=== 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: b261977e9e408dd060ac96be327c7e498359dcdc ################# 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 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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. # # Last update: 21.06.2023 # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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.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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/01 job id: 47389165 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/02 job id: 47389168 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/03 job id: 47389170 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/04 job id: 47389172 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/05 job id: 47389175 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/06 job id: 47389178 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/07 job id: 47389180 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/08 job id: 47389182 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/09 job id: 47389184 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/10 job id: 47389186 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/11 job id: 47389188 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/12 job id: 47389190 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/13 job id: 47389192 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/14 job id: 47389195 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/15 job id: 47389198 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/16 job id: 47389200 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/17 job id: 47389202 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/18 job id: 47389203 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/19 job id: 47389205 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/20 job id: 47389207 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/21 job id: 47389209 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/22 job id: 47389213 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/23 job id: 47389216 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/24 job id: 47389219 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/25 job id: 47389221 --- 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/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/26 job id: 47389224 --- 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 ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/27 job id: 47389228 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ 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]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.037 135.037 135.038 farming_run 1 2.0 134.517 134.519 135.007 135.011 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465995E+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 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.024 115.486 115.486 qs_energies 1 2.0 0.000 0.000 115.263 115.265 mp2_main 1 3.0 0.000 0.000 113.071 113.073 mp2_gpw_main 1 4.0 0.019 0.025 111.972 111.974 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.926 93.296 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.881 55.250 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.173 41.325 46.480 get_2c_integrals 1 6.0 0.008 0.009 36.940 37.874 integrate_v_rspace 273 8.0 0.432 0.448 24.725 29.656 pw_transfer 6555 10.6 0.372 0.396 27.418 28.167 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.094 26.737 grid_integrate_task_list 273 9.0 20.541 25.916 20.541 25.916 fft_wrap_pw1pw2_100 2178 12.4 1.189 1.365 23.614 24.254 compute_2c_integrals 1 7.0 0.002 0.002 19.630 19.632 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.881 19.354 mp2_eri_2c_integrate_gpw 1 9.0 2.392 2.438 18.879 19.351 rpa_ri_compute_en 1 5.0 0.019 0.021 18.937 19.095 cp_fm_cholesky_decompose 12 8.2 17.303 18.228 17.303 18.228 cholesky_decomp 1 7.0 0.000 0.000 16.156 17.083 fft3d_s 5443 13.4 16.167 16.648 16.189 16.670 ao_to_mo_and_store_B_mult_1 272 7.0 10.722 15.329 10.722 15.329 calculate_wavefunction 272 8.0 5.406 5.548 12.531 13.137 rpa_num_int 1 6.0 0.000 0.000 10.676 10.686 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.656 10.685 calc_mat_Q 8 8.0 0.000 0.000 9.463 9.574 contract_S_to_Q 8 9.0 0.000 0.000 8.885 8.995 parallel_gemm_fm 14 9.1 0.000 0.000 8.486 8.587 parallel_gemm_fm_cosma 14 10.1 8.486 8.587 8.486 8.587 calc_potential_gpw 544 9.5 0.005 0.005 8.221 8.568 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.198 8.522 potential_pw2rs 545 10.0 0.107 0.110 7.694 8.385 create_integ_mat 1 6.0 0.014 0.027 7.796 7.806 collocate_single_gaussian 272 10.0 0.039 0.042 7.481 7.778 array2fm 1 7.0 0.000 0.000 6.802 7.231 pw_scatter_s 2720 13.7 4.440 4.621 4.440 4.621 pw_gather_s 2722 13.2 3.890 4.245 3.890 4.245 array2fm_buffer_send 1 8.0 2.968 3.149 2.968 3.149 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.974239, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2811.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 424 9. 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 0.036 394.791 394.793 farming_run 1 2.0 394.085 394.091 394.754 394.756 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.228190E+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 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 9758476. 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 0.009 0.029 208.683 208.684 qs_energies 1 2.0 0.000 0.000 208.461 208.469 scf_env_do_scf 1 3.0 0.000 0.000 106.040 106.040 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.148 105.155 rebuild_ks_matrix 4 6.0 0.000 0.000 105.147 105.154 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.147 105.154 hfx_ks_matrix 4 8.0 0.001 0.001 104.763 104.768 integrate_four_center 4 9.0 0.144 0.454 104.763 104.767 mp2_main 1 3.0 0.000 0.000 102.131 102.139 mp2_gpw_main 1 4.0 0.031 0.044 101.247 101.261 integrate_four_center_main 4 10.0 0.130 0.611 96.504 98.936 integrate_four_center_bin 263 11.0 96.374 98.495 96.374 98.495 init_scf_loop 1 4.0 0.000 0.000 91.868 91.868 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.418 75.572 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.967 55.120 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.157 41.502 46.697 integrate_v_rspace 95 8.0 0.395 0.569 27.888 32.931 pw_transfer 2240 10.6 0.143 0.163 29.959 30.319 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.963 29.319 grid_integrate_task_list 95 9.0 23.177 28.437 23.177 28.437 mp2_ri_gpw_compute_en 1 5.0 0.054 0.065 26.683 28.325 fft_wrap_pw1pw2_100 730 12.4 1.286 1.412 26.666 27.042 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.842 1.901 24.895 24.904 ao_to_mo_and_store_B_mult_1 91 7.0 10.785 24.355 10.785 24.355 get_2c_integrals 1 6.0 0.000 0.000 20.345 20.386 compute_2c_integrals 1 7.0 0.002 0.003 19.325 19.328 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.918 19.194 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.853 18.917 19.193 fft3d_s 1823 13.4 18.462 18.822 18.475 18.835 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.170 14.171 calculate_wavefunction 91 8.0 2.014 2.048 9.736 9.971 potential_pw2rs 186 10.0 0.034 0.035 8.669 9.253 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.595 8.750 9.106 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.281 8.639 local_gemm 172 8.0 8.192 8.555 8.192 8.555 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.516 7.911 8.379 collocate_single_gaussian 91 10.0 0.017 0.020 7.912 8.167 calc_potential_gpw 182 9.5 0.002 0.002 7.920 8.083 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.418 6.346 6.418 mp_sendrecv_dm3 2068 8.0 5.942 6.408 5.942 6.408 pw_gather_s 912 13.2 4.903 5.422 4.903 5.422 mp_sync 37 10.5 2.954 4.826 2.954 4.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.236380, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 452.153344E+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 10329 270. MP_Sync 530 MP_Alltoall 2083 977600. 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 0.069 0.120 60.676 60.677 qs_mol_dyn_low 1 2.0 0.004 0.014 60.222 60.264 qs_forces 11 3.9 0.003 0.008 60.097 60.098 qs_energies 11 4.9 0.001 0.007 58.483 58.497 scf_env_do_scf 11 5.9 0.000 0.001 51.737 51.737 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 49.163 49.164 qs_scf_new_mos 108 7.5 0.000 0.001 37.380 37.658 qs_scf_loop_do_ot 108 8.5 0.001 0.001 37.380 37.658 dbcsr_multiply_generic 2286 12.5 0.095 0.099 36.611 37.176 ot_scf_mini 108 9.5 0.002 0.002 35.557 35.741 velocity_verlet 10 3.0 0.001 0.002 29.514 29.516 multiply_cannon 2286 13.5 0.189 0.194 27.523 29.175 multiply_cannon_loop 2286 14.5 1.485 1.574 26.549 28.208 ot_mini 108 10.5 0.001 0.001 21.473 21.710 qs_ot_get_derivative 108 11.5 0.001 0.001 18.323 18.498 mp_waitall_1 245248 16.5 10.208 16.912 10.208 16.912 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.747 14.339 multiply_cannon_multrec 54864 15.5 4.181 6.416 7.719 11.212 qs_ot_get_p 119 10.4 0.001 0.001 9.140 9.459 rebuild_ks_matrix 119 8.3 0.000 0.000 8.800 8.954 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.014 8.800 8.954 mp_sum_l 7287 12.8 6.336 8.086 6.336 8.086 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.762 7.914 multiply_cannon_sync_h2d 54864 15.5 5.762 7.155 5.762 7.155 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 6.282 6.750 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.142 6.247 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.786 5.849 init_scf_run 11 5.9 0.000 0.001 5.364 5.364 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.363 5.364 qs_rho_update_rho_low 119 7.7 0.001 0.001 5.180 5.299 calculate_rho_elec 119 8.7 0.012 0.017 5.180 5.299 sum_up_and_integrate 119 10.3 0.012 0.015 5.230 5.239 integrate_v_rspace 119 11.3 0.002 0.003 5.218 5.228 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.912 4.913 dbcsr_mm_accdrv_process 76910 16.1 1.132 1.789 3.458 4.866 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.583 4.585 cp_fm_redistribute_end 50 14.0 2.339 4.549 2.350 4.555 cp_fm_diag_elpa_base 50 14.0 2.195 4.413 2.200 4.425 rs_pw_transfer 974 11.9 0.012 0.012 3.669 3.816 multiply_cannon_metrocomm1 54864 15.5 0.053 0.057 2.216 3.766 calculate_dm_sparse 119 9.5 0.000 0.001 3.093 3.237 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.005 3.221 apply_single 119 13.6 0.000 0.000 3.005 3.220 density_rs2pw 119 9.7 0.004 0.005 3.016 3.154 jit_kernel_multiply 13 15.8 2.267 2.995 2.267 2.995 calculate_first_density_matrix 1 7.0 0.001 0.003 2.958 2.962 ot_diis_step 108 11.5 0.006 0.006 2.820 2.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.783 2.786 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.696 2.765 potential_pw2rs 119 12.3 0.004 0.004 2.521 2.567 make_m2s 4572 13.5 0.054 0.056 2.465 2.554 init_scf_loop 11 6.9 0.001 0.004 2.544 2.546 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.500 pw_transfer 1439 11.6 0.053 0.059 2.377 2.474 make_images 4572 14.5 0.133 0.139 2.382 2.470 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.301 2.402 acc_transpose_blocks 54864 15.5 0.228 0.248 1.878 2.393 wfi_extrapolate 11 7.9 0.001 0.001 2.323 2.323 mp_sum_d 4135 12.0 1.563 2.182 1.563 2.182 grid_integrate_task_list 119 12.3 1.991 2.092 1.991 2.092 fft3d_ps 1201 14.6 0.374 0.483 1.945 2.041 mp_alltoall_d11v 2130 13.8 1.690 1.882 1.690 1.882 fft_wrap_pw1pw2_140 487 13.2 0.183 0.202 1.733 1.835 mp_waitany 12084 13.8 1.387 1.527 1.387 1.527 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.454 1.470 grid_collocate_task_list 119 9.7 1.359 1.447 1.359 1.447 make_images_sizes 4572 15.5 0.004 0.004 1.084 1.365 mp_alltoall_i44 4572 16.5 1.080 1.361 1.080 1.361 prepare_preconditioner 11 7.9 0.000 0.000 1.275 1.308 make_preconditioner 11 8.9 0.000 0.001 1.275 1.308 dbcsr_dot_sd 1205 11.9 0.049 0.061 0.856 1.274 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.208 1.252 acc_transpose_blocks_kernels 54864 16.5 0.245 0.380 0.887 1.222 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.677000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 487.460864E+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 10306 303. MP_Sync 54 MP_Alltoall 2060 808431. 