=== 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: b0e0e430d9aee02af32c52f46e573f4f07a4ec1f ################# 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 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.04.2023) # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/01 job id: 46622917 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/02 job id: 46622919 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/03 job id: 46622920 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/04 job id: 46622921 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/05 job id: 46622922 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/06 job id: 46622923 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/07 job id: 46622924 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/08 job id: 46622925 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/09 job id: 46622926 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/10 job id: 46622927 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/11 job id: 46622928 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/12 job id: 46622929 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/13 job id: 46622930 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/14 job id: 46622931 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/15 job id: 46622932 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/16 job id: 46622933 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/17 job id: 46622934 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/18 job id: 46622935 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/19 job id: 46622936 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/20 job id: 46622937 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/21 job id: 46622938 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/22 job id: 46622939 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/23 job id: 46622940 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/24 job id: 46622941 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/25 job id: 46622942 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/26 job id: 46622943 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/27 job id: 46622944 --- 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/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 0.024 0.035 136.629 136.629 farming_run 1 2.0 136.111 136.112 136.593 136.595 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465504E+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 0.009 0.026 116.386 116.386 qs_energies 1 2.0 0.000 0.000 116.151 116.153 mp2_main 1 3.0 0.000 0.000 113.716 113.718 mp2_gpw_main 1 4.0 0.019 0.025 112.430 112.432 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.009 93.382 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.436 55.810 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.759 47.083 get_2c_integrals 1 6.0 0.008 0.010 36.986 37.400 integrate_v_rspace 273 8.0 0.439 0.452 25.114 30.231 pw_transfer 6555 10.6 0.376 0.385 27.364 27.813 grid_integrate_task_list 273 9.0 20.917 26.536 20.917 26.536 fft_wrap_pw1pw2 5465 11.4 0.045 0.047 26.064 26.495 fft_wrap_pw1pw2_100 2178 12.4 1.164 1.244 23.581 24.016 rpa_ri_compute_en 1 5.0 0.019 0.021 19.311 19.455 compute_2c_integrals 1 7.0 0.002 0.002 19.291 19.293 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.780 18.914 mp2_eri_2c_integrate_gpw 1 9.0 2.395 2.439 18.778 18.911 cp_fm_cholesky_decompose 12 8.2 17.870 18.323 17.870 18.323 cholesky_decomp 1 7.0 0.000 0.000 16.541 16.983 fft3d_s 5443 13.4 16.154 16.618 16.176 16.640 ao_to_mo_and_store_B_mult_1 272 7.0 10.845 15.554 10.845 15.554 calculate_wavefunction 272 8.0 5.436 5.584 12.574 13.204 rpa_num_int 1 6.0 0.000 0.000 10.863 10.863 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.836 10.853 calc_mat_Q 8 8.0 0.000 0.000 9.473 9.556 contract_S_to_Q 8 9.0 0.000 0.000 8.895 8.981 parallel_gemm_fm 14 9.1 0.000 0.000 8.485 8.582 parallel_gemm_fm_cosma 14 10.1 8.484 8.582 8.484 8.582 calc_potential_gpw 544 9.5 0.005 0.005 8.210 8.580 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.149 8.369 potential_pw2rs 545 10.0 0.107 0.108 7.667 8.317 create_integ_mat 1 6.0 0.022 0.026 7.980 7.980 collocate_single_gaussian 272 10.0 0.039 0.042 7.419 7.670 array2fm 1 7.0 0.000 0.000 6.797 7.281 pw_scatter_s 2720 13.7 4.456 4.601 4.456 4.601 pw_gather_s 2722 13.2 3.879 4.255 3.879 4.255 array2fm_buffer_send 1 8.0 2.893 3.140 2.893 3.140 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.428505, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 0.031 0.042 399.234 399.237 farming_run 1 2.0 398.181 398.218 399.179 399.183 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.221030E+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 7383115. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.037 210.152 210.153 qs_energies 1 2.0 0.003 0.008 209.898 209.908 scf_env_do_scf 1 3.0 0.000 0.000 107.270 107.270 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.236 106.243 rebuild_ks_matrix 4 6.0 0.000 0.000 106.234 106.242 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.066 106.234 106.242 hfx_ks_matrix 4 8.0 0.001 0.001 105.787 105.792 integrate_four_center 4 9.0 0.143 0.456 105.787 105.791 mp2_main 1 3.0 0.017 0.050 102.313 102.326 mp2_gpw_main 1 4.0 0.052 0.099 101.327 101.358 integrate_four_center_main 4 10.0 0.130 0.543 96.705 100.120 integrate_four_center_bin 267 11.0 96.576 99.910 96.576 99.910 init_scf_loop 1 4.0 0.000 0.000 92.919 92.920 mp2_ri_gpw_compute_in 1 5.0 0.088 0.136 74.812 75.902 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.019 54.415 55.490 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.164 42.245 47.316 integrate_v_rspace 95 8.0 0.397 0.576 28.665 33.541 pw_transfer 2240 10.6 0.145 0.166 29.919 30.319 ao_to_mo_and_store_B_mult_1 91 7.0 10.496 29.505 10.496 29.505 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.945 29.376 grid_integrate_task_list 95 9.0 23.899 28.985 23.899 28.985 mp2_ri_gpw_compute_en 1 5.0 0.057 0.068 26.276 27.966 fft_wrap_pw1pw2_100 730 12.4 1.260 1.403 26.627 27.152 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.837 1.884 24.510 24.520 get_2c_integrals 1 6.0 0.004 0.043 20.278 20.323 compute_2c_integrals 1 7.0 0.004 0.013 19.242 19.272 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.811 19.100 mp2_eri_2c_integrate_gpw 1 9.0 1.739 1.864 18.810 19.099 fft3d_s 1823 13.4 18.438 18.903 18.451 18.916 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.348 14.348 calculate_wavefunction 91 8.0 2.042 2.092 9.761 9.973 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.582 8.722 9.211 potential_pw2rs 186 10.0 0.033 0.035 8.684 9.197 local_gemm 172 8.0 8.163 8.633 8.163 8.633 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.213 8.598 collocate_single_gaussian 91 10.0 0.018 0.040 7.861 8.100 calc_potential_gpw 182 9.5 0.002 0.002 7.887 8.076 mp2_ri_gpw_compute_en_comm 22 7.0 0.537 0.555 7.547 7.976 mp_sync 37 10.5 3.665 6.719 3.665 6.719 mp2_ri_gpw_compute_en_ener 172 7.0 6.359 6.422 6.359 6.422 mp_sendrecv_dm3 2068 8.0 5.481 5.888 5.481 5.888 pw_gather_s 912 13.2 4.907 5.298 4.907 5.298 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.326070, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1508.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 448.720896E+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 588782. 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 0.013 0.030 52.372 52.374 qs_mol_dyn_low 1 2.0 0.006 0.031 52.101 52.109 qs_forces 11 3.9 0.002 0.002 51.984 51.985 qs_energies 11 4.9 0.001 0.001 50.515 50.531 scf_env_do_scf 11 5.9 0.001 0.001 44.576 44.576 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.522 42.522 qs_scf_new_mos 108 7.5 0.000 0.001 32.498 32.802 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.497 32.802 dbcsr_multiply_generic 2286 12.5 0.093 0.095 32.347 32.734 ot_scf_mini 108 9.5 0.002 0.002 30.898 31.089 multiply_cannon 2286 13.5 0.180 0.188 25.391 26.963 multiply_cannon_loop 2286 14.5 1.462 1.550 24.713 26.298 velocity_verlet 10 3.0 0.001 0.005 25.407 25.410 ot_mini 108 10.5 0.001 0.001 18.533 18.784 qs_ot_get_derivative 108 11.5 0.001 0.001 15.687 15.866 mp_waitall_1 245248 16.5 7.949 13.919 7.949 13.919 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.698 12.345 multiply_cannon_multrec 54864 15.5 4.275 6.728 7.539 10.949 qs_ot_get_p 119 10.4 0.001 0.001 7.879 8.138 rebuild_ks_matrix 119 8.3 0.000 0.000 7.830 7.963 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.829 7.963 multiply_cannon_sync_h2d 54864 15.5 6.020 7.245 6.020 7.245 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.895 7.015 mp_sum_l 7287 12.8 5.052 6.585 5.052 6.585 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.465 5.899 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.145 5.257 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.182 5.227 init_scf_run 11 5.9 0.000 0.001 4.686 4.687 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.686 4.686 sum_up_and_integrate 119 10.3 0.012 0.014 4.474 4.481 integrate_v_rspace 119 11.3 0.002 0.003 4.462 4.470 dbcsr_mm_accdrv_process 76910 16.1 1.140 1.802 3.186 4.467 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.447 4.448 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.139 4.288 calculate_rho_elec 119 8.7 0.012 0.017 4.138 4.288 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.251 4.252 cp_fm_redistribute_end 50 14.0 2.169 4.223 2.175 4.226 cp_fm_diag_elpa_base 50 14.0 2.045 4.121 2.049 4.128 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.856 3.049 apply_single 119 13.6 0.000 0.000 2.856 3.048 calculate_dm_sparse 119 9.5 0.000 0.001 2.807 2.948 rs_pw_transfer 974 11.9 0.012 0.014 2.685 2.825 multiply_cannon_metrocomm1 54864 15.5 0.052 0.058 1.581 2.682 jit_kernel_multiply 13 15.8 1.986 2.581 1.986 2.581 ot_diis_step 108 11.5 0.006 0.006 2.580 2.580 calculate_first_density_matrix 1 7.0 0.000 0.001 2.525 2.530 