.452 8.261 9.929 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.258 9.880 pw_transfer 1535 11.6 0.087 0.105 9.504 9.568 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.236 9.450 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.278 9.351 mp_alltoall_i22 712 14.1 5.479 9.196 5.479 9.196 density_rs2pw 127 9.7 0.006 0.006 7.124 8.735 fft_wrap_pw1pw2_140 519 13.2 0.480 0.491 8.255 8.343 grid_integrate_task_list 127 12.3 7.470 7.941 7.470 7.941 wfi_extrapolate 11 7.9 0.001 0.001 7.655 7.655 cp_fm_cholesky_decompose 22 10.9 7.463 7.589 7.463 7.589 fft3d_ps 1281 14.7 2.759 2.835 7.510 7.573 multiply_cannon_metrocomm4 24850 15.5 0.075 0.084 2.780 7.214 mp_irecv_dv 75445 16.2 2.636 6.960 2.636 6.960 rs_pw_transfer 1038 11.9 0.013 0.015 4.649 6.316 calculate_dm_sparse 127 9.5 0.001 0.001 6.116 6.222 grid_collocate_task_list 127 9.7 5.025 5.788 5.025 5.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.449 5.491 mp_alltoall_d11v 2401 14.1 4.783 5.387 4.783 5.387 potential_pw2rs 127 12.3 0.022 0.022 4.624 4.641 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.512 4.517 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.387 4.438 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.327 4.417 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.125 4.177 mp_waitany 14840 13.9 2.258 3.911 2.258 3.911 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.288000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1074.454545, yerr=15.155939 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/19/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 5.820057E+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 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.497448E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3989 58394. MP_Allreduce 10970 1176. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.040 172.739 172.740 qs_mol_dyn_low 1 2.0 0.003 0.003 172.233 172.247 qs_forces 11 3.9 0.003 0.003 172.059 172.067 qs_energies 11 4.9 0.002 0.002 164.440 164.447 scf_env_do_scf 11 5.9 0.001 0.001 147.234 147.240 velocity_verlet 10 3.0 0.002 0.002 112.423 112.431 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 112.154 112.156 dbcsr_multiply_generic 2485 12.5 0.189 0.193 73.349 73.897 qs_scf_new_mos 116 7.6 0.001 0.001 73.534 73.625 qs_scf_loop_do_ot 116 8.6 0.001 0.001 73.533 73.625 ot_scf_mini 116 9.6 0.004 0.004 69.152 69.222 multiply_cannon 2485 13.5 0.579 0.622 53.863 58.637 multiply_cannon_loop 2485 14.5 0.439 0.446 49.359 50.377 ot_mini 116 10.6 0.001 0.001 39.133 39.198 init_scf_loop 11 6.9 0.000 0.000 34.931 34.933 mp_waitall_1 124680 16.7 25.667 33.170 25.667 33.170 prepare_preconditioner 11 7.9 0.000 0.000 31.016 31.049 make_preconditioner 11 8.9 0.000 0.000 31.016 31.049 rebuild_ks_matrix 127 8.3 0.001 0.001 29.870 29.981 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 29.870 29.980 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.942 29.209 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.106 27.204 multiply_cannon_multrec 9940 15.5 10.292 14.088 17.593 20.473 ot_diis_step 116 11.6 0.019 0.020 19.637 19.637 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.199 19.523 apply_single 127 13.6 0.001 0.001 19.198 19.523 qs_ot_get_derivative 116 11.6 0.001 0.002 19.430 19.497 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.223 19.272 make_m2s 4970 13.5 0.066 0.071 15.961 18.082 make_images 4970 14.5 2.254 2.633 15.656 17.770 cp_fm_cholesky_invert 11 10.9 17.630 17.636 17.630 17.636 qs_ot_get_p 127 10.4 0.001 0.001 17.286 17.370 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.410 15.458 calculate_rho_elec 127 8.7 0.257 0.266 15.409 15.457 sum_up_and_integrate 127 10.3 0.180 0.188 15.032 15.080 integrate_v_rspace 127 11.3 