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 0.031 0.069 43.944 43.946 qs_mol_dyn_low 1 2.0 0.003 0.005 43.565 43.573 qs_forces 11 3.9 0.004 0.021 43.462 43.464 qs_energies 11 4.9 0.002 0.006 41.595 41.599 scf_env_do_scf 11 5.9 0.001 0.002 35.757 35.758 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 32.757 32.759 dbcsr_multiply_generic 2286 12.5 0.102 0.105 22.876 23.270 qs_scf_new_mos 108 7.5 0.001 0.001 21.718 21.969 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.717 21.969 ot_scf_mini 108 9.5 0.003 0.004 20.776 20.952 velocity_verlet 10 3.0 0.009 0.035 20.731 20.732 multiply_cannon 2286 13.5 0.208 0.217 17.053 18.660 multiply_cannon_loop 2286 14.5 0.909 0.974 15.732 17.206 ot_mini 108 10.5 0.001 0.001 12.662 12.902 mp_waitall_1 200699 16.5 6.743 12.182 6.743 12.182 qs_ot_get_derivative 108 11.5 0.001 0.001 10.099 10.279 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.287 9.838 multiply_cannon_multrec 27432 15.5 1.996 4.518 6.179 9.068 rebuild_ks_matrix 119 8.3 0.000 0.000 8.287 8.430 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 8.287 8.429 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.318 7.449 dbcsr_mm_accdrv_process 47894 16.0 3.118 5.405 4.112 6.050 qs_ot_get_p 119 10.4 0.001 0.001 5.103 5.338 sum_up_and_integrate 119 10.3 0.025 0.028 5.175 5.184 integrate_v_rspace 119 11.3 0.002 0.003 5.150 5.161 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.913 4.957 calculate_rho_elec 119 8.7 0.021 0.024 4.912 4.957 rs_pw_transfer 974 11.9 0.010 0.011 4.194 4.686 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.831 4.684 init_scf_run 11 5.9 0.000 0.001 4.540 4.541 scf_env_initial_rho_setup 11 6.9 0.001 0.003 4.540 4.540 mp_sum_l 7287 12.8 2.276 4.283 2.276 4.283 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.121 4.182 apply_single 119 13.6 0.000 0.000 3.121 4.181 density_rs2pw 119 9.7 0.004 0.005 3.152 3.650 make_m2s 4572 13.5 0.053 0.056 3.125 3.404 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.373 3.392 make_images 4572 14.5 0.200 0.238 3.034 3.312 init_scf_loop 11 6.9 0.001 0.004 2.970 2.973 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.895 2.896 potential_pw2rs 119 12.3 0.006 0.006 2.836 2.871 multiply_cannon_sync_h2d 27432 15.5 2.175 2.869 2.175 2.869 calculate_first_density_matrix 1 7.0 0.001 0.003 2.793 2.795 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.469 2.562 ot_diis_step 108 11.5 0.011 0.012 2.504 2.505 pw_transfer 1439 11.6 0.066 0.070 2.339 2.409 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.400 2.402 cp_fm_redistribute_end 50 14.0 1.217 2.363 1.223 2.367 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.338 2.340 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.246 2.317 cp_fm_diag_elpa_base 50 14.0 1.108 2.248 1.138 2.293 calculate_dm_sparse 119 9.5 0.000 0.001 2.181 2.260 jit_kernel_multiply 9 16.1 0.942 2.151 0.942 2.151 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.060 2.101 grid_integrate_task_list 119 12.3 1.840 1.957 1.840 1.957 prepare_preconditioner 11 7.9 0.000 0.000 1.910 1.936 make_preconditioner 11 8.9 0.000 0.001 1.910 1.936 make_images_data 4572 15.5 0.045 0.051 1.328 1.928 fft3d_ps 1201 14.6 0.522 0.580 1.798 1.875 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.785 1.858 mp_alltoall_d11v 2130 13.8 1.483 1.667 1.483 1.667 wfi_extrapolate 11 7.9 0.001 0.001 1.661 1.661 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.137 1.659 fft_wrap_pw1pw2_140 487 13.2 0.202 0.212 1.582 1.643 mp_allgather_i34 2286 14.5 0.760 1.572 0.760 1.572 acc_transpose_blocks 27432 15.5 0.110 0.115 1.236 1.555 grid_collocate_task_list 119 9.7 1.272 1.430 1.272 1.430 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.359 1.370 make_images_sizes 4572 15.5 0.005 0.005 0.980 1.334 mp_alltoall_i44 4572 16.5 0.975 1.330 0.975 1.330 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.241 1.290 rs_pw_transfer_PW2RS_50 119 14.3 0.589 0.608 1.062 1.208 mp_sendrecv_dv 16779 12.7 1.128 1.167 1.128 1.167 mp_sum_d 4135 12.0 0.675 1.125 0.675 1.125 mp_waitany 5720 13.7 0.605 1.124 0.605 1.124 rs_pw_transfer_RS2PW_140 130 11.5 0.140 0.149 0.621 1.104 mp_alltoall_z22v 1201 16.6 0.944 1.076 0.944 1.076 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.032 1.046 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.002 1.003 rs_pw_transfer_RS2PW_50 119 11.7 0.278 0.294 0.866 0.950 acc_transpose_blocks_kernels 27432 16.5 0.182 0.272 0.691 0.922 dbcsr_complete_redistribute 329 12.2 0.085 0.114 0.813 0.893 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.946000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.818182, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 520.519680E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 706427. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.038 39.486 39.486 qs_mol_dyn_low 1 2.0 0.003 0.003 39.257 39.265 qs_forces 11 3.9 0.002 0.002 39.192 39.196 qs_energies 11 4.9 0.001 0.001 37.456 37.463 scf_env_do_scf 11 5.9 0.000 0.001 31.852 31.853 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 28.464 28.465 dbcsr_multiply_generic 2286 12.5 0.097 0.098 18.847 19.029 velocity_verlet 10 3.0 0.001 0.001 18.885 18.886 qs_scf_new_mos 108 7.5 0.001 0.001 18.214 18.235 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.213 18.235 ot_scf_mini 108 9.5 0.002 0.002 17.367 17.388 multiply_cannon 2286 13.5 0.197 0.205 14.307 15.027 multiply_cannon_loop 2286 14.5 0.641 0.673 13.229 13.926 ot_mini 108 10.5 0.001 0.001 10.205 10.240 qs_ot_get_derivative 108 11.5 0.001 0.001 8.471 8.488 rebuild_ks_matrix 119 8.3 0.000 0.000 7.502 7.561 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.501 7.560 multiply_cannon_multrec 18288 15.5 1.920 2.849 7.184 7.497 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.641 6.695 dbcsr_mm_accdrv_process 38222 16.0 4.156 5.880 5.180 5.998 mp_waitall_1 158411 16.6 4.206 5.062 4.206 5.062 sum_up_and_integrate 119 10.3 0.031 0.031 4.942 4.946 integrate_v_rspace 119 11.3 0.002 0.003 4.912 4.919 qs_ot_get_p 119 10.4 0.001 0.001 4.695 4.743 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.558 4.563 calculate_rho_elec 119 8.7 0.031 0.032 4.557 4.562 init_scf_run 11 5.9 0.000 0.001 4.301 4.301 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.301 4.301 rs_pw_transfer 974 11.9 0.009 0.010 3.830 4.078 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.141 3.781 init_scf_loop 11 6.9 0.000 0.000 3.360 3.361 density_rs2pw 119 9.7 0.004 0.004 3.006 3.272 qs_ot_p2m_diag 50 11.0 0.012 0.013 3.140 3.146 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.875 2.894 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.790 2.790 calculate_first_density_matrix 1 7.0 0.000 0.000 2.762 2.763 make_m2s 4572 13.5 0.044 0.045 2.540 2.742 potential_pw2rs 119 12.3 0.007 0.009 2.654 2.674 make_images 4572 14.5 0.191 0.204 2.454 2.657 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.161 2.508 apply_single 119 13.6 0.000 0.000 2.161 2.508 prepare_preconditioner 11 7.9 0.000 0.000 2.466 2.472 make_preconditioner 11 8.9 0.000 0.000 2.466 2.472 pw_transfer 1439 11.6 0.066 0.069 2.371 2.394 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.301 2.391 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.354 2.364 cp_fm_diag_elpa_base 50 14.0 2.322 2.338 2.352 2.362 jit_kernel_multiply 10 15.8 0.973 2.320 0.973 2.320 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.278 2.299 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.149 2.151 calculate_dm_sparse 119 9.5 0.000 0.000 2.094 2.119 mp_sum_l 7287 12.8 1.612 2.030 1.612 2.030 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.973 1.984 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.940 1.956 grid_integrate_task_list 119 12.3 1.792 1.889 1.792 1.889 fft3d_ps 1201 14.6 0.531 0.550 1.768 1.791 ot_diis_step 108 11.5 0.011 0.011 1.659 1.659 fft_wrap_pw1pw2_140 487 13.2 0.251 0.257 1.568 1.586 multiply_cannon_sync_h2d 18288 15.5 1.279 1.489 1.279 1.489 wfi_extrapolate 11 7.9 0.001 0.001 1.475 1.475 make_images_data 4572 15.5 0.045 0.049 1.148 1.420 grid_collocate_task_list 119 9.7 1.245 1.381 1.245 1.381 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.356 1.363 acc_transpose_blocks 18288 15.5 0.076 0.079 1.339 1.361 cp_fm_cholesky_invert 11 10.9 1.175 1.182 1.175 1.182 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.961 1.157 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.101 1.152 mp_alltoall_z22v 1201 16.6 1.029 1.086 1.029 1.086 mp_alltoall_d11v 2130 13.8 0.949 1.076 0.949 1.076 mp_sendrecv_dv 11067 12.7 0.997 1.014 0.997 1.014 make_images_sizes 4572 15.5 0.005 0.005 0.707 1.001 mp_allgather_i34 2286 14.5 0.537 0.999 0.537 0.999 mp_alltoall_i44 4572 16.5 0.702 0.996 0.702 0.996 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.977 0.980 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.937 0.944 rs_pw_transfer_PW2RS_50 119 14.3 0.401 0.412 0.865 0.909 acc_transpose_blocks_kernels 18288 16.5 0.212 0.220 0.888 0.899 mp_waitany 9880 13.7 0.628 0.896 0.628 0.896 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.581 0.841 rs_pw_transfer_RS2PW_50 119 11.7 0.243 0.251 0.780 0.804 dbcsr_complete_redistribute 329 12.2 0.093 0.104 0.686 0.801 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=39.486000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.363636, yerr=1.822722 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 554.823680E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 37.424 37.424 qs_mol_dyn_low 1 2.0 0.003 0.003 36.725 36.733 qs_forces 11 3.9 0.002 0.003 36.670 36.671 qs_energies 11 4.9 0.001 0.001 34.901 34.906 scf_env_do_scf 11 5.9 0.000 0.001 29.764 29.765 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 26.122 26.122 dbcsr_multiply_generic 2286 12.5 0.098 0.102 18.716 18.894 velocity_verlet 10 3.0 0.001 0.001 18.719 18.721 qs_scf_new_mos 108 7.5 0.001 0.001 17.168 17.223 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.167 17.222 ot_scf_mini 108 9.5 0.002 0.003 16.218 16.272 multiply_cannon 2286 13.5 0.224 0.262 14.777 15.308 multiply_cannon_loop 2286 14.5 0.948 0.981 13.779 14.280 ot_mini 108 10.5 0.001 0.001 9.833 9.899 multiply_cannon_multrec 27432 15.5 2.381 3.266 8.635 9.113 qs_ot_get_derivative 108 11.5 0.001 0.001 7.999 8.053 dbcsr_mm_accdrv_process 47916 15.9 5.213 7.128 6.162 7.349 rebuild_ks_matrix 119 8.3 0.000 0.000 6.875 6.934 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.874 6.933 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.098 6.149 sum_up_and_integrate 119 10.3 0.035 0.037 4.056 4.064 integrate_v_rspace 119 11.3 0.003 0.003 4.021 4.029 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.836 3.868 calculate_rho_elec 119 8.7 0.040 0.046 3.836 3.868 qs_ot_get_p 119 10.4 0.001 0.001 3.730 3.801 init_scf_run 11 5.9 0.000 0.001 3.798 3.798 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.797 3.798 init_scf_loop 11 6.9 0.000 0.000 3.620 3.620 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.818 3.293 mp_waitall_1 137007 16.6 2.238 2.882 2.238 2.882 prepare_preconditioner 11 7.9 0.000 0.000 2.746 2.755 make_preconditioner 11 8.9 0.000 0.000 2.746 2.755 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.344 2.676 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.126 2.603 apply_single 119 13.6 0.000 0.000 2.126 2.603 make_m2s 4572 13.5 0.054 0.055 2.392 2.516 rs_pw_transfer 974 11.9 0.009 0.009 2.328 2.507 make_images 4572 14.5 0.271 0.335 2.285 2.408 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.369 2.377 density_rs2pw 119 9.7 0.004 0.004 2.189 2.376 calculate_first_density_matrix 1 7.0 0.000 0.000 2.326 2.327 pw_transfer 1439 11.6 0.066 0.070 2.152 2.188 calculate_dm_sparse 119 9.5 0.000 0.000 2.041 2.110 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.059 2.099 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.012 2.039 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.035 2.035 grid_integrate_task_list 119 12.3 1.829 1.894 1.829 1.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.871 1.873 jit_kernel_multiply 10 15.8 0.889 1.815 0.889 1.815 ot_diis_step 108 11.5 0.012 0.012 1.792 1.793 potential_pw2rs 119 12.3 0.009 0.010 1.781 1.786 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.761 1.769 cp_fm_diag_elpa_base 50 14.0 1.724 1.741 1.758 1.766 fft_wrap_pw1pw2_140 487 13.2 0.289 0.304 1.670 1.712 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.690 1.702 mp_sum_l 7287 12.8 1.124 1.679 1.124 1.679 fft3d_ps 1201 14.6 0.558 0.608 1.497 1.521 acc_transpose_blocks 27432 15.5 0.111 0.113 1.487 1.516 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 0.895 1.503 wfi_extrapolate 11 7.9 0.001 0.001 1.408 1.408 grid_collocate_task_list 119 9.7 1.249 1.364 1.249 1.364 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.217 1.229 cp_fm_upper_to_full 72 14.2 0.835 1.189 0.835 1.189 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.128 1.144 dbcsr_complete_redistribute 329 12.2 0.118 0.143 0.809 1.086 make_images_data 4572 15.5 0.045 0.048 0.945 1.082 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.065 1.066 multiply_cannon_sync_h2d 27432 15.5 0.990 1.034 0.990 1.034 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.814 1.031 mp_alltoall_d11v 2130 13.8 0.821 0.932 0.821 0.932 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.870 0.876 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.870 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 0.850 0.868 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.585 0.856 cp_fm_cholesky_invert 11 10.9 0.852 0.855 0.852 0.855 mp_alltoall_z22v 1201 16.6 0.823 0.848 0.823 0.848 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.283 0.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.424000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=2.248967 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 601.911296E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 30.439 30.439 qs_mol_dyn_low 1 2.0 0.003 0.003 30.194 30.202 qs_forces 11 3.9 0.002 0.002 30.081 30.082 qs_energies 11 4.9 0.001 0.001 28.316 28.319 scf_env_do_scf 11 5.9 0.000 0.001 22.951 22.951 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 20.314 20.315 velocity_verlet 10 3.0 0.001 0.001 15.447 15.453 dbcsr_multiply_generic 2286 12.5 0.094 0.099 13.068 13.156 qs_scf_new_mos 108 7.5 0.001 0.001 11.706 11.732 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.706 11.731 ot_scf_mini 108 9.5 0.002 0.002 10.991 11.021 multiply_cannon 2286 13.5 0.229 0.238 10.187 10.717 multiply_cannon_loop 2286 14.5 0.332 0.343 9.246 9.460 rebuild_ks_matrix 119 8.3 0.000 0.000 6.349 6.376 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.349 6.376 ot_mini 108 10.5 0.001 0.001 6.045 6.079 multiply_cannon_multrec 9144 15.5 1.617 1.862 5.819 6.029 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.659 5.684 qs_ot_get_derivative 108 11.5 0.001 0.001 4.716 4.746 dbcsr_mm_accdrv_process 12550 15.8 3.402 4.133 4.101 4.204 sum_up_and_integrate 119 10.3 0.038 0.042 3.974 3.980 integrate_v_rspace 119 11.3 0.003 0.003 3.936 3.943 init_scf_run 11 5.9 0.000 0.001 3.907 3.907 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.907 3.907 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.831 3.840 calculate_rho_elec 119 8.7 0.060 0.061 3.831 3.839 qs_ot_get_p 119 10.4 0.001 0.001 3.092 3.129 calculate_first_density_matrix 1 7.0 0.000 0.000 2.673 2.673 init_scf_loop 11 6.9 0.000 0.000 2.612 2.614 mp_waitall_1 115863 16.7 1.843 2.299 1.843 2.299 pw_transfer 1439 11.6 0.066 0.070 2.269 2.282 density_rs2pw 119 9.7 0.004 0.004 2.099 2.242 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.175 2.189 rs_pw_transfer 974 11.9 0.008 0.008 2.023 2.153 make_m2s 4572 13.5 0.034 0.035 1.959 2.113 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.034 2.035 make_images 4572 14.5 0.267 0.302 1.869 2.021 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.012 2.015 grid_integrate_task_list 119 12.3 1.856 1.920 1.856 1.920 prepare_preconditioner 11 7.9 0.000 0.000 1.868 1.873 make_preconditioner 11 8.9 0.000 0.000 1.868 1.873 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.809 1.823 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.802 1.802 calculate_dm_sparse 119 9.5 0.000 0.000 1.783 1.799 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.753 1.783 fft_wrap_pw1pw2_140 487 13.2 0.365 0.374 1.706 1.721 potential_pw2rs 119 12.3 0.010 0.011 1.668 1.674 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.644 1.665 jit_kernel_multiply 9 15.8 0.661 1.637 0.661 1.637 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.536 1.544 cp_fm_diag_elpa_base 50 14.0 1.506 1.525 1.534 1.542 fft3d_ps 1201 14.6 0.563 0.575 1.509 1.521 grid_collocate_task_list 119 9.7 1.299 1.422 1.299 1.422 acc_transpose_blocks 9144 15.5 0.038 0.039 1.308 1.319 ot_diis_step 108 11.5 0.012 0.012 1.308 1.309 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.259 1.274 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.238 1.240 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.177 1.208 apply_single 119 13.6 0.000 0.000 1.176 1.207 wfi_extrapolate 11 7.9 0.001 0.001 1.179 1.179 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.163 1.168 make_images_data 4572 15.5 0.039 0.042 0.912 1.110 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 1.063 1.071 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.822 1.070 cp_fm_cholesky_invert 11 10.9 0.955 0.958 0.955 0.958 jit_kernel_transpose 5 15.6 0.946 0.953 0.946 0.953 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.866 0.914 mp_alltoall_d11v 2130 13.8 0.815 0.873 0.815 0.873 mp_alltoall_z22v 1201 16.6 0.820 0.854 0.820 0.854 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.818 0.821 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.770 0.779 multiply_cannon_sync_h2d 9144 15.5 0.703 0.774 0.703 0.774 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.692 0.739 mp_allgather_i34 2286 14.5 0.271 0.705 0.271 0.705 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.493 0.614 mp_waitany 5200 13.7 0.492 0.612 0.492 0.612 dbcsr_complete_redistribute 329 12.2 0.169 0.213 0.584 0.612 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.439000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.181818, yerr=4.839575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 762.322944E+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 10154 429. 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 0.049 0.061 43.688 43.689 qs_mol_dyn_low 1 2.0 0.003 0.003 43.299 43.306 qs_forces 11 3.9 0.002 0.002 43.241 43.242 qs_energies 11 4.9 0.001 0.001 41.194 41.198 scf_env_do_scf 11 5.9 0.001 0.001 35.094 35.094 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.831 26.832 velocity_verlet 10 3.0 0.001 0.001 24.473 24.478 dbcsr_multiply_generic 2286 12.5 0.102 0.103 18.205 18.399 qs_scf_new_mos 108 7.5 0.001 0.001 16.689 16.784 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.689 16.783 ot_scf_mini 108 9.5 0.002 0.002 15.567 15.670 multiply_cannon 2286 13.5 0.304 0.315 14.056 15.048 multiply_cannon_loop 2286 14.5 0.347 0.352 12.715 13.681 ot_mini 108 10.5 0.001 0.001 9.254 9.376 multiply_cannon_multrec 9144 15.5 3.513 5.076 8.731 8.853 init_scf_loop 11 6.9 0.000 0.000 8.233 8.235 rebuild_ks_matrix 119 8.3 0.000 0.000 7.543 7.696 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.542 7.695 qs_ot_get_derivative 108 11.5 0.001 0.001 7.167 7.270 prepare_preconditioner 11 7.9 0.000 0.000 7.217 7.233 make_preconditioner 11 8.9 0.000 0.000 7.217 7.233 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.770 7.105 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.806 6.943 dbcsr_mm_accdrv_process 12550 15.8 4.183 5.695 5.092 6.497 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.520 4.587 cp_fm_upper_to_full 72 14.2 3.204 4.586 3.204 4.586 calculate_rho_elec 119 8.7 0.119 0.122 4.520 4.586 sum_up_and_integrate 119 10.3 0.064 0.066 4.170 4.178 integrate_v_rspace 119 11.3 0.003 0.004 4.106 4.113 init_scf_run 11 5.9 0.000 0.001 4.063 4.063 scf_env_initial_rho_setup 11 6.9 0.008 0.008 4.063 4.063 mp_waitall_1 94719 16.7 2.742 3.789 2.742 3.789 qs_ot_get_p 119 10.4 0.001 0.001 3.482 3.619 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.613 3.021 dbcsr_complete_redistribute 329 12.2 0.286 0.292 2.040 2.899 pw_transfer 1439 11.6 0.068 0.069 2.809 2.814 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.710 2.715 make_m2s 4572 13.5 0.038 0.038 2.478 2.705 make_images 4572 14.5 0.349 0.379 2.357 2.583 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.698 2.538 calculate_first_density_matrix 1 7.0 0.000 0.000 2.443 2.511 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.229 2.507 apply_single 119 13.6 0.000 0.000 2.229 2.507 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.481 2.414 calculate_dm_sparse 119 9.5 0.000 0.000 2.323 2.377 density_rs2pw 119 9.7 0.004 0.004 2.334 2.359 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.442 2.274 fft_wrap_pw1pw2_140 487 13.2 0.619 0.623 2.261 2.268 mp_alltoall_i22 627 13.8 1.429 2.235 1.429 2.235 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.131 2.134 grid_integrate_task_list 119 12.3 2.069 2.089 2.069 2.089 ot_diis_step 108 11.5 0.014 0.014 2.055 2.056 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.883 1.884 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.798 1.851 mp_sum_l 7287 12.8 1.127 1.849 1.127 1.849 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.838 1.840 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.760 1.761 fft3d_ps 1201 14.6 0.599 0.609 1.725 1.731 cp_fm_cholesky_invert 11 10.9 1.699 1.703 1.699 1.703 rs_pw_transfer 974 11.9 0.009 0.009 1.648 1.673 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.608 1.647 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.239 1.539 potential_pw2rs 119 12.3 0.014 0.014 1.529 1.533 grid_collocate_task_list 119 9.7 1.514 1.525 1.514 1.525 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.515 1.515 cp_fm_diag_elpa_base 50 14.0 1.366 1.420 1.513 1.513 make_images_data 4572 15.5 0.042 0.045 1.214 1.510 wfi_extrapolate 11 7.9 0.001 0.001 1.468 1.468 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.351 1.412 mp_alltoall_d11v 2130 13.8 1.246 1.315 1.246 1.315 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.192 1.216 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.096 1.113 jit_kernel_multiply 6 15.6 0.881 1.081 0.881 1.081 multiply_cannon_sync_h2d 9144 15.5 1.044 1.047 1.044 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.982 1.033 mp_alltoall_z22v 1201 16.6 0.990 1.009 0.990 1.009 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.952 0.967 acc_transpose_blocks 9144 15.5 0.039 0.039 0.947 0.957 qs_create_task_list 11 7.9 0.001 0.001 0.939 0.952 generate_qs_task_list 11 8.9 0.368 0.388 0.938 0.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.689000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.272727, yerr=13.843876 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.522432E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1798545. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.067 83.945 83.946 qs_mol_dyn_low 1 2.0 0.003 0.004 83.511 83.520 qs_forces 11 3.9 0.003 0.005 83.442 83.443 qs_energies 11 4.9 0.003 0.020 80.502 80.515 scf_env_do_scf 11 5.9 0.001 0.001 71.280 71.283 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 65.622 65.623 dbcsr_multiply_generic 2055 12.4 0.107 0.111 51.262 51.483 qs_scf_new_mos 99 7.5 0.001 0.001 47.763 47.876 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.762 47.876 ot_scf_mini 99 9.5 0.002 0.003 45.338 45.424 velocity_verlet 10 3.0 0.001 0.001 43.997 43.998 multiply_cannon 2055 13.4 0.179 0.185 42.412 43.358 multiply_cannon_loop 2055 14.4 1.519 1.548 41.420 42.410 ot_mini 99 10.5 0.001 0.001 26.676 26.749 qs_ot_get_derivative 99 11.5 0.001 0.001 19.931 20.000 multiply_cannon_multrec 49320 15.4 12.395 13.097 17.248 17.901 rebuild_ks_matrix 110 8.3 0.000 0.000 14.860 14.980 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.859 14.980 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.985 13.086 mp_waitall_1 220248 16.4 11.154 11.928 11.154 11.928 multiply_cannon_sync_h2d 49320 15.4 10.275 10.877 10.275 10.877 qs_ot_get_p 110 10.4 0.001 0.001 10.119 10.221 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.745 8.277 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.612 7.760 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.202 7.677 apply_single 110 13.6 0.000 0.001 7.201 7.677 sum_up_and_integrate 110 10.3 0.036 0.043 7.338 7.352 integrate_v_rspace 110 11.3 0.003 0.004 7.302 7.325 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.944 7.081 calculate_rho_elec 110 8.6 0.021 0.026 6.944 7.080 init_scf_run 11 5.9 0.000 0.001 7.021 7.022 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.021 7.021 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.870 6.907 ot_diis_step 99 11.5 0.005 0.005 6.553 6.553 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.937 5.937 init_scf_loop 11 6.9 0.001 0.004 5.623 5.624 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.485 5.532 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.363 5.399 cp_fm_diag_elpa_base 48 14.0 5.350 5.386 5.361 5.397 mp_sum_l 6594 12.7 4.315 4.980 4.315 4.980 dbcsr_mm_accdrv_process 87628 16.1 1.908 1.975 4.727 4.956 rs_pw_transfer 902 11.9 0.012 0.013 3.970 4.603 density_rs2pw 110 9.6 0.004 0.005 3.640 4.220 make_m2s 4110 13.4 0.061 0.066 4.075 4.194 wfi_extrapolate 11 7.9 0.001 0.001 4.181 4.181 make_images 4110 14.4 0.177 0.189 3.979 4.101 calculate_dm_sparse 110 9.5 0.001 0.001 3.816 3.916 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.519 3.681 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.627 3.630 prepare_preconditioner 11 7.9 0.000 0.000 3.438 3.455 make_preconditioner 11 8.9 0.000 0.002 3.438 3.455 grid_integrate_task_list 110 12.3 3.277 3.432 3.277 3.432 pw_transfer 1331 11.6 0.055 0.066 3.327 3.418 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.281 3.364 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.238 3.331 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.245 3.298 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.216 3.271 potential_pw2rs 110 12.3 0.006 0.007 2.803 2.835 fft_wrap_pw1pw2_140 451 13.1 0.457 0.499 2.716 2.815 calculate_first_density_matrix 1 7.0 0.000 0.002 2.735 2.741 mp_alltoall_d11v 2046 13.8 2.163 2.627 2.163 2.627 jit_kernel_multiply 13 15.9 2.544 2.575 2.544 2.575 mp_waitany 14300 13.8 1.893 2.543 1.893 2.543 fft3d_ps 1111 14.6 0.783 0.872 2.426 2.499 grid_collocate_task_list 110 9.6 2.156 2.403 2.156 2.403 acc_transpose_blocks 49320 15.4 0.209 0.217 2.109 2.173 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.008 2.025 make_images_data 4110 15.4 0.043 0.046 1.849 1.976 mp_sum_d 3889 11.9 1.374 1.924 1.374 1.924 hybrid_alltoall_any 4261 16.3 0.083 0.480 1.601 1.865 cp_fm_cholesky_invert 11 10.9 1.853 1.857 1.853 1.857 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.678 1.700 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.946000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.923261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 590.372864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1723921. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.060 74.891 74.892 qs_mol_dyn_low 1 2.0 0.007 0.034 74.414 74.425 qs_forces 11 3.9 0.005 0.018 74.339 74.344 qs_energies 11 4.9 0.004 0.028 70.840 70.849 scf_env_do_scf 11 5.9 0.000 0.001 61.643 61.647 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 52.939 52.940 velocity_verlet 10 3.0 0.001 0.001 39.006 39.007 dbcsr_multiply_generic 2055 12.4 0.117 0.121 38.801 38.977 qs_scf_new_mos 99 7.5 0.001 0.001 35.295 35.401 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.294 35.400 ot_scf_mini 99 9.5 0.003 0.003 33.556 33.658 multiply_cannon 2055 13.4 0.221 0.240 31.337 32.599 multiply_cannon_loop 2055 14.4 0.927 0.950 29.860 30.858 ot_mini 99 10.5 0.001 0.001 19.174 19.282 multiply_cannon_multrec 24660 15.4 7.632 9.050 13.831 15.279 rebuild_ks_matrix 110 8.3 0.000 0.000 14.614 14.765 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 14.613 14.765 qs_ot_get_derivative 99 11.5 0.001 0.002 13.270 13.366 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.828 12.961 mp_waitall_1 176588 16.5 8.472 10.839 8.472 10.839 init_scf_loop 11 6.9 0.000 0.001 8.654 8.656 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.279 7.815 multiply_cannon_sync_h2d 24660 15.4 6.938 7.805 6.938 7.805 qs_ot_get_p 110 10.4 0.001 0.001 7.615 7.762 sum_up_and_integrate 110 10.3 0.053 0.060 7.395 7.408 integrate_v_rspace 110 11.3 0.002 0.003 7.342 7.354 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.600 7.341 apply_single 110 13.6 0.000 0.001 6.600 7.341 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.929 6.941 calculate_rho_elec 110 8.6 0.040 0.048 6.929 6.940 init_scf_run 11 5.9 0.000 0.001 6.667 6.667 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.666 6.667 prepare_preconditioner 11 7.9 0.000 0.000 6.547 6.562 make_preconditioner 11 8.9 0.000 0.000 6.547 6.562 dbcsr_mm_accdrv_process 52282 16.1 4.512 5.405 6.039 6.405 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.086 6.263 ot_diis_step 99 11.5 0.010 0.012 5.849 5.849 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.944 5.656 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.480 5.505 make_m2s 4110 13.4 0.057 0.060 4.808 5.310 make_images 4110 14.4 0.402 0.446 4.699 5.197 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.007 5.008 rs_pw_transfer 902 11.9 0.012 0.013 3.954 4.647 density_rs2pw 110 9.6 0.004 0.005 3.989 4.644 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.310 4.322 cp_fm_diag_elpa_base 48 14.0 4.251 4.280 4.307 4.318 pw_transfer 1331 11.6 0.067 0.077 4.128 4.288 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.020 4.181 wfi_extrapolate 11 7.9 0.001 0.002 3.794 3.795 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.517 3.571 grid_integrate_task_list 110 12.3 3.150 3.415 3.150 3.415 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.338 3.342 fft_wrap_pw1pw2_140 451 13.1 0.517 0.536 3.169 3.326 fft3d_ps 1111 14.6 1.112 1.335 2.997 3.169 make_images_data 4110 15.4 0.047 0.050 2.628 3.163 potential_pw2rs 110 12.3 0.008 0.009 3.107 3.140 calculate_dm_sparse 110 9.5 0.001 0.001 3.101 3.129 cp_fm_cholesky_invert 11 10.9 3.077 3.085 3.077 3.085 hybrid_alltoall_any 4261 16.3 0.103 0.444 2.278 3.055 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.938 2.991 mp_sum_l 6594 12.7 2.056 2.796 2.056 2.796 calculate_first_density_matrix 1 7.0 0.001 0.005 2.766 2.769 mp_alltoall_d11v 2046 13.8 2.010 2.540 2.010 2.540 grid_collocate_task_list 110 9.6 2.098 2.540 2.098 2.540 jit_kernel_multiply 11 16.3 1.178 2.097 1.178 2.097 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.051 2.081 mp_waitany 10164 13.8 1.390 2.043 1.390 2.043 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.945 1.961 qs_energies_init_hamiltonians 11 5.9 0.002 0.006 1.947 1.952 mp_allgather_i34 2055 14.4 0.802 1.871 0.802 1.871 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.802 1.863 cp_fm_cholesky_decompose 22 10.9 1.761 1.768 1.761 1.768 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.216 1.057 1.716 mp_alltoall_z22v 1111 16.6 1.593 1.686 1.593 1.686 mp_irecv_dv 57340 16.2 0.675 1.648 0.675 1.648 dbcsr_complete_redistribute 325 12.2 0.232 0.288 1.331 1.624 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.601 1.611 acc_transpose_blocks 24660 15.4 0.112 0.116 1.583 1.607 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=74.892000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.909091, yerr=6.734640 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/11/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 404.681598E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 3346752 0.0% 0.0% 100.0% average stack size 0.0 0.0 294.1 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 661.520384E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.079 67.263 67.264 qs_mol_dyn_low 1 2.0 0.003 0.003 66.717 66.727 qs_forces 11 3.9 0.003 0.003 66.653 66.654 qs_energies 11 4.9 0.001 0.001 63.316 63.328 scf_env_do_scf 11 5.9 0.000 0.001 54.934 54.934 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 44.934 44.935 velocity_verlet 10 3.0 0.001 0.001 36.247 36.248 dbcsr_multiply_generic 2055 12.4 0.110 0.116 30.471 31.113 qs_scf_new_mos 99 7.5 0.001 0.001 28.600 28.707 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.599 28.706 ot_scf_mini 99 9.5 0.002 0.003 27.256 27.378 multiply_cannon 2055 13.4 0.213 0.225 22.733 24.627 multiply_cannon_loop 2055 14.4 0.621 0.637 21.286 23.166 ot_mini 99 10.5 0.001 0.001 15.071 15.208 rebuild_ks_matrix 110 8.3 0.000 0.000 13.139 13.399 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.138 13.399 mp_waitall_1 139946 16.5 8.104 12.270 8.104 12.270 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.572 11.815 qs_ot_get_derivative 99 11.5 0.001 0.001 10.474 10.594 multiply_cannon_multrec 16440 15.4 3.872 4.815 9.575 10.472 init_scf_loop 11 6.9 0.000 0.000 9.959 9.960 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.636 8.710 prepare_preconditioner 11 7.9 0.000 0.000 8.105 8.126 make_preconditioner 11 8.9 0.000 0.000 8.105 8.126 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.375 7.770 qs_ot_get_p 110 10.4 0.001 0.001 7.069 7.248 sum_up_and_integrate 110 10.3 0.061 0.061 7.129 7.145 integrate_v_rspace 110 11.3 0.003 0.003 7.068 7.085 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.734 6.755 calculate_rho_elec 110 8.6 0.059 0.059 6.733 6.754 init_scf_run 11 5.9 0.000 0.001 5.913 5.913 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.913 5.913 dbcsr_mm_accdrv_process 34862 16.1 4.672 5.252 5.560 5.736 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.999 5.413 apply_single 110 13.6 0.000 0.000 4.999 5.413 make_m2s 4110 13.4 0.050 0.051 4.867 5.208 make_images 4110 14.4 0.394 0.515 4.751 5.094 qs_ot_p2m_diag 48 11.0 0.042 0.044 5.025 5.039 density_rs2pw 110 9.6 0.004 0.005 3.695 4.967 rs_pw_transfer 902 11.9 0.010 0.011 3.438 4.693 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.596 4.597 ot_diis_step 99 11.5 0.010 0.011 4.511 4.511 multiply_cannon_sync_h2d 16440 15.4 3.685 4.236 3.685 4.236 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.525 4.145 pw_transfer 1331 11.6 0.066 0.076 4.035 4.048 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.928 3.942 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.874 3.883 cp_fm_diag_elpa_base 48 14.0 3.796 3.831 3.872 3.881 cp_fm_cholesky_invert 11 10.9 3.432 3.441 3.432 3.441 grid_integrate_task_list 110 12.3 3.192 3.432 3.192 3.432 wfi_extrapolate 11 7.9 0.001 0.001 3.318 3.318 make_images_data 4110 15.4 0.043 0.047 2.773 3.302 fft_wrap_pw1pw2_140 451 13.1 0.634 0.644 3.229 3.247 hybrid_alltoall_any 4261 16.3 0.106 0.380 2.457 3.215 mp_sum_l 6594 12.7 2.265 3.116 2.265 3.116 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.920 2.990 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.927 2.929 potential_pw2rs 110 12.3 0.011 0.012 2.791 2.802 fft3d_ps 1111 14.6 1.091 1.099 2.753 2.765 calculate_dm_sparse 110 9.5 0.001 0.001 2.629 2.663 mp_waitany 17072 13.8 1.278 2.598 1.278 2.598 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.486 2.578 calculate_first_density_matrix 1 7.0 0.000 0.000 2.498 2.499 grid_collocate_task_list 110 9.6 2.118 2.498 2.118 2.498 multiply_cannon_metrocomm4 14385 15.4 0.046 0.050 0.886 2.496 mp_alltoall_d11v 2046 13.8 2.120 2.427 2.120 2.427 mp_irecv_dv 48980 15.7 0.816 2.368 0.816 2.368 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.180 0.948 2.212 dbcsr_complete_redistribute 325 12.2 0.318 0.375 1.659 2.147 cp_fm_cholesky_decompose 22 10.9 2.042 2.062 2.042 2.062 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.978 1.978 mp_allgather_i34 2055 14.4 0.769 1.931 0.769 1.931 cp_fm_upper_to_full 70 14.2 1.425 1.847 1.425 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.764 1.786 jit_kernel_multiply 9 16.4 0.500 1.721 0.500 1.721 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.120 1.600 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.474 1.517 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.483 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.363 1.480 mp_alltoall_z22v 1111 16.6 1.451 1.474 1.451 1.474 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=67.264000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.818182, yerr=7.894972 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 735.809536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.130 72.736 72.737 qs_mol_dyn_low 1 2.0 0.004 0.014 72.177 72.186 qs_forces 11 3.9 0.003 0.003 72.090 72.092 qs_energies 11 4.9 0.003 0.018 68.533 68.538 scf_env_do_scf 11 5.9 0.001 0.001 59.703 59.705 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 46.108 46.110 velocity_verlet 10 3.0 0.001 0.001 41.005 41.009 dbcsr_multiply_generic 2055 12.4 0.114 0.120 31.116 31.402 qs_scf_new_mos 99 7.5 0.001 0.001 29.841 29.944 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.841 29.943 ot_scf_mini 99 9.5 0.003 0.004 28.117 28.222 multiply_cannon 2055 13.4 0.245 0.263 23.013 24.255 multiply_cannon_loop 2055 14.4 0.892 0.915 21.242 21.999 ot_mini 99 10.5 0.001 0.001 15.395 15.515 multiply_cannon_multrec 24660 15.4 4.226 6.849 12.648 13.851 init_scf_loop 11 6.9 0.001 0.006 13.538 13.540 rebuild_ks_matrix 110 8.3 0.000 0.000 12.858 12.965 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.858 12.965 prepare_preconditioner 11 7.9 0.000 0.000 11.711 11.726 make_preconditioner 11 8.9 0.001 0.006 11.711 11.725 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.350 11.445 make_full_inverse_cholesky 11 9.9 0.000 0.001 9.855 11.395 qs_ot_get_derivative 99 11.5 0.001 0.001 11.072 11.178 dbcsr_mm_accdrv_process 52304 16.0 6.884 8.352 8.274 9.214 mp_waitall_1 121746 16.5 5.230 7.855 5.230 7.855 qs_ot_get_p 110 10.4 0.001 0.001 7.451 7.591 sum_up_and_integrate 110 10.3 0.067 0.071 7.100 7.114 integrate_v_rspace 110 11.3 0.003 0.003 7.033 7.047 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.873 6.886 calculate_rho_elec 110 8.6 0.078 0.082 6.872 6.885 make_m2s 4110 13.4 0.059 0.062 6.138 6.730 make_images 4110 14.4 0.578 0.702 5.995 6.588 init_scf_run 11 5.9 0.000 0.001 5.933 5.933 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.933 5.933 qs_ot_p2m_diag 48 11.0 0.055 0.065 5.341 5.357 cp_fm_upper_to_full 70 14.2 3.504 5.095 3.504 5.095 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.800 4.801 ot_diis_step 99 11.5 0.011 0.011 4.258 4.259 density_rs2pw 110 9.6 0.004 0.004 3.731 4.253 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.150 4.233 apply_single 110 13.6 0.000 0.000 4.149 4.232 pw_transfer 1331 11.6 0.066 0.077 4.090 4.134 dbcsr_complete_redistribute 325 12.2 0.407 0.453 2.929 4.132 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.064 4.073 cp_fm_diag_elpa_base 48 14.0 3.889 3.957 4.061 4.069 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.983 4.032 rs_pw_transfer 902 11.9 0.010 0.011 3.300 3.835 cp_fm_cholesky_invert 11 10.9 3.821 3.833 3.821 3.833 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.535 3.821 make_images_data 4110 15.4 0.046 0.049 3.117 3.820 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.660 3.714 hybrid_alltoall_any 4261 16.3 0.121 0.458 2.665 3.705 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.416 3.470 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.278 3.467 grid_integrate_task_list 110 12.3 3.257 3.437 3.257 3.437 multiply_cannon_sync_h2d 24660 15.4 3.177 3.389 3.177 3.389 wfi_extrapolate 11 7.9 0.001 0.001 3.315 3.315 fft_wrap_pw1pw2_140 451 13.1 0.665 0.686 3.219 3.269 calculate_dm_sparse 110 9.5 0.001 0.001 3.038 3.077 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.846 3.016 mp_alltoall_i22 605 13.7 1.775 3.010 1.775 3.010 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.878 2.880 fft3d_ps 1111 14.6 1.084 1.115 2.749 2.780 potential_pw2rs 110 12.3 0.013 0.013 2.685 2.710 calculate_first_density_matrix 1 7.0 0.000 0.002 2.506 2.509 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.471 2.503 grid_collocate_task_list 110 9.6 2.221 2.501 2.221 2.501 mp_alltoall_d11v 2046 13.8 2.040 2.326 2.040 2.326 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.272 2.275 cp_fm_cholesky_decompose 22 10.9 2.098 2.142 2.098 2.142 mp_allgather_i34 2055 14.4 0.869 1.969 0.869 1.969 mp_sum_l 6594 12.7 1.323 1.946 1.323 1.946 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.889 1.914 jit_kernel_multiply 10 15.8 1.057 1.805 1.057 1.805 mp_waitany 13376 13.8 1.245 1.784 1.245 1.784 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.758 1.770 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.598 1.702 multiply_cannon_metrocomm4 20550 15.4 0.060 0.064 0.847 1.614 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.578 acc_transpose_blocks 24660 15.4 0.104 0.106 1.528 1.560 mp_irecv_dv 62702 16.1 0.747 1.535 0.747 1.535 mp_alltoall_z22v 1111 16.6 1.452 1.524 1.452 1.524 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=72.737000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=696.818182, yerr=10.607186 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 831.688704E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.030 56.883 56.884 qs_mol_dyn_low 1 2.0 0.003 0.003 56.653 56.662 qs_forces 11 3.9 0.003 0.003 55.893 55.894 qs_energies 11 4.9 0.001 0.001 52.177 52.182 scf_env_do_scf 11 5.9 0.000 0.001 43.791 43.791 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.743 35.743 velocity_verlet 10 3.0 0.001 0.001 32.367 32.413 dbcsr_multiply_generic 2055 12.4 0.104 0.108 22.825 22.975 qs_scf_new_mos 99 7.5 0.001 0.001 20.493 20.557 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.492 20.556 ot_scf_mini 99 9.5 0.002 0.002 19.245 19.261 multiply_cannon 2055 13.4 0.246 0.260 17.288 18.709 multiply_cannon_loop 2055 14.4 0.324 0.337 15.916 16.243 rebuild_ks_matrix 110 8.3 0.000 0.000 12.094 12.120 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.094 12.120 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.724 10.747 ot_mini 99 10.5 0.001 0.001 10.391 10.402 multiply_cannon_multrec 8220 15.4 3.236 4.472 7.448 8.412 init_scf_loop 11 6.9 0.000 0.000 8.001 8.002 mp_waitall_1 103326 16.6 6.082 7.871 6.082 7.871 sum_up_and_integrate 110 10.3 0.080 0.082 6.659 6.672 qs_ot_get_derivative 99 11.5 0.001 0.001 6.591 6.607 integrate_v_rspace 110 11.3 0.003 0.003 6.579 6.592 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.375 6.396 calculate_rho_elec 110 8.6 0.115 0.116 6.374 6.396 prepare_preconditioner 11 7.9 0.000 0.000 6.306 6.310 make_preconditioner 11 8.9 0.000 0.000 6.306 6.310 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.889 5.963 init_scf_run 11 5.9 0.000 0.001 5.231 5.231 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.231 5.231 dbcsr_mm_accdrv_process 17442 15.9 2.830 3.636 4.084 5.021 qs_ot_get_p 110 10.4 0.001 0.001 4.916 4.933 make_m2s 4110 13.4 0.039 0.040 4.277 4.544 make_images 4110 14.4 0.638 0.692 4.146 4.412 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.916 4.272 pw_transfer 1331 11.6 0.066 0.072 3.971 3.990 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.863 3.886 ot_diis_step 99 11.5 0.012 0.012 3.777 3.777 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.692 3.706 apply_single 110 13.6 0.000 0.000 3.692 3.706 grid_integrate_task_list 110 12.3 3.365 3.556 3.365 3.556 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.520 3.523 density_rs2pw 110 9.6 0.004 0.004 3.091 3.470 fft_wrap_pw1pw2_140 451 13.1 0.833 0.844 3.273 3.303 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.217 3.218 cp_fm_cholesky_invert 11 10.9 3.062 3.065 3.062 3.065 multiply_cannon_sync_h2d 8220 15.4 2.911 3.031 2.911 3.031 make_images_data 4110 15.4 0.038 0.044 2.438 2.833 hybrid_alltoall_any 4261 16.3 0.201 0.864 2.357 2.812 wfi_extrapolate 11 7.9 0.001 0.001 2.806 2.806 rs_pw_transfer 902 11.9 0.010 0.011 2.330 2.745 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.676 2.683 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.681 2.682 cp_fm_diag_elpa_base 48 14.0 2.615 2.639 2.674 2.681 grid_collocate_task_list 110 9.6 2.316 2.575 2.316 2.