0.004 0.005 14.852 14.906 qs_ot_p2m_diag 82 11.4 0.489 0.495 13.608 13.625 cp_dbcsr_syevd 82 12.4 0.005 0.005 12.494 12.495 make_images_data 4970 15.5 0.051 0.060 9.780 12.214 hybrid_alltoall_any 5155 16.4 0.814 3.628 9.630 11.924 multiply_cannon_sync_h2d 9940 15.5 11.558 11.905 11.558 11.905 init_scf_run 11 5.9 0.000 0.001 10.408 10.408 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.408 10.408 pw_transfer 1535 11.6 0.085 0.094 9.872 9.917 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.652 9.702 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.511 9.524 cp_fm_diag_elpa_base 82 14.4 9.273 9.353 9.507 9.519 fft_wrap_pw1pw2_140 519 13.2 0.492 0.508 8.525 8.576 grid_integrate_task_list 127 12.3 7.719 8.065 7.719 8.065 density_rs2pw 127 9.7 0.005 0.006 7.039 7.909 cp_fm_cholesky_decompose 22 10.9 7.771 7.870 7.771 7.870 fft3d_ps 1281 14.7 2.688 2.763 7.825 7.850 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.792 7.845 dbcsr_mm_accdrv_process 20590 16.1 2.727 3.506 6.932 7.580 mp_allgather_i34 2485 14.5 2.731 7.564 2.731 7.564 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.402 7.424 wfi_extrapolate 11 7.9 0.001 0.001 7.356 7.356 calculate_dm_sparse 127 9.5 0.001 0.001 6.057 6.143 grid_collocate_task_list 127 9.7 5.352 6.010 5.352 6.010 mp_alltoall_d11v 2401 14.1 4.920 5.729 4.920 5.729 dbcsr_complete_redistribute 393 12.7 2.123 2.174 5.105 5.483 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.298 5.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.159 5.168 rs_pw_transfer 1038 11.9 0.012 0.013 4.075 4.983 potential_pw2rs 127 12.3 0.026 0.028 4.534 4.546 multiply_cannon_metrocomm4 7455 15.5 0.024 0.026 1.832 4.146 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.074 4.114 mp_irecv_dv 28618 15.9 1.796 4.072 1.796 4.072 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.571 3.919 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.662 3.674 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.637 3.667 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.531 3.643 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.316 3.635 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=172.740000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1395.727273, yerr=39.854487 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.038024E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58564. MP_Allreduce 11104 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.068 299.210 299.211 qs_mol_dyn_low 1 2.0 0.003 0.003 298.639 298.651 qs_forces 11 3.9 0.003 0.003 298.541 298.545 qs_energies 11 4.9 0.002 0.002 289.087 289.101 scf_env_do_scf 11 5.9 0.001 0.001 266.818 266.833 velocity_verlet 10 3.0 0.002 0.002 216.318 216.326 scf_env_do_scf_inner_loop 118 6.6 0.004 0.011 141.276 141.278 init_scf_loop 11 6.9 0.000 0.000 125.287 125.289 prepare_preconditioner 11 7.9 0.000 0.000 120.139 120.173 make_preconditioner 11 8.9 0.000 0.000 120.139 120.173 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.844 117.344 qs_scf_new_mos 118 7.6 0.001 0.001 89.835 89.960 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.834 89.959 ot_scf_mini 118 9.6 0.004 0.004 85.059 85.098 dbcsr_multiply_generic 2529 12.6 0.215 0.221 83.124 83.674 cp_fm_upper_to_full 106 14.8 53.944 77.307 53.944 77.307 multiply_cannon 2529 13.6 0.711 0.758 58.719 59.351 multiply_cannon_loop 2529 14.6 0.478 0.485 55.154 56.450 ot_mini 118 10.6 0.001 0.001 44.135 44.172 dbcsr_complete_redistribute 397 12.7 3.970 4.031 29.891 42.832 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.538 39.540 rebuild_ks_matrix 129 8.3 0.001 0.001 38.836 38.881 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 38.836 38.880 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.250 37.209 