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.523 2.533 calculate_dm_sparse 110 9.5 0.001 0.001 2.471 2.526 fft3d_ps 1111 14.6 1.136 1.165 2.410 2.424 calculate_first_density_matrix 1 7.0 0.000 0.000 2.319 2.322 potential_pw2rs 110 12.3 0.015 0.016 2.187 2.193 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.140 2.161 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.767 1.979 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.923 1.937 mp_alltoall_d11v 2046 13.8 1.635 1.861 1.635 1.861 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.825 1.838 cp_fm_cholesky_decompose 22 10.9 1.799 1.812 1.799 1.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.703 1.706 mp_allgather_i34 2055 14.4 0.558 1.658 0.558 1.658 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.520 1.640 dbcsr_complete_redistribute 325 12.2 0.570 0.600 1.470 1.581 mp_waitany 9240 13.8 1.122 1.539 1.122 1.539 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.426 1.443 jit_kernel_multiply 8 15.8 0.943 1.407 0.943 1.407 qs_create_task_list 11 7.9 0.011 0.018 1.233 1.338 generate_qs_task_list 11 8.9 0.373 0.441 1.222 1.337 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.826 1.233 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.841 1.233 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.157 1.179 rs_gather_matrices 110 12.3 0.323 0.362 0.937 1.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.884000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.909091, yerr=11.719871 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 1.371906E+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 9752 812. 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 0.027 0.044 94.143 94.144 qs_mol_dyn_low 1 2.0 0.003 0.003 93.757 93.766 qs_forces 11 3.9 0.021 0.021 93.631 93.633 qs_energies 11 4.9 0.001 0.001 89.281 89.284 scf_env_do_scf 11 5.9 0.001 0.001 78.592 78.592 velocity_verlet 10 3.0 0.001 0.001 59.826 59.839 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 47.981 47.983 init_scf_loop 11 6.9 0.000 0.000 30.531 30.533 dbcsr_multiply_generic 2055 12.4 0.119 0.124 30.082 30.224 qs_scf_new_mos 99 7.5 0.001 0.001 28.650 28.699 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.649 28.698 prepare_preconditioner 11 7.9 0.000 0.000 28.363 28.376 make_preconditioner 11 8.9 0.000 0.000 28.363 28.376 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.397 27.798 ot_scf_mini 99 9.5 0.002 0.002 26.787 26.806 multiply_cannon 2055 13.4 0.347 0.372 22.063 22.867 multiply_cannon_loop 2055 14.4 0.343 0.346 20.026 20.452 cp_fm_upper_to_full 70 14.2 12.884 18.550 12.884 18.550 rebuild_ks_matrix 110 8.3 0.000 0.001 14.808 14.868 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.807 14.868 ot_mini 99 10.5 0.001 0.001 14.578 14.617 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.366 13.419 dbcsr_complete_redistribute 325 12.2 1.015 1.028 7.698 10.990 qs_ot_get_derivative 99 11.5 0.001 0.001 9.929 9.950 mp_waitall_1 84994 16.7 8.517 9.936 8.517 9.936 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.597 9.882 multiply_cannon_multrec 8220 15.4 4.393 4.618 9.558 9.658 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.951 9.191 mp_alltoall_i22 605 13.7 5.584 8.894 5.584 8.894 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.337 8.368 calculate_rho_elec 110 8.6 0.226 0.227 8.336 8.367 sum_up_and_integrate 110 10.3 0.150 0.151 7.843 7.859 integrate_v_rspace 110 11.3 0.003 0.004 7.692 7.708 qs_ot_get_p 110 10.4 0.001 0.001 6.769 6.823 make_m2s 4110 13.4 0.042 0.043 5.990 6.652 cp_fm_cholesky_invert 11 10.9 6.479 6.485 6.479 6.485 make_images 4110 14.4 0.879 0.931 5.799 6.460 init_scf_run 11 5.9 0.000 0.001 6.305 6.305 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.305 6.305 pw_transfer 1331 11.6 0.075 0.075 5.535 5.551 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.071 5.446 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.418 5.433 dbcsr_mm_accdrv_process 11614 15.7 3.179 3.612 5.023 5.242 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.710 5.189 apply_single 110 13.6 0.000 0.000 4.710 5.188 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.990 5.002 fft_wrap_pw1pw2_140 451 13.1 1.335 1.340 4.638 4.653 ot_diis_step 99 11.5 0.015 0.016 4.605 4.605 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.485 4.487 make_images_data 4110 15.4 0.041 0.044 3.294 4.303 density_rs2pw 110 9.6 0.004 0.004 4.227 4.267 hybrid_alltoall_any 4261 16.3 0.257 0.552 3.208 4.224 multiply_cannon_sync_h2d 8220 15.4 3.951 3.955 3.951 3.955 grid_integrate_task_list 110 12.3 3.695 3.771 3.695 3.771 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.767 3.767 cp_fm_diag_elpa_base 48 14.0 3.210 3.423 3.765 3.765 qs_energies_init_hamiltonians 11 5.9 0.009 0.011 3.718 3.719 wfi_extrapolate 11 7.9 0.001 0.001 3.601 3.602 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.126 3.561 fft3d_ps 1111 14.6 1.307 1.313 3.350 3.364 calculate_dm_sparse 110 9.5 0.001 0.001 3.253 3.278 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.954 2.958 rs_pw_transfer 902 11.9 0.010 0.011 2.822 2.877 potential_pw2rs 110 12.3 0.021 0.022 2.837 2.847 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.754 2.768 grid_collocate_task_list 110 9.6 2.685 2.705 2.685 2.705 calculate_first_density_matrix 1 7.0 0.000 0.000 2.568 2.568 cp_fm_cholesky_decompose 22 10.9 2.550 2.564 2.550 2.564 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.404 2.419 mp_alltoall_d11v 2046 13.8 2.318 2.390 2.318 2.390 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.239 2.290 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.191 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.043 2.051 mp_allgather_i34 2055 14.4 0.784 1.994 0.784 1.994 qs_create_task_list 11 7.9 0.000 0.000 1.900 1.947 generate_qs_task_list 11 8.9 0.731 0.788 1.900 1.946 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.864 1.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.144000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1232.909091, yerr=59.561205 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.091776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 1505569. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.069 209.219 209.221 qs_mol_dyn_low 1 2.0 0.009 0.049 208.605 208.619 qs_forces 11 3.9 0.005 0.005 208.468 208.470 qs_energies 11 4.9 0.001 0.002 202.762 202.775 scf_env_do_scf 11 5.9 0.001 0.001 185.996 186.000 scf_env_do_scf_inner_loop 117 6.6 0.003 0.007 164.854 164.856 dbcsr_multiply_generic 2507 12.6 0.177 0.182 124.290 125.205 velocity_verlet 10 3.0 0.001 0.001 124.810 124.812 qs_scf_new_mos 117 7.6 0.001 0.001 124.436 124.661 qs_scf_loop_do_ot 117 8.6 0.001 0.001 124.435 124.660 ot_scf_mini 117 9.6 0.003 0.003 117.808 117.999 multiply_cannon 2507 13.6 0.238 0.252 100.926 102.986 multiply_cannon_loop 2507 14.6 2.086 2.121 98.716 100.762 ot_mini 117 10.6 0.001 0.001 65.498 65.744 multiply_cannon_multrec 60168 15.6 33.400 35.516 41.555 43.139 qs_ot_get_derivative 117 11.6 0.001 0.002 40.742 40.934 rebuild_ks_matrix 128 8.3 0.001 0.001 33.909 34.454 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.031 33.909 34.453 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.370 30.883 mp_waitall_1 267128 16.5 28.419 30.856 28.419 30.856 qs_ot_get_p 128 10.4 0.001 0.001 30.330 30.647 multiply_cannon_sync_h2d 60168 15.6 27.504 30.003 27.504 30.003 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.310 25.448 apply_single 128 13.6 0.001 0.001 24.310 25.447 ot_diis_step 117 11.6 0.007 0.008 24.482 24.484 qs_ot_p2m_diag 83 11.4 0.079 0.106 23.668 23.721 cp_dbcsr_syevd 83 12.4 0.005 0.006 21.093 21.094 init_scf_loop 11 6.9 0.000 0.001 21.065 21.066 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.331 19.477 cp_fm_diag_elpa 83 13.4 0.000 0.001 18.056 18.094 cp_fm_diag_elpa_base 83 14.4 17.974 18.017 18.051 18.091 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 15.512 17.561 prepare_preconditioner 11 7.9 0.000 0.000 16.421 16.454 make_preconditioner 11 8.9 0.000 0.000 16.421 16.454 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.648 15.838 make_m2s 5014 13.6 0.106 0.115 14.091 14.552 sum_up_and_integrate 128 10.3 0.089 0.106 14.381 14.393 make_images 5014 14.6 0.404 0.422 13.911 14.381 integrate_v_rspace 128 11.3 0.004 0.005 14.292 14.309 qs_rho_update_rho_low 128 7.7 0.001 0.002 14.079 14.190 calculate_rho_elec 128 8.7 0.046 0.065 14.078 14.190 init_scf_run 11 5.9 0.000 0.001 12.502 12.502 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.502 12.502 density_rs2pw 128 9.7 0.006 0.007 7.318 10.673 rs_pw_transfer 1046 11.9 0.017 0.019 6.073 9.468 mp_sum_l 7950 12.9 8.031 9.367 8.031 9.367 cp_fm_cholesky_invert 11 10.9 9.333 9.341 9.333 9.341 wfi_extrapolate 11 7.9 0.001 0.001 9.203 9.203 calculate_dm_sparse 128 9.5 0.001 0.001 8.429 8.521 multiply_cannon_metrocomm1 60168 15.6 0.091 0.096 6.134 8.368 dbcsr_mm_accdrv_process 124484 16.2 3.113 3.244 7.714 8.246 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.944 8.084 pw_transfer 1547 11.6 0.075 0.089 7.779 7.953 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.818 7.914 make_images_data 5014 15.6 0.067 0.074 6.958 7.869 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.576 7.748 grid_integrate_task_list 128 12.3 7.055 7.585 7.055 7.585 hybrid_alltoall_any 5200 16.5 0.292 2.262 6.075 7.265 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.004 6.721 6.733 fft_wrap_pw1pw2_140 523 13.2 1.278 1.327 6.479 6.662 mp_waitany 16020 13.9 2.705 6.067 2.705 6.067 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.844 5.974 mp_alltoall_d11v 2415 14.1 4.322 5.931 4.322 5.931 grid_collocate_task_list 128 9.7 4.672 5.930 4.672 5.930 fft3d_ps 1291 14.7 2.135 2.769 5.333 5.641 rs_pw_transfer_RS2PW_140 139 11.5 0.287 0.306 2.185 5.568 cp_fm_cholesky_decompose 22 10.9 4.885 4.897 4.885 4.897 potential_pw2rs 128 12.3 0.009 0.011 4.722 4.750 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=209.221000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.818182, yerr=6.057575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 829.976576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57964. MP_Allreduce 11128 958. MP_Sync 87 MP_Alltoall 1969 6638201. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.072 0.128 205.276 205.278 qs_mol_dyn_low 1 2.0 0.004 0.013 204.642 204.654 qs_forces 11 3.9 0.013 0.075 204.514 204.518 qs_energies 11 4.9 0.005 0.031 197.380 197.397 scf_env_do_scf 11 5.9 0.001 0.001 180.323 180.334 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 143.876 143.880 velocity_verlet 10 3.0 0.003 0.011 128.269 128.270 qs_scf_new_mos 117 7.6 0.001 0.001 101.602 101.956 qs_scf_loop_do_ot 117 8.6 0.001 0.002 101.601 101.955 dbcsr_multiply_generic 2507 12.6 0.190 0.195 99.551 100.776 ot_scf_mini 117 9.6 0.004 0.006 96.731 97.125 multiply_cannon 2507 13.6 0.478 0.529 78.261 82.572 multiply_cannon_loop 2507 14.6 1.253 1.292 74.566 77.024 ot_mini 117 10.6 0.001 0.001 51.328 51.659 mp_waitall_1 214728 16.6 25.787 40.746 25.787 40.746 init_scf_loop 11 6.9 0.001 0.006 36.344 36.348 multiply_cannon_multrec 30084 15.6 22.129 26.332 31.727 36.237 rebuild_ks_matrix 128 8.3 0.001 0.001 34.811 35.286 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.029 34.810 35.286 prepare_preconditioner 11 7.9 0.000 0.000 31.732 31.782 make_preconditioner 11 8.9 0.000 0.001 31.732 31.781 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.240 31.677 make_full_inverse_cholesky 11 9.9 0.000 0.001 30.342 30.895 qs_ot_get_derivative 117 11.6 0.001 0.002 29.245 29.646 multiply_cannon_metrocomm3 30084 15.6 0.091 0.095 15.749 29.249 qs_ot_get_p 128 10.4 0.001 0.002 26.811 27.189 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 22.118 23.225 apply_single 128 13.6 0.001 0.001 22.118 23.225 ot_diis_step 117 11.6 0.014 0.023 21.899 21.901 multiply_cannon_sync_h2d 30084 15.6 19.435 21.875 19.435 21.875 qs_ot_p2m_diag 83 11.4 0.188 0.216 21.490 21.528 cp_dbcsr_syevd 83 12.4 0.006 0.006 20.207 20.209 cp_fm_cholesky_invert 11 10.9 18.587 18.602 18.587 18.602 make_m2s 5014 13.6 0.093 0.098 15.302 17.040 make_images 5014 14.6 1.164 1.353 15.089 16.827 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.669 16.700 cp_fm_diag_elpa_base 83 14.4 16.375 16.501 16.662 16.695 sum_up_and_integrate 128 10.3 0.116 0.133 15.830 15.869 integrate_v_rspace 128 11.3 0.004 0.004 15.713 15.758 qs_rho_update_rho_low 128 7.7 0.001 0.002 15.282 15.319 calculate_rho_elec 128 8.7 0.089 0.106 15.281 15.318 init_scf_run 11 5.9 0.000 0.001 12.070 12.072 scf_env_initial_rho_setup 11 6.9 0.000 0.002 12.070 12.071 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.559 11.827 make_images_data 5014 15.6 0.064 0.074 9.128 11.161 density_rs2pw 128 9.7 0.007 0.007 8.559 11.037 multiply_cannon_metrocomm4 27577 15.6 0.105 0.119 3.733 10.574 hybrid_alltoall_any 5200 16.5 0.345 1.520 7.758 10.547 mp_irecv_dv 69486 16.3 3.533 10.180 3.533 10.180 rs_pw_transfer 1046 11.9 0.014 0.016 7.381 9.876 dbcsr_mm_accdrv_process 62242 16.2 4.456 5.449 9.056 9.660 pw_transfer 1547 11.6 0.087 0.103 9.418 9.490 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.191 9.268 wfi_extrapolate 11 7.9 0.001 0.001 8.835 8.835 cp_fm_cholesky_decompose 22 10.9 8.196 8.284 8.196 8.284 fft_wrap_pw1pw2_140 523 13.2 1.332 1.352 7.725 7.824 grid_integrate_task_list 128 12.3 7.134 7.683 7.134 7.683 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.430 7.206 calculate_dm_sparse 128 9.5 0.001 0.001 6.565 6.713 fft3d_ps 1291 14.7 2.808 2.990 6.499 6.561 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.242 6.264 potential_pw2rs 128 12.3 0.016 0.019 6.045 6.080 grid_collocate_task_list 128 9.7 4.791 5.957 4.791 5.957 mp_sum_l 7950 12.9 4.242 5.862 4.242 5.862 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.502 5.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.339 5.481 mp_allgather_i34 2507 14.6 2.265 5.446 2.265 5.446 mp_waitany 11748 13.9 2.678 5.202 2.678 5.202 mp_alltoall_d11v 2415 14.1 4.339 5.154 4.339 5.154 rs_pw_transfer_RS2PW_140 139 11.5 0.352 0.377 2.200 4.666 dbcsr_complete_redistribute 395 12.7 0.771 0.865 3.366 4.300 mp_sum_d 4479 12.1 2.801 4.150 2.801 4.150 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=205.278000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/17/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 2.928533E+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 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 940.232704E+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 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11057 1000. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.056 192.737 192.738 qs_mol_dyn_low 1 2.0 0.003 0.004 192.239 192.253 qs_forces 11 3.9 0.021 0.039 192.130 192.133 qs_energies 11 4.9 0.007 0.049 185.205 185.244 scf_env_do_scf 11 5.9 0.001 0.002 168.431 168.431 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 128.993 128.994 velocity_verlet 10 3.0 0.001 0.001 122.626 122.628 qs_scf_new_mos 117 7.6 0.001 0.001 88.739 89.071 qs_scf_loop_do_ot 117 8.6 0.001 0.002 88.738 89.070 ot_scf_mini 117 9.6 0.003 0.004 84.556 84.957 dbcsr_multiply_generic 2507 12.6 0.179 0.185 83.557 84.585 multiply_cannon 2507 13.6 0.498 0.515 62.451 66.613 multiply_cannon_loop 2507 14.6 0.864 0.889 59.092 62.000 ot_mini 117 10.6 0.001 0.001 43.911 44.320 init_scf_loop 11 6.9 0.001 0.007 39.311 39.313 mp_waitall_1 170520 16.6 26.050 35.482 26.050 35.482 prepare_preconditioner 11 7.9 0.000 0.001 35.142 35.198 make_preconditioner 11 8.9 0.000 0.001 35.142 35.198 make_full_inverse_cholesky 11 9.9 0.000 0.001 32.643 34.152 rebuild_ks_matrix 128 8.3 0.001 0.001 31.961 32.405 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 31.960 32.404 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.698 29.103 qs_ot_get_p 128 10.4 0.001 0.001 25.816 26.391 multiply_cannon_multrec 20056 15.6 13.584 16.971 22.240 25.508 qs_ot_get_derivative 117 11.6 0.001 0.002 24.162 24.563 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 14.901 24.030 qs_ot_p2m_diag 83 11.4 0.266 0.273 21.064 21.071 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.785 20.771 apply_single 128 13.6 0.001 0.001 19.785 20.770 cp_dbcsr_syevd 83 12.4 0.005 0.006 19.929 19.930 ot_diis_step 117 11.6 0.017 0.018 19.642 19.643 cp_fm_cholesky_invert 11 10.9 16.976 16.987 16.976 16.987 make_m2s 5014 13.6 0.082 0.086 15.789 16.735 make_images 5014 14.6 1.181 1.274 15.555 16.499 cp_fm_diag_elpa 83 13.4 0.000 0.001 16.131 16.148 cp_fm_diag_elpa_base 83 14.4 15.661 15.866 16.126 16.143 multiply_cannon_sync_h2d 20056 15.6 14.244 15.762 14.244 15.762 sum_up_and_integrate 128 10.3 0.131 0.145 15.647 15.674 integrate_v_rspace 128 11.3 0.004 0.004 15.515 15.546 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.457 15.496 calculate_rho_elec 128 8.7 0.132 0.148 15.456 15.495 init_scf_run 11 5.9 0.000 0.001 11.519 11.519 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.518 11.519 make_images_data 5014 15.6 0.059 0.067 9.634 10.932 hybrid_alltoall_any 5200 16.5 0.438 2.004 8.313 10.106 density_rs2pw 128 9.7 0.006 0.007 8.166 10.097 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.615 9.908 pw_transfer 1547 11.6 0.087 0.104 9.397 9.504 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 9.171 9.282 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.443 8.930 cp_fm_cholesky_decompose 22 10.9 8.812 8.843 8.812 8.843 mp_irecv_dv 50230 16.2 3.323 8.684 3.323 8.684 rs_pw_transfer 1046 11.9 0.014 0.014 6.613 8.510 dbcsr_mm_accdrv_process 41502 16.2 4.413 5.091 8.106 8.220 fft_wrap_pw1pw2_140 523 13.2 1.410 1.433 7.769 7.895 wfi_extrapolate 11 7.9 0.001 0.001 7.870 7.870 grid_integrate_task_list 128 12.3 7.217 7.782 7.217 7.782 cp_fm_upper_to_full 105 14.8 5.916 7.587 5.916 7.587 dbcsr_complete_redistribute 395 12.7 1.170 1.196 5.019 6.948 fft3d_ps 1291 14.7 2.744 2.974 6.293 6.370 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.987 5.992 calculate_dm_sparse 128 9.5 0.001 0.001 5.823 5.914 grid_collocate_task_list 128 9.7 4.980 5.875 4.980 5.875 potential_pw2rs 128 12.3 0.021 0.024 5.625 5.641 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.687 5.608 mp_alltoall_d11v 2415 14.1 4.869 5.569 4.869 5.569 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.833 5.512 mp_sum_l 7950 12.9 3.518 5.167 3.518 5.167 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.025 5.161 mp_allgather_i34 2507 14.6 1.873 4.869 1.873 4.869 mp_waitany 11748 13.9 2.517 4.410 2.517 4.410 transfer_fm_to_dbcsr 11 9.9 0.019 0.029 2.480 4.335 mp_alltoall_i22 716 14.1 2.124 4.193 2.124 4.193 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.160 4.186 rs_pw_transfer_RS2PW_140 139 11.5 0.337 0.361 2.003 3.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=192.738000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=892.000000, yerr=10.126473 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 4.353788E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.143566E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379345520 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.109 192.908 192.910 qs_mol_dyn_low 1 2.0 0.003 0.003 192.303 192.316 qs_forces 11 3.9 0.004 0.004 192.159 192.165 qs_energies 11 4.9 0.007 0.046 184.892 184.901 scf_env_do_scf 11 5.9 0.001 0.001 167.009 167.017 velocity_verlet 10 3.0 0.007 0.045 127.156 127.158 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.724 119.725 qs_scf_new_mos 117 7.6 0.001 0.001 81.343 81.610 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.343 81.610 dbcsr_multiply_generic 2507 12.6 0.186 0.192 79.848 80.562 ot_scf_mini 117 9.6 0.003 0.004 76.804 77.095 multiply_cannon 2507 13.6 0.552 0.589 54.827 58.685 multiply_cannon_loop 2507 14.6 1.194 1.222 51.058 52.706 init_scf_loop 11 6.9 0.001 0.004 47.151 47.153 prepare_preconditioner 11 7.9 0.000 0.000 43.003 43.024 make_preconditioner 11 8.9 0.000 0.001 43.003 43.024 ot_mini 117 10.6 0.001 0.001 42.522 42.812 make_full_inverse_cholesky 11 9.9 0.011 0.021 36.540 41.636 multiply_cannon_multrec 30084 15.6 14.227 19.620 26.254 31.205 rebuild_ks_matrix 128 8.3 0.001 0.001 30.475 30.748 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 30.475 30.748 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.441 27.674 mp_waitall_1 147882 16.7 17.538 27.139 17.538 27.139 qs_ot_get_derivative 117 11.6 0.001 0.002 22.894 23.197 make_m2s 5014 13.6 0.098 0.102 20.597 21.663 make_images 5014 14.6 1.940 2.297 20.291 21.363 qs_ot_get_p 128 10.4 0.001 0.001 20.471 20.778 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.026 19.563 apply_single 128 13.6 0.001 0.001 19.025 19.563 ot_diis_step 117 11.6 0.017 0.019 19.501 19.504 cp_fm_upper_to_full 105 14.8 11.502 17.008 11.502 17.008 cp_fm_cholesky_invert 11 10.9 16.591 16.600 16.591 16.600 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.236 16.289 sum_up_and_integrate 128 10.3 0.140 0.152 15.103 15.126 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.026 15.069 calculate_rho_elec 128 8.7 0.176 0.192 15.026 15.068 integrate_v_rspace 128 11.3 0.004 0.004 14.963 14.990 multiply_cannon_metrocomm3 30084 15.6 0.048 0.051 6.193 14.971 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.942 14.943 make_images_data 5014 15.6 0.062 0.067 11.194 13.221 dbcsr_complete_redistribute 395 12.7 1.499 1.608 9.167 13.029 multiply_cannon_sync_h2d 30084 15.6 11.742 12.556 11.742 12.556 hybrid_alltoall_any 5200 16.5 0.518 2.086 9.862 12.236 dbcsr_mm_accdrv_process 62264 16.2 7.383 8.487 11.603 12.046 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.757 11.768 init_scf_run 11 5.9 0.000 0.001 11.760 11.762 scf_env_initial_rho_setup 11 6.9 0.009 0.013 11.760 11.761 cp_fm_diag_elpa_base 83 14.4 10.771 11.101 11.750 11.759 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.723 11.549 transfer_fm_to_dbcsr 11 9.9 0.001 0.003 6.440 10.199 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.519 9.731 pw_transfer 1547 11.6 0.087 0.101 9.552 9.635 mp_alltoall_i22 716 14.1 5.685 9.440 5.685 9.440 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.324 9.414 density_rs2pw 128 9.7 0.006 0.007 7.336 8.648 fft_wrap_pw1pw2_140 523 13.2 1.558 1.590 8.226 8.322 grid_integrate_task_list 128 12.3 7.470 7.987 7.470 7.987 wfi_extrapolate 11 7.9 0.001 0.001 7.870 7.871 cp_fm_cholesky_decompose 22 10.9 7.774 7.865 7.774 7.865 multiply_cannon_metrocomm4 25070 15.6 0.080 0.091 2.765 6.898 mp_irecv_dv 76098 16.2 2.618 6.632 2.618 6.632 calculate_dm_sparse 128 9.5 0.001 0.001 6.348 6.430 rs_pw_transfer 1046 11.9 0.013 0.014 5.004 6.422 fft3d_ps 1291 14.7 2.832 2.913 6.249 6.296 mp_alltoall_d11v 2415 14.1 5.155 6.077 5.155 6.077 grid_collocate_task_list 128 9.7 5.153 5.926 5.153 5.926 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.504 5.562 potential_pw2rs 128 12.3 0.023 0.023 4.805 4.827 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.524 4.526 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.390 4.474 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.398 4.467 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.197 4.268 mp_allgather_i34 2507 14.6 1.882 4.035 1.882 4.035 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.910000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1076.272727, yerr=21.692441 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 1.496584E+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 11050 1168. 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 0.045 0.062 177.951 177.952 qs_mol_dyn_low 1 2.0 0.003 0.003 177.400 177.413 qs_forces 11 3.9 0.004 0.004 177.286 177.291 qs_energies 11 4.9 0.001 0.001 169.646 169.653 scf_env_do_scf 11 5.9 0.001 0.001 151.204 151.209 velocity_verlet 10 3.0 0.001 0.001 114.973 114.977 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.637 114.638 qs_scf_new_mos 116 7.6 0.001 0.001 75.843 75.940 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.842 75.940 dbcsr_multiply_generic 2485 12.5 0.178 0.190 74.303 74.764 ot_scf_mini 116 9.6 0.003 0.004 71.418 71.469 multiply_cannon 2485 13.5 0.577 0.638 54.323 58.667 multiply_cannon_loop 2485 14.5 0.443 0.460 49.780 50.741 ot_mini 116 10.6 0.001 0.001 39.681 39.728 init_scf_loop 11 6.9 0.000 0.000 36.419 36.421 mp_waitall_1 124680 16.7 25.865 32.694 25.865 32.694 prepare_preconditioner 11 7.9 0.000 0.000 32.397 32.424 make_preconditioner 11 8.9 0.000 0.000 32.397 32.424 make_full_inverse_cholesky 11 9.9 0.016 0.027 30.311 30.579 rebuild_ks_matrix 127 8.3 0.001 0.001 30.078 30.195 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.077 30.194 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.281 27.391 multiply_cannon_multrec 9940 15.5 10.330 14.602 17.924 20.611 qs_ot_get_derivative 116 11.6 0.001 0.002 19.916 19.973 ot_diis_step 116 11.6 0.019 0.020 19.695 19.696 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.251 19.567 apply_single 127 13.6 0.001 0.001 19.251 19.566 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 12.031 19.160 qs_ot_get_p 127 10.4 0.001 0.001 18.883 18.943 make_m2s 4970 13.5 0.066 0.070 16.353 18.565 cp_fm_cholesky_invert 11 10.9 18.492 18.498 18.492 18.498 make_images 4970 14.5 2.292 2.734 16.048 18.261 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.535 15.570 calculate_rho_elec 127 8.7 0.257 0.267 15.534 15.570 sum_up_and_integrate 127 10.3 0.180 0.189 15.223 15.265 integrate_v_rspace 127 11.3 0.004 0.004 15.043 15.094 qs_ot_p2m_diag 82 11.4 0.490 0.495 15.073 15.088 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.923 13.925 make_images_data 4970 15.5 0.050 0.060 9.952 12.286 hybrid_alltoall_any 5155 16.4 0.812 3.804 9.722 12.128 multiply_cannon_sync_h2d 9940 15.5 11.563 11.907 11.563 11.907 init_scf_run 11 5.9 0.000 0.001 11.510 11.510 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.510 11.510 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.831 10.841 cp_fm_diag_elpa_base 82 14.4 10.564 10.666 10.823 10.834 pw_transfer 1535 11.6 0.086 0.097 9.695 9.726 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.472 9.508 fft_wrap_pw1pw2_140 519 13.2 1.895 1.934 8.349 8.390 cp_fm_cholesky_decompose 22 10.9 8.234 8.354 8.234 8.354 grid_integrate_task_list 127 12.3 7.708 8.146 7.708 8.146 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 7.913 7.950 density_rs2pw 127 9.7 0.006 0.006 7.049 7.940 dbcsr_mm_accdrv_process 20590 16.1 2.736 3.544 7.224 7.937 wfi_extrapolate 11 7.9 0.001 0.001 7.641 7.641 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.540 7.535 mp_allgather_i34 2485 14.5 2.754 6.998 2.754 6.998 calculate_dm_sparse 127 9.5 0.001 0.001 6.193 6.275 mp_alltoall_d11v 2401 14.1 5.231 6.167 5.231 6.167 grid_collocate_task_list 127 9.7 5.421 6.095 5.421 6.095 fft3d_ps 1281 14.7 2.746 2.824 5.978 6.007 dbcsr_complete_redistribute 393 12.7 2.124 2.169 5.239 5.633 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.360 5.383 qs_energies_init_hamiltonians 11 5.9 0.003 0.019 5.331 5.332 rs_pw_transfer 1038 11.9 0.012 0.013 4.311 5.218 potential_pw2rs 127 12.3 0.026 0.027 4.615 4.632 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.243 4.263 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.565 3.885 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.833 3.863 mp_irecv_dv 28618 15.9 1.799 3.797 1.799 3.797 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.696 3.726 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.698 3.704 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.388 3.699 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.611 3.692 calculate_first_density_matrix 1 7.0 0.000 0.000 3.632 3.635 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.584 3.600 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.952000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1394.090909, yerr=43.652459 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.109695E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59127. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.069 306.094 306.094 qs_mol_dyn_low 1 2.0 0.003 0.003 305.469 305.482 qs_forces 11 3.9 0.004 0.004 305.244 305.246 qs_energies 11 4.9 0.001 0.001 295.783 295.794 scf_env_do_scf 11 5.9 0.001 0.001 273.388 273.398 velocity_verlet 10 3.0 0.001 0.001 219.261 219.285 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 144.618 144.621 init_scf_loop 11 6.9 0.000 0.000 128.514 128.517 prepare_preconditioner 11 7.9 0.000 0.000 123.402 123.431 make_preconditioner 11 8.9 0.000 0.000 123.402 123.431 make_full_inverse_cholesky 11 9.9 0.034 0.039 98.395 120.534 qs_scf_new_mos 116 7.6 0.001 0.001 93.431 93.604 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.430 93.603 ot_scf_mini 116 9.6 0.004 0.004 88.619 88.673 dbcsr_multiply_generic 2485 12.5 0.209 0.221 84.605 85.236 cp_fm_upper_to_full 104 14.8 54.229 78.234 54.229 78.234 multiply_cannon 2485 13.5 0.706 0.764 59.521 60.059 multiply_cannon_loop 2485 14.5 0.470 0.480 55.692 57.261 ot_mini 116 10.6 0.001 0.001 45.450 45.504 dbcsr_complete_redistribute 393 12.7 3.969 4.015 30.690 44.668 copy_fm_to_dbcsr 208 11.6 0.001 0.002 27.287 41.296 transfer_fm_to_dbcsr 11 9.9 0.030 0.032 24.960 38.794 rebuild_ks_matrix 127 8.3 0.001 0.001 38.668 38.753 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.668 38.753 mp_alltoall_i22 712 14.1 22.780 36.973 22.780 36.973 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.560 35.632 cp_fm_cholesky_invert 11 10.9 34.609 34.615 34.609 34.615 mp_waitall_1 102768 16.8 29.543 33.675 29.543 33.675 qs_ot_get_p 127 10.4 0.001 0.001 27.449 27.495 qs_ot_get_derivative 116 11.6 0.002 0.002 24.853 24.902 qs_ot_p2m_diag 82 11.4 0.868 0.874 23.188 23.222 cp_dbcsr_syevd 82 12.4 0.006 0.006 21.404 21.408 make_m2s 4970 13.5 0.075 0.078 20.267 21.362 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 19.730 21.160 make_images 4970 14.5 3.727 3.855 19.791 20.888 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.772 20.788 calculate_rho_elec 127 8.7 0.477 0.478 20.772 20.787 ot_diis_step 116 11.6 0.022 0.023 20.555 20.555 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.743 20.058 apply_single 127 13.6 0.001 0.001 19.743 20.057 sum_up_and_integrate 127 10.3 0.318 0.320 19.865 19.942 integrate_v_rspace 127 11.3 0.004 0.004 19.546 19.621 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.142 18.143 cp_fm_diag_elpa_base 82 14.4 13.707 15.369 18.136 18.137 multiply_cannon_multrec 9940 15.5 10.448 12.282 17.847 17.899 multiply_cannon_sync_h2d 9940 15.5 15.539 15.557 15.539 15.557 make_images_data 4970 15.5 0.057 0.062 10.913 12.875 hybrid_alltoall_any 5155 16.4 1.289 3.014 10.903 12.758 init_scf_run 11 5.9 0.000 0.001 12.428 12.429 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.428 12.428 pw_transfer 1535 11.6 0.092 0.093 11.736 11.747 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 11.503 11.512 fft_wrap_pw1pw2_140 519 13.2 3.044 3.069 10.156 10.169 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.689 9.736 mp_alltoall_d11v 2401 14.1 8.644 9.546 8.644 9.546 cp_fm_cholesky_decompose 22 10.9 9.381 9.469 9.381 9.469 wfi_extrapolate 11 7.9 0.001 0.001 9.164 9.164 dbcsr_mm_accdrv_process 20590 16.0 3.753 5.396 7.163 8.994 grid_integrate_task_list 127 12.3 8.532 8.725 8.532 8.725 density_rs2pw 127 9.7 0.005 0.005 8.241 8.389 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 7.914 7.915 calculate_dm_sparse 127 9.5 0.001 0.001 6.687 6.824 fft3d_ps 1281 14.7 2.825 2.838 6.624 6.653 grid_collocate_task_list 127 9.7 6.390 6.467 6.390 6.467 rs_scatter_matrices 138 9.7 3.655 4.588 6.118 6.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.225 6.252 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=306.094000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2814.000000, yerr=148.928293 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 1.259594E+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 3139 6114. MP_Sync 4 MP_Alltoall 54 25001975. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.091 85.771 85.782 qs_energies 1 2.0 0.000 0.001 85.067 85.079 ls_scf 1 3.0 0.002 0.012 84.145 84.155 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.836 72.990 multiply_cannon 111 7.7 0.017 0.020 56.007 57.423 multiply_cannon_loop 111 8.7 0.211 0.228 52.645 53.943 ls_scf_main 1 4.0 0.001 0.022 52.580 52.582 density_matrix_trs4 2 5.0 0.002 0.011 46.933 46.997 ls_scf_init_scf 1 4.0 0.000 0.001 28.493 28.494 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.292 27.361 mp_waitall_1 11031 10.9 22.577 25.538 22.577 25.538 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.168 25.186 multiply_cannon_multrec 2664 9.7 8.192 8.951 15.576 17.344 multiply_cannon_sync_h2d 2664 9.7 13.578 15.800 13.578 15.800 make_m2s 222 7.7 0.010 0.012 13.109 13.565 make_images 222 8.7 0.101 0.111 13.087 13.544 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.755 12.798 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.503 8.957 make_images_data 222 9.7 0.005 0.005 7.652 8.161 dbcsr_mm_accdrv_process 4760 10.4 0.501 0.623 7.001 7.999 hybrid_alltoall_any 227 10.6 0.218 1.847 6.611 7.549 dbcsr_mm_accdrv_process_sort 4760 11.4 6.299 7.214 6.299 7.214 calculate_norms 4752 9.8 5.512 6.230 5.512 6.230 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.046 5.151 mp_sum_l 887 5.1 3.247 4.801 3.247 4.801 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.070 4.346 mp_irecv_dv 6231 10.9 2.054 4.329 2.054 4.329 make_images_sizes 222 9.7 0.000 0.000 0.704 3.817 mp_alltoall_i44 222 10.7 0.704 3.817 0.704 3.817 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.391 3.663 arnoldi_extremal 4 6.8 0.000 0.001 3.320 3.343 arnoldi_normal_ev 4 7.8 0.001 0.003 3.319 3.343 build_subspace 16 8.4 0.009 0.012 3.221 3.223 ls_scf_post 1 4.0 0.001 0.008 3.070 3.082 ls_scf_store_result 1 5.0 0.000 0.000 2.866 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.392 2.765 dbcsr_merge_single_wm 555 10.7 0.453 0.613 2.383 2.757 make_images_pack 222 9.7 2.212 2.637 2.214 2.639 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.334 2.578 dbcsr_sort_data 658 11.4 2.188 2.507 2.188 2.507 ls_scf_dm_to_ks 2 5.0 0.004 0.033 2.381 2.458 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.451 2.063 2.453 buffer_matrices_ensure_size 222 8.7 1.754 2.066 1.754 2.066 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.850 1.851 rebuild_ks_matrix 3 7.3 0.000 0.000 1.840 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.840 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.782000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 2.102858E+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 3138 10075. 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 0.032 0.050 92.214 92.215 qs_energies 1 2.0 0.000 0.000 91.564 91.572 ls_scf 1 3.0 0.000 0.000 90.233 90.241 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.195 76.520 multiply_cannon 111 7.7 0.029 0.038 53.597 57.648 ls_scf_main 1 4.0 0.000 0.000 55.795 55.800 multiply_cannon_loop 111 8.7 0.117 0.125 50.303 53.746 density_matrix_trs4 2 5.0 0.002 0.003 50.013 50.219 mp_waitall_1 9105 10.9 21.509 30.886 21.509 30.886 ls_scf_init_scf 1 4.0 0.000 0.000 30.874 30.875 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.667 29.763 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.272 27.294 multiply_cannon_multrec 1332 9.7 13.199 16.494 22.505 26.889 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.884 21.026 make_m2s 222 7.7 0.008 0.009 15.685 16.451 make_images 222 8.7 1.588 1.961 15.655 16.422 dbcsr_mm_accdrv_process 4041 10.4 0.297 0.458 8.905 10.442 make_images_data 222 9.7 0.004 0.005 9.047 10.086 dbcsr_mm_accdrv_process_sort 4041 11.4 8.496 9.993 8.496 9.993 hybrid_alltoall_any 227 10.6 0.522 2.449 8.425 9.524 mp_sum_l 887 5.1 5.605 8.992 5.605 8.992 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.217 7.784 mp_irecv_dv 3311 11.0 3.196 7.736 3.196 7.736 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.339 7.122 calculate_norms 2376 9.8 6.006 6.745 6.006 6.745 multiply_cannon_sync_h2d 1332 9.7 4.829 5.998 4.829 5.998 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.072 5.309 arnoldi_extremal 4 6.8 0.001 0.005 4.652 4.682 arnoldi_normal_ev 4 7.8 0.002 0.006 4.651 4.682 build_subspace 16 8.4 0.014 0.021 4.397 4.399 ls_scf_post 1 4.0 0.000 0.000 3.564 3.573 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.164 3.392 ls_scf_store_result 1 5.0 0.000 0.000 3.276 3.372 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.222 2.741 3.224 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.643 2.744 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.211 2.664 mp_allgather_i34 111 8.7 0.983 2.529 0.983 2.529 make_images_pack 222 9.7 2.033 2.433 2.036 2.436 dbcsr_sort_data 436 11.2 1.842 2.123 1.842 2.123 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.877 1.879 rebuild_ks_matrix 3 7.3 0.000 0.000 1.864 1.866 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.864 1.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.215000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1752.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 2.759987E+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 3138 10896. 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 0.050 0.066 94.708 94.710 qs_energies 1 2.0 0.000 0.000 94.092 94.123 ls_scf 1 3.0 0.000 0.001 92.662 92.693 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.230 77.577 ls_scf_main 1 4.0 0.000 0.002 58.091 58.096 multiply_cannon 111 7.7 0.047 0.105 53.199 56.738 multiply_cannon_loop 111 8.7 0.100 0.106 49.550 53.761 density_matrix_trs4 2 5.0 0.002 0.004 51.989 52.153 mp_waitall_1 7281 11.0 24.110 34.170 24.110 34.170 ls_scf_init_scf 1 4.0 0.000 0.001 30.994 30.995 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.784 29.857 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.385 27.396 multiply_cannon_multrec 888 9.7 12.638 15.076 21.259 24.483 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.168 23.228 make_m2s 222 7.7 0.007 0.008 17.024 18.302 make_images 222 8.7 1.981 2.314 16.983 18.261 make_images_data 222 9.7 0.004 0.005 9.825 11.022 hybrid_alltoall_any 227 10.6 0.621 2.865 9.464 10.795 dbcsr_mm_accdrv_process 3754 10.4 0.246 0.413 8.131 9.360 mp_sum_l 887 5.1 5.428 8.952 5.428 8.952 dbcsr_mm_accdrv_process_sort 3754 11.4 7.760 8.947 7.760 8.947 multiply_cannon_sync_h2d 888 9.7 6.016 7.265 6.016 7.265 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.742 7.126 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.215 7.000 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.423 6.735 mp_irecv_dv 2335 11.1 2.408 6.677 2.408 6.677 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.992 5.223 arnoldi_extremal 4 6.8 0.000 0.001 5.101 5.117 arnoldi_normal_ev 4 7.8 0.001 0.005 5.101 5.116 build_subspace 16 8.4 0.014 0.020 4.799 4.806 calculate_norms 1584 9.8 4.318 4.696 4.318 4.696 mp_allgather_i34 111 8.7 1.462 3.845 1.462 3.845 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.463 3.779 ls_scf_post 1 4.0 0.004 0.033 3.577 3.609 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.595 3.020 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.309 3.398 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.948 3.079 make_images_sizes 222 9.7 0.000 0.000 0.865 2.291 mp_alltoall_i44 222 10.7 0.864 2.291 0.864 2.291 make_images_pack 222 9.7 1.816 2.128 1.819 2.130 dbcsr_sort_data 325 11.1 1.884 2.124 1.884 2.124 dbcsr_data_release 9322 10.9 1.336 2.056 1.336 2.056 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.924 1.927 rebuild_ks_matrix 3 7.3 0.000 0.000 1.906 1.908 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.906 1.908 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.710000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2168.