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.802 35.853 mp_alltoall_i22 720 14.1 22.056 34.992 22.056 34.992 cp_fm_cholesky_invert 11 10.9 32.995 33.002 32.995 33.002 mp_waitall_1 104580 16.8 28.063 32.076 28.063 32.076 qs_ot_get_p 129 10.4 0.001 0.001 25.369 25.379 qs_ot_get_derivative 118 11.6 0.002 0.002 24.130 24.170 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.379 21.405 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.816 20.842 calculate_rho_elec 129 8.7 0.486 0.487 20.815 20.842 make_m2s 5058 13.6 0.077 0.079 19.884 20.789 make_images 5058 14.6 3.818 3.997 19.399 20.308 sum_up_and_integrate 129 10.3 0.324 0.325 19.945 20.025 ot_diis_step 118 11.6 0.022 0.023 19.980 19.981 multiply_cannon_metrocomm3 10116 15.6 0.024 0.024 18.772 19.884 integrate_v_rspace 129 11.3 0.004 0.004 19.621 19.700 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.653 19.654 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.166 19.367 apply_single 129 13.6 0.001 0.001 19.165 19.366 multiply_cannon_multrec 10116 15.6 10.562 12.389 18.036 18.130 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.466 16.468 cp_fm_diag_elpa_base 84 14.4 12.037 13.672 16.463 16.463 multiply_cannon_sync_h2d 10116 15.6 15.782 15.815 15.782 15.815 hybrid_alltoall_any 5245 16.5 1.310 3.058 10.747 12.692 make_images_data 5058 15.6 0.061 0.067 10.566 12.298 pw_transfer 1559 11.6 0.093 0.094 12.288 12.298 init_scf_run 11 5.9 0.000 0.001 12.206 12.206 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.206 12.206 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 12.051 12.060 fft_wrap_pw1pw2_140 527 13.2 0.546 0.549 10.713 10.724 fft3d_ps 1301 14.7 2.762 2.769 10.070 10.078 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.456 9.492 mp_alltoall_d11v 2429 14.1 8.224 9.232 8.224 9.232 wfi_extrapolate 11 7.9 0.001 0.001 9.126 9.126 dbcsr_mm_accdrv_process 20934 16.1 3.862 5.695 7.234 9.111 grid_integrate_task_list 129 12.3 8.653 8.812 8.653 8.812 cp_fm_cholesky_decompose 22 10.9 8.725 8.765 8.725 8.765 density_rs2pw 129 9.7 0.005 0.005 8.314 8.449 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.056 8.059 calculate_dm_sparse 129 9.5 0.001 0.001 6.611 6.718 grid_collocate_task_list 129 9.7 6.410 6.448 6.410 6.448 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.344 6.422 rs_scatter_matrices 140 9.7 3.667 4.516 6.063 6.284 copy_dbcsr_to_fm 187 11.8 0.004 0.005 6.011 6.086 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=299.211000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2683.545455, yerr=180.076430 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 1.261396E+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 7111212. 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 0.018 0.036 84.931 84.932 qs_energies 1 2.0 0.000 0.000 84.502 84.512 ls_scf 1 3.0 0.000 0.000 83.585 83.595 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.639 72.808 multiply_cannon 111 7.7 0.017 0.020 55.967 57.421 multiply_cannon_loop 111 8.7 0.201 0.217 52.453 54.162 ls_scf_main 1 4.0 0.000 0.000 52.248 52.248 density_matrix_trs4 2 5.0 0.002 0.003 46.721 46.804 ls_scf_init_scf 1 4.0 0.000 0.000 28.310 28.311 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.178 27.231 mp_waitall_1 11031 10.9 22.613 25.136 22.613 25.136 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.050 25.071 multiply_cannon_multrec 2664 9.7 8.174 8.908 15.475 17.213 multiply_cannon_sync_h2d 2664 9.7 13.505 15.783 13.505 15.783 make_m2s 222 7.7 0.009 0.011 13.055 13.600 make_images 222 8.7 0.100 0.110 13.033 13.579 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.739 13.158 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.557 8.533 make_images_data 222 9.7 0.004 0.005 7.651 8.153 dbcsr_mm_accdrv_process 