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 3.384893E+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 3138 13030. 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 0.046 0.127 98.419 98.421 qs_energies 1 2.0 0.000 0.000 97.293 97.301 ls_scf 1 3.0 0.000 0.002 95.547 95.556 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.680 78.941 ls_scf_main 1 4.0 0.000 0.001 59.207 59.209 multiply_cannon 111 7.7 0.063 0.161 51.651 56.739 density_matrix_trs4 2 5.0 0.002 0.004 52.998 53.133 multiply_cannon_loop 111 8.7 0.115 0.128 46.515 49.670 ls_scf_init_scf 1 4.0 0.000 0.001 33.035 33.038 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.648 31.720 mp_waitall_1 6369 11.0 22.758 29.140 22.758 29.140 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.058 29.072 multiply_cannon_multrec 1332 9.7 14.187 17.526 22.134 24.604 make_m2s 222 7.7 0.008 0.010 21.289 22.788 make_images 222 8.7 3.147 3.608 21.237 22.739 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.021 17.197 make_images_data 222 9.7 0.004 0.005 11.965 13.601 hybrid_alltoall_any 227 10.6 0.799 3.733 11.290 13.001 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.409 7.596 9.113 dbcsr_mm_accdrv_process_sort 3641 11.4 7.223 8.697 7.223 8.697 mp_sum_l 887 5.1 4.190 7.659 4.190 7.659 multiply_cannon_sync_h2d 1332 9.7 5.485 6.240 5.485 6.240 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.100 6.066 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.191 6.011 mp_irecv_dv 3229 10.9 2.077 5.981 2.077 5.981 arnoldi_extremal 4 6.8 0.000 0.001 5.376 5.392 arnoldi_normal_ev 4 7.8 0.001 0.005 5.376 5.391 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.603 5.247 build_subspace 16 8.4 0.014 0.021 5.032 5.041 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.543 4.710 mp_allgather_i34 111 8.7 2.246 4.655 2.246 4.655 calculate_norms 2376 9.8 4.194 4.534 4.194 4.534 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.642 3.950 dbcsr_matrix_vector_mult_local 304 10.0 3.155 3.680 3.157 3.682 dbcsr_sort_data 658 11.4 3.045 3.378 3.045 3.378 ls_scf_post 1 4.0 0.000 0.001 3.305 3.312 dbcsr_special_finalize 555 9.7 0.006 0.007 2.782 3.175 dbcsr_merge_single_wm 555 10.7 0.534 0.668 2.773 3.166 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.087 3.145 ls_scf_store_result 1 5.0 0.000 0.000 2.994 3.063 dbcsr_data_release 10477 10.7 1.582 2.414 1.582 2.414 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.046 2.049 rebuild_ks_matrix 3 7.3 0.000 0.000 2.023 2.026 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.007 2.023 2.026 dbcsr_finalize 304 7.8 0.050 0.062 1.817 2.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.421000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2741.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 4.658258E+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 3129 15263. 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 0.043 0.061 99.967 99.968 qs_energies 1 2.0 0.000 0.000 99.004 99.008 ls_scf 1 3.0 0.000 0.000 97.024 97.034 dbcsr_multiply_generic 111 6.7 0.018 0.019 78.367 78.589 ls_scf_main 1 4.0 0.000 0.000 62.443 62.444 multiply_cannon 111 7.7 0.098 0.193 55.247 60.425 density_matrix_trs4 2 5.0 0.002 0.003 55.246 55.334 multiply_cannon_loop 111 8.7 0.071 0.082 50.835 52.761 mp_waitall_1 5436 11.0 26.623 31.447 26.623 31.447 ls_scf_init_scf 1 4.0 0.000 0.000 30.941 30.946 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.642 29.678 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.451 27.460 multiply_cannon_multrec 444 9.7 14.068 16.736 21.184 22.707 make_m2s 222 7.7 0.006 0.007 18.201 20.586 make_images 222 8.7 3.733 4.420 18.137 20.524 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.896 16.721 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.327 13.298 make_images_data 222 9.7 0.003 0.004 10.200 12.714 hybrid_alltoall_any 227 10.6 0.791 3.782 9.860 12.616 multiply_cannon_sync_h2d 444 9.7 6.530 8.023 6.530 8.023 dbcsr_mm_accdrv_process 3003 10.4 0.189 0.365 6.815 7.943 dbcsr_mm_accdrv_process_sort 3003 11.4 6.492 7.595 6.492 7.595 mp_allgather_i34 111 8.7 2.635 7.022 2.635 7.022 arnoldi_extremal 4 6.8 0.000 0.001 5.918 5.928 arnoldi_normal_ev 4 7.8 0.002 0.005 5.918 5.928 build_subspace 16 8.4 0.015 0.019 5.529 5.539 mp_sum_l 887 5.1 2.965 4.804 2.965 4.804 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.650 4.753 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.186 4.356 dbcsr_matrix_vector_mult_local 304 10.0 3.609 4.058 3.611 4.060 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.690 3.951 mp_irecv_dv 1241 11.2 1.671 3.928 1.671 3.928 make_images_sizes 222 9.7 0.000 0.000 1.324 3.794 mp_alltoall_i44 222 10.7 1.324 3.794 1.324 3.794 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.059 3.691 calculate_norms 792 9.8 3.557 3.689 3.557 3.689 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.572 3.670 ls_scf_post 1 4.0 0.000 0.000 3.640 3.644 ls_scf_store_result 1 5.0 0.000 0.000 3.412 3.465 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.287 dbcsr_merge_all 275 8.9 0.475 0.525 2.052 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.089 2.090 rebuild_ks_matrix 3 7.3 0.000 0.000 2.057 2.058 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.057 2.058 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.968000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_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 8.872395E+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 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.084 0.099 113.199 113.201 qs_energies 1 2.0 0.000 0.000 111.703 111.719 ls_scf 1 3.0 0.000 0.000 108.757 108.772 dbcsr_multiply_generic 111 6.7 0.023 0.027 81.818 81.946 ls_scf_main 1 4.0 0.000 0.000 65.964 65.982 density_matrix_trs4 2 5.0 0.002 0.003 56.570 56.634 multiply_cannon 111 7.7 0.124 0.199 54.270 55.891 multiply_cannon_loop 111 8.7 0.068 0.070 50.860 52.102 ls_scf_init_scf 1 4.0 0.000 0.000 39.080 39.083 ls_scf_init_matrix_S 1 5.0 0.000 0.000 37.410 37.427 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 34.547 34.554 mp_waitall_1 4527 11.1 22.381 26.264 22.381 26.264 make_m2s 222 7.7 0.007 0.007 23.720 24.603 make_images 222 8.7 4.599 4.989 23.610 24.492 multiply_cannon_multrec 444 9.7 17.897 18.549 22.546 23.120 hybrid_alltoall_any 227 10.6 1.665 3.648 12.759 15.484 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.933 15.371 make_images_data 222 9.7 0.003 0.004 12.883 15.260 multiply_cannon_sync_h2d 444 9.7 8.849 8.899 8.849 8.899 arnoldi_extremal 4 6.8 0.001 0.001 7.466 7.478 arnoldi_normal_ev 4 7.8 0.003 0.009 7.466 7.477 build_subspace 16 8.4 0.026 0.035 6.916 6.926 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.401 5.609 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.305 5.400 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.062 5.313 compute_matrix_preconditioner 1 6.0 0.002 0.002 5.262 5.270 dbcsr_matrix_vector_mult_local 304 10.0 4.853 5.130 4.855 5.132 acc_transpose_blocks 444 9.7 0.002 0.002 4.112 4.706 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 4.077 4.671 jit_kernel_transpose 1 13.0 4.072 4.667 4.072 4.667 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.304 4.484 4.611 dbcsr_mm_accdrv_process_sort 1814 11.4 4.185 4.315 4.185 4.315 ls_scf_post 1 4.0 0.000 0.000 3.713 3.727 make_images_sizes 222 9.7 0.000 0.000 1.480 3.609 mp_alltoall_i44 222 10.7 1.480 3.609 1.480 3.609 ls_scf_store_result 1 5.0 0.000 0.000 3.431 3.442 calculate_norms 792 9.8 3.240 3.275 3.240 3.275 dbcsr_finalize 304 7.8 0.082 0.090 3.097 3.168 mp_allgather_i34 111 8.7 0.991 3.114 0.991 3.114 dbcsr_merge_all 275 8.9 0.894 0.920 2.878 2.947 dbcsr_complete_redistribute 5 7.6 1.430 1.468 2.810 2.940 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.915 2.915 dbcsr_data_release 12724 10.6 2.330 2.850 2.330 2.850 matrix_ls_to_qs 2 6.0 0.000 0.000 2.458 2.603 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.586 2.589 rebuild_ks_matrix 3 7.3 0.000 0.000 2.519 2.523 qs_ks_build_kohn_sham_matrix 3 8.3 0.016 0.019 2.519 2.523 dbcsr_sort_data 325 11.1 2.443 2.496 2.443 2.496 mp_sum_l 887 5.1 1.517 2.444 1.517 2.444 dbcsr_new_transposed 4 7.5 0.243 0.252 2.386 2.408 mp_alltoall_d11v 48 9.2 2.198 2.292 2.198 2.292 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=113.201000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6838.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b261977e9e408dd060ac96be327c7e498359dcdc_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 589.856768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 3860318. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.151 283.190 283.192 qs_mol_dyn_low 1 2.0 0.006 0.047 281.877 281.899 qs_forces 5 3.8 0.006 0.028 281.696 281.699 qs_energies 5 4.8 0.002 0.040 278.216 278.257 scf_env_do_scf 5 5.8 0.000 0.001 259.519 259.522 scf_env_do_scf_inner_loop 105 6.6 0.004 0.029 225.149 225.163 qs_scf_new_mos 105 7.6 0.001 0.001 176.238 176.482 qs_scf_loop_do_ot 105 8.6 0.001 0.001 176.238 176.481 ot_scf_mini 105 9.6 0.003 0.004 166.341 166.532 dbcsr_multiply_generic 1445 12.2 0.127 0.134 138.197 138.955 multiply_cannon 1445 13.2 0.277 0.290 117.205 119.346 velocity_verlet 4 3.0 0.001 0.025 116.794 116.795 multiply_cannon_loop 1445 14.2 2.394 2.522 114.794 116.683 qs_ot_get_p 112 10.4 0.001 0.001 74.163 74.468 qs_ot_p2m_diag 40 11.0 0.020 0.031 62.201 62.272 ot_mini 105 10.6 0.001 0.012 61.658 61.778 cp_dbcsr_syevd 40 12.0 0.002 0.003 58.645 58.647 cp_fm_syevd 40 13.0 0.000 0.000 52.208 52.380 mp_waitall_1 488190 16.1 37.057 46.718 37.057 46.718 cp_fm_redistribute_end 40 14.0 23.383 46.630 23.393 46.636 cp_fm_syevd_base 40 14.0 23.224 46.475 23.224 46.475 multiply_cannon_multrec 69360 15.2 31.450 37.080 39.171 44.886 qs_ot_get_derivative 55 11.6 0.001 0.004 39.943 40.106 multiply_cannon_metrocomm3 69360 15.2 0.195 0.205 25.841 35.562 multiply_cannon_sync_h2d 69360 15.2 30.302 34.387 30.302 34.387 init_scf_loop 7 6.6 0.002 0.048 34.320 34.327 rebuild_ks_matrix 110 8.4 0.000 0.000 32.884 33.068 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.060 32.884 33.068 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.102 30.270 prepare_preconditioner 7 7.6 0.000 0.001 28.582 28.614 make_preconditioner 7 8.6 0.001 0.025 28.582 28.614 qs_rho_update_rho_low 110 7.6 0.003 0.060 23.303 23.660 calculate_rho_elec 110 8.6 0.030 0.034 23.300 23.659 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.016 23.318 apply_single 62 13.6 0.000 0.000 23.016 23.318 ot_new_cg_direction 55 11.6 0.001 0.002 20.983 20.985 make_full_inverse_cholesky 7 9.6 0.000 0.009 19.985 20.046 rs_pw_transfer 690 11.5 0.011 0.012 18.586 19.992 density_rs2pw 110 9.6 0.006 0.007 16.852 18.282 init_scf_run 5 5.8 0.000 0.001 15.797 15.798 scf_env_initial_rho_setup 5 6.8 0.002 0.017 15.797 15.798 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.231 15.381 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.363 14.455 pw_transfer 1645 12.4 0.086 0.111 13.598 13.849 fft_wrap_pw1pw2 1425 13.5 0.014 0.018 13.452 13.709 cp_fm_cholesky_invert 7 10.6 13.614 13.628 13.614 13.628 mp_sum_l 4764 12.2 12.515 13.244 12.515 13.244 calculate_dm_sparse 110 9.5 0.000 0.001 11.494 11.723 sum_up_and_integrate 60 10.3 0.028 0.035 11.662 11.678 integrate_v_rspace 60 11.3 0.002 0.010 11.634 11.649 fft_wrap_pw1pw2_240 915 15.0 1.239 1.336 11.190 11.392 check_diag 80 13.5 8.595 8.898 11.146 11.319 qs_vxc_create 110 10.4 0.003 0.049 11.044 11.107 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 10.682 10.751 calculate_first_density_matrix 1 7.0 0.000 0.006 10.294 10.314 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 10.293 10.310 acc_transpose_blocks 69360 15.2 0.359 0.381 9.648 10.062 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.832 9.888 fft3d_pb 915 16.0 2.390 2.697 9.194 9.552 multiply_cannon_metrocomm1 69360 15.2 0.092 0.099 4.563 8.705 make_m2s 2890 13.2 0.082 0.091 8.091 8.577 make_images 2890 14.2 0.244 0.264 7.980 8.467 dbcsr_mm_accdrv_process 154766 15.8 4.001 4.197 7.591 8.401 make_full_single_inverse 7 9.6 0.002 0.069 8.223 8.256 acc_transpose_blocks_kernels 69360 16.2 0.857 0.897 7.726 8.045 xc_rho_set_and_dset_create 110 12.4 0.078 0.154 7.633 7.890 xc_vxc_pw_create 60 11.3 0.039 0.067 7.516 7.579 potential_pw2rs 60 12.3 0.003 0.003 7.432 7.471 jit_kernel_transpose 5 15.0 6.870 7.151 6.870 7.151 rs_pw_transfer_RS2PW_30 110 11.6 1.591 1.668 6.813 7.054 mp_alltoall_z22v 2340 17.7 6.314 6.775 6.314 6.775 xc_pw_derive 510 13.4 0.006 0.007 6.701 6.773 mp_sendrecv_dv 168740 12.6 6.464 6.562 6.464 6.562 cp_fm_cholesky_decompose 14 10.2 5.808 5.816 5.808 5.816 mp_waitany 7680 13.5 4.304 5.780 4.304 5.780 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=283.192000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.200000, yerr=3.487119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b261977e9e408dd060ac96be327c7e498359dcdc Summary: empty Status: OK