4760 10.4 0.504 0.617 6.920 7.888 hybrid_alltoall_any 227 10.6 0.215 1.838 6.573 7.847 dbcsr_mm_accdrv_process_sort 4760 11.4 6.214 7.091 6.214 7.091 calculate_norms 4752 9.8 5.522 6.096 5.522 6.096 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.052 5.172 mp_sum_l 807 5.4 3.079 4.908 3.079 4.908 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.306 3.773 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.065 3.535 mp_irecv_dv 6231 10.9 2.046 3.508 2.046 3.508 make_images_sizes 222 9.7 0.000 0.000 0.629 3.239 mp_alltoall_i44 222 10.7 0.628 3.239 0.628 3.239 arnoldi_extremal 4 6.8 0.000 0.000 3.204 3.235 arnoldi_normal_ev 4 7.8 0.001 0.003 3.204 3.234 build_subspace 16 8.4 0.009 0.012 3.098 3.101 ls_scf_post 1 4.0 0.000 0.000 3.027 3.037 ls_scf_store_result 1 5.0 0.000 0.000 2.845 2.887 dbcsr_special_finalize 555 9.7 0.005 0.006 2.421 2.790 dbcsr_merge_single_wm 555 10.7 0.487 0.631 2.412 2.782 make_images_pack 222 9.7 2.208 2.632 2.210 2.633 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.313 2.567 dbcsr_sort_data 658 11.4 2.184 2.490 2.184 2.490 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.471 2.069 2.472 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.284 2.353 buffer_matrices_ensure_size 222 8.7 1.813 2.154 1.813 2.154 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.806 1.807 rebuild_ks_matrix 3 7.3 0.000 0.000 1.797 1.798 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.797 1.798 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.932000, 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/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 2.119934E+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 0.029 0.044 91.427 91.428 qs_energies 1 2.0 0.000 0.000 90.974 90.979 ls_scf 1 3.0 0.000 0.000 89.656 89.661 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.677 76.002 multiply_cannon 111 7.7 0.028 0.036 53.488 57.563 ls_scf_main 1 4.0 0.000 0.000 55.269 55.274 multiply_cannon_loop 111 8.7 0.116 0.122 50.225 53.312 density_matrix_trs4 2 5.0 0.002 0.003 49.506 49.705 ls_scf_init_scf 1 4.0 0.000 0.000 30.827 30.828 mp_waitall_1 9105 10.9 21.678 30.202 21.678 30.202 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.616 29.725 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.206 27.219 multiply_cannon_multrec 1332 9.7 13.014 16.781 22.208 27.073 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.211 21.086 make_m2s 222 7.7 0.006 0.007 15.413 16.161 make_images 222 8.7 1.566 1.916 15.383 16.133 dbcsr_mm_accdrv_process 4041 10.4 0.280 0.457 8.792 10.386 dbcsr_mm_accdrv_process_sort 4041 11.4 8.383 9.933 8.383 9.933 make_images_data 222 9.7 0.004 0.004 8.875 9.896 hybrid_alltoall_any 227 10.6 0.521 2.459 8.272 9.415 mp_sum_l 807 5.4 5.481 8.477 5.481 8.477 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.228 7.676 mp_irecv_dv 3311 11.0 3.208 7.628 3.208 7.628 calculate_norms 2376 9.8 5.988 6.705 5.988 6.705 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.233 6.680 multiply_cannon_sync_h2d 1332 9.7 4.789 5.886 4.789 5.886 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.082 5.265 arnoldi_extremal 4 6.8 0.000 0.000 4.612 4.640 arnoldi_normal_ev 4 7.8 0.002 0.005 4.612 4.640 build_subspace 16 8.4 0.014 0.021 4.352 4.358 ls_scf_post 1 4.0 0.000 0.000 3.560 3.565 ls_scf_store_result 1 5.0 0.000 0.000 3.275 3.383 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.127 3.363 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.230 2.741 3.232 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.221 2.776 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.620 2.694 mp_allgather_i34 111 8.7 0.970 2.549 0.970 2.549 make_images_pack 222 9.7 2.020 2.431 2.022 2.433 dbcsr_sort_data 436 11.2 1.818 2.019 1.818 2.019 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.892 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.879 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.876 1.879 dbcsr_data_new 4174 10.1 1.612 1.845 1.612 1.845 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.428000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1769.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 2.685735E+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 0.054 0.161 94.892 94.893 qs_energies 1 2.0 0.012 0.051 94.102 94.173 ls_scf 1 3.0 0.000 0.001 92.667 92.708 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.361 77.661 ls_scf_main 1 4.0 0.000 0.003 58.106 58.111 multiply_cannon 111 7.7 0.039 0.085 53.175 57.065 multiply_cannon_loop 111 8.7 0.099 0.104 49.637 53.692 density_matrix_trs4 2 5.0 0.002 0.010 52.126 52.274 mp_waitall_1 7281 11.0 24.530 34.614 24.530 34.614 ls_scf_init_scf 1 4.0 0.001 0.006 30.983 30.985 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.725 29.796 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.314 27.333 multiply_cannon_multrec 888 9.7 12.478 16.064 20.993 25.166 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.419 23.787 make_m2s 222 7.7 0.006 0.006 17.320 18.523 make_images 222 8.7 1.968 2.269 17.282 18.483 hybrid_alltoall_any 227 10.6 0.620 2.863 9.543 11.034 make_images_data 222 9.7 0.003 0.004 9.916 10.925 dbcsr_mm_accdrv_process 3754 10.4 0.252 0.416 8.025 9.216 mp_sum_l 807 5.4 5.360 9.171 5.360 9.171 dbcsr_mm_accdrv_process_sort 3754 11.4 7.654 8.800 7.654 8.800 multiply_cannon_sync_h2d 888 9.7 5.956 8.045 5.956 8.045 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.082 7.354 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.521 6.942 mp_irecv_dv 2335 11.1 2.505 6.898 2.505 6.898 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.817 6.425 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.935 5.127 arnoldi_extremal 4 6.8 0.000 0.000 5.050 5.074 arnoldi_normal_ev 4 7.8 0.001 0.005 5.050 5.074 build_subspace 16 8.4 0.014 0.020 4.735 4.741 calculate_norms 1584 9.8 4.282 4.705 4.282 4.705 mp_allgather_i34 111 8.7 1.375 3.890 1.375 3.890 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.412 3.752 ls_scf_post 1 4.0 0.011 0.050 3.579 3.622 dbcsr_matrix_vector_mult_local 304 10.0 3.012 3.594 3.014 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.311 3.392 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.862 2.940 make_images_sizes 222 9.7 0.000 0.000 1.084 2.309 mp_alltoall_i44 222 10.7 1.084 2.308 1.084 2.308 make_images_pack 222 9.7 1.817 2.141 1.819 2.144 dbcsr_sort_data 325 11.1 1.894 2.143 1.894 2.143 dbcsr_data_release 9322 10.9 1.323 1.960 1.323 1.960 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.957 1.958 rebuild_ks_matrix 3 7.3 0.000 0.000 1.939 1.941 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.005 1.939 1.940 dbcsr_finalize 304 7.8 0.026 0.032 1.616 1.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.893000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2212.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 3.307454E+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 0.062 0.190 97.524 97.526 qs_energies 1 2.0 0.000 0.001 96.718 96.724 ls_scf 1 3.0 0.008 0.033 95.064 95.071 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.398 78.638 ls_scf_main 1 4.0 0.004 0.025 58.757 58.770 multiply_cannon 111 7.7 0.054 0.084 51.692 56.238 density_matrix_trs4 2 5.0 0.003 0.005 52.562 52.695 multiply_cannon_loop 111 8.7 0.115 0.126 46.673 49.546 ls_scf_init_scf 1 4.0 0.000 0.001 33.067 33.079 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.536 31.601 mp_waitall_1 6369 11.0 22.727 29.248 22.727 29.248 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.020 29.035 multiply_cannon_multrec 1332 9.7 14.232 17.133 22.061 24.890 make_m2s 222 7.7 0.007 0.009 21.061 22.445 make_images 222 8.7 3.126 3.597 21.010 22.398 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.409 17.214 make_images_data 222 9.7 0.004 0.005 11.728 13.325 hybrid_alltoall_any 227 10.6 0.808 3.795 11.130 12.798 dbcsr_mm_accdrv_process 3641 10.4 0.228 0.403 7.466 8.972 dbcsr_mm_accdrv_process_sort 3641 11.4 7.093 8.563 7.093 8.563 mp_sum_l 807 5.4 4.089 7.389 4.089 7.389 multiply_cannon_sync_h2d 1332 9.7 5.550 6.314 5.550 6.314 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.090 6.035 mp_irecv_dv 3229 10.9 2.068 5.966 2.068 5.966 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.120 5.801 arnoldi_extremal 4 6.8 0.000 0.000 5.178 5.193 arnoldi_normal_ev 4 7.8 0.001 0.004 5.178 5.193 build_subspace 16 8.4 0.014 0.021 4.835 4.847 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.564 4.744 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.423 4.653 calculate_norms 2376 9.8 4.190 4.548 4.190 4.548 mp_allgather_i34 111 8.7 2.152 4.450 2.152 4.450 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.556 3.852 dbcsr_matrix_vector_mult_local 304 10.0 3.179 3.674 3.181 3.676 dbcsr_sort_data 658 11.4 3.055 3.527 3.055 3.527 ls_scf_post 1 4.0 0.000 0.001 3.231 3.237 dbcsr_special_finalize 555 9.7 0.006 0.007 2.796 3.220 dbcsr_merge_single_wm 555 10.7 0.534 0.656 2.787 3.212 ls_scf_dm_to_ks 2 5.0 0.002 0.028 3.041 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.942 3.009 dbcsr_data_release 10477 10.7 1.588 2.455 1.588 2.455 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.092 2.104 rebuild_ks_matrix 3 7.3 0.000 0.000 2.068 2.081 qs_ks_build_kohn_sham_matrix 3 8.3 0.017 0.067 2.068 2.080 dbcsr_finalize 304 7.8 0.049 0.061 1.803 2.060 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.526000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2711.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 4.630122E+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 0.072 0.106 99.565 99.566 qs_energies 1 2.0 0.000 0.000 98.635 98.640 ls_scf 1 3.0 0.000 0.000 96.699 96.708 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.991 78.199 ls_scf_main 1 4.0 0.000 0.000 62.138 62.139 multiply_cannon 111 7.7 0.083 0.160 54.960 59.659 density_matrix_trs4 2 5.0 0.002 0.002 54.868 54.958 multiply_cannon_loop 111 8.7 0.069 0.077 50.552 52.865 mp_waitall_1 5436 11.0 26.625 32.507 26.625 32.507 ls_scf_init_scf 1 4.0 0.000 0.000 30.952 30.957 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.621 29.670 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.435 27.445 multiply_cannon_multrec 444 9.7 13.977 16.737 21.004 23.561 make_m2s 222 7.7 0.004 0.005 17.880 20.274 make_images 222 8.7 3.713 4.395 17.819 20.214 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.890 16.272 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.319 15.040 make_images_data 222 9.7 0.003 0.004 10.217 12.471 hybrid_alltoall_any 227 10.6 0.788 3.755 9.979 12.398 multiply_cannon_sync_h2d 444 9.7 6.522 8.142 6.522 8.142 dbcsr_mm_accdrv_process 3003 10.4 0.165 0.344 6.732 7.876 dbcsr_mm_accdrv_process_sort 3003 11.4 6.410 7.508 6.410 7.508 mp_allgather_i34 111 8.7 2.649 6.975 2.649 6.975 arnoldi_extremal 4 6.8 0.000 0.000 5.817 5.829 arnoldi_normal_ev 4 7.8 0.002 0.005 5.817 5.829 build_subspace 16 8.4 0.015 0.020 5.410 5.424 mp_sum_l 807 5.4 3.225 4.831 3.225 4.831 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.568 4.713 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.172 4.361 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.645 4.347 mp_irecv_dv 1241 11.2 1.626 4.323 1.626 4.323 dbcsr_matrix_vector_mult_local 304 10.0 3.675 4.143 3.677 4.145 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.715 3.795 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.341 3.710 calculate_norms 792 9.8 3.530 3.670 3.530 3.670 ls_scf_post 1 4.0 0.000 0.000 3.608 3.613 make_images_sizes 222 9.7 0.000 0.000 1.026 3.598 mp_alltoall_i44 222 10.7 1.026 3.597 1.026 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.435 dbcsr_finalize 304 7.8 0.062 0.078 2.197 2.256 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.241 2.242 rebuild_ks_matrix 3 7.3 0.000 0.000 2.208 2.210 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.208 2.210 dbcsr_merge_all 275 8.9 0.473 0.520 2.046 2.107 dbcsr_data_release 10123 10.8 1.324 2.016 1.324 2.016 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.566000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3609.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1_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 8.785457E+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 0.081 0.102 108.613 108.613 qs_energies 1 2.0 0.000 0.000 107.194 107.207 ls_scf 1 3.0 0.000 0.000 104.240 104.252 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.659 77.774 ls_scf_main 1 4.0 0.000 0.000 66.287 66.287 density_matrix_trs4 2 5.0 0.002 0.003 56.947 57.007 multiply_cannon 111 7.7 0.141 0.289 50.034 51.988 multiply_cannon_loop 111 8.7 0.068 0.070 46.518 47.500 ls_scf_init_scf 1 4.0 0.000 0.000 34.296 34.296 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.638 32.649 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.799 29.807 mp_waitall_1 4527 11.1 22.366 26.278 22.366 26.278 make_m2s 222 7.7 0.005 0.005 23.932 24.988 make_images 222 8.7 4.570 4.992 23.827 24.880 multiply_cannon_multrec 444 9.7 17.877 18.666 22.485 23.042 hybrid_alltoall_any 227 10.6 1.659 3.624 12.989 15.735 make_images_data 222 9.7 0.003 0.003 13.209 15.722 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.580 11.281 multiply_cannon_sync_h2d 444 9.7 8.847 8.894 8.847 8.894 arnoldi_extremal 4 6.8 0.000 0.000 7.369 7.380 arnoldi_normal_ev 4 7.8 0.003 0.009 7.369 7.379 build_subspace 16 8.4 0.026 0.037 6.794 6.803 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.437 5.580 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.281 5.374 dbcsr_matrix_vector_mult_local 304 10.0 5.002 5.299 5.004 5.301 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.031 5.277 dbcsr_mm_accdrv_process 1814 10.4 0.257 0.327 4.426 4.566 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.246 4.123 4.246 mp_allgather_i34 111 8.7 1.174 3.680 1.174 3.680 ls_scf_post 1 4.0 0.000 0.000 3.658 3.669 make_images_sizes 222 9.7 0.000 0.000 1.423 3.502 mp_alltoall_i44 222 10.7 1.423 3.502 1.423 3.502 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.416 calculate_norms 792 9.8 3.239 3.283 3.239 3.283 dbcsr_finalize 304 7.8 0.082 0.090 3.087 3.184 dbcsr_merge_all 275 8.9 0.891 0.914 2.873 2.965 qs_energies_init_hamiltonians 1 3.0 0.002 0.006 2.924 2.924 dbcsr_complete_redistribute 5 7.6 1.436 1.474 2.776 2.889 dbcsr_data_release 12724 10.6 2.323 2.837 2.323 2.837 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.587 2.588 matrix_ls_to_qs 2 6.0 0.000 0.000 2.442 2.566 rebuild_ks_matrix 3 7.3 0.000 0.000 2.521 2.523 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.521 2.523 dbcsr_sort_data 325 11.1 2.440 2.512 2.440 2.512 dbcsr_new_transposed 4 7.5 0.243 0.251 2.304 2.330 dbcsr_frobenius_norm 74 6.6 2.056 2.132 2.192 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.201 dbcsr_add_anytype 103 7.2 0.859 0.891 2.131 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.613000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6858.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 685d0dbee4de4cc9de9525d9fb9a5bb641aab1e1 Summary: empty Status: OK