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.318 2.375 density_rs2pw 119 9.7 0.004 0.005 2.165 2.321 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.280 2.282 acc_transpose_blocks 54864 15.5 0.224 0.246 1.702 2.140 grid_integrate_task_list 119 12.3 2.023 2.123 2.023 2.123 wfi_extrapolate 11 7.9 0.001 0.001 2.101 2.101 init_scf_loop 11 6.9 0.000 0.000 2.037 2.038 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.991 2.038 mp_sum_d 4135 12.0 1.251 1.888 1.251 1.888 potential_pw2rs 119 12.3 0.004 0.004 1.828 1.836 pw_transfer 1439 11.6 0.052 0.056 1.705 1.773 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.630 1.699 make_m2s 4572 13.5 0.053 0.056 1.593 1.650 make_images 4572 14.5 0.133 0.138 1.511 1.568 mp_alltoall_d11v 2130 13.8 1.241 1.474 1.241 1.474 mp_waitany 12084 13.8 1.242 1.444 1.242 1.444 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.365 1.383 grid_collocate_task_list 119 9.7 1.288 1.360 1.288 1.360 fft3d_ps 1201 14.6 0.373 0.479 1.273 1.339 fft_wrap_pw1pw2_140 487 13.2 0.183 0.200 1.262 1.330 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.681 1.077 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.374000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 484.061184E+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 1030290. 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 0.030 0.057 38.651 38.653 qs_mol_dyn_low 1 2.0 0.003 0.003 38.282 38.288 qs_forces 11 3.9 0.002 0.002 38.202 38.203 qs_energies 11 4.9 0.003 0.016 36.419 36.423 scf_env_do_scf 11 5.9 0.001 0.001 31.172 31.172 scf_env_do_scf_inner_loop 108 6.5 0.002 0.008 28.741 28.742 dbcsr_multiply_generic 2286 12.5 0.100 0.103 21.239 21.588 qs_scf_new_mos 108 7.5 0.001 0.001 19.732 19.988 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.731 19.988 ot_scf_mini 108 9.5 0.003 0.004 18.855 19.033 multiply_cannon 2286 13.5 0.206 0.216 16.459 18.006 velocity_verlet 10 3.0 0.012 0.023 17.934 17.935 multiply_cannon_loop 2286 14.5 0.890 0.965 15.360 16.729 ot_mini 108 10.5 0.001 0.001 11.604 11.843 mp_waitall_1 200699 16.5 5.561 11.004 5.561 11.004 multiply_cannon_metrocomm3 27432 15.5 0.066 0.069 4.181 9.633 qs_ot_get_derivative 108 11.5 0.001 0.001 9.151 9.332 multiply_cannon_multrec 27432 15.5 1.974 4.379 6.010 8.821 rebuild_ks_matrix 119 8.3 0.000 0.000 7.198 7.333 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.198 7.333 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.271 6.394 dbcsr_mm_accdrv_process 47894 16.0 3.078 5.331 3.967 5.850 qs_ot_get_p 119 10.4 0.001 0.002 4.396 4.634 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.578 4.427 sum_up_and_integrate 119 10.3 0.024 0.027 4.172 4.187 integrate_v_rspace 119 11.3 0.002 0.003 4.148 4.163 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.047 4.124 apply_single 119 13.6 0.000 0.000 3.047 4.124 init_scf_run 11 5.9 0.000 0.001 4.003 4.004 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.002 4.004 mp_sum_l 7287 12.8 1.995 3.950 1.995 3.950 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.688 3.721 calculate_rho_elec 119 8.7 0.021 0.024 3.687 3.720 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.984 3.002 multiply_cannon_sync_h2d 27432 15.5 2.217 2.920 2.217 2.920 rs_pw_transfer 974 11.9 0.023 0.037 2.503 2.901 make_m2s 4572 13.5 0.052 0.054 2.352 2.569 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.561 2.561 calculate_first_density_matrix 1 7.0 0.000 0.003 2.523 2.525 make_images 4572 14.5 0.200 0.235 2.264 2.480 density_rs2pw 119 9.7 0.004 0.004 2.018 2.442 init_scf_loop 11 6.9 0.001 0.005 2.412 2.414 ot_diis_step 108 11.5 0.010 0.011 2.405 2.405 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.252 2.253 cp_fm_redistribute_end 50 14.0 1.141 2.228 1.145 2.231 cp_fm_diag_elpa_base 50 14.0 1.054 2.137 1.082 2.179 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.066 2.161 calculate_dm_sparse 119 9.5 0.000 0.001 2.080 2.158 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.000 2.003 pw_transfer 1439 11.6 0.065 0.069 1.935 1.978 grid_integrate_task_list 119 12.3 1.834 1.939 1.834 1.939 potential_pw2rs 119 12.3 0.006 0.006 1.908 1.925 jit_kernel_multiply 9 16.1 0.837 1.895 0.837 1.895 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.844 1.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.790 1.830 acc_transpose_blocks 27432 15.5 0.108 0.113 1.217 1.565 prepare_preconditioner 11 7.9 0.000 0.000 1.508 1.534 make_preconditioner 11 8.9 0.000 0.002 1.508 1.534 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.484 1.527 make_images_data 4572 15.5 0.045 0.051 1.104 1.517 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.409 1.464 fft3d_ps 1201 14.6 0.519 0.573 1.395 1.431 wfi_extrapolate 11 7.9 0.001 0.001 1.429 1.429 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.955 1.414 grid_collocate_task_list 119 9.7 1.230 1.370 1.230 1.370 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.308 1.315 mp_allgather_i34 2286 14.5 0.532 1.269 0.532 1.269 mp_alltoall_d11v 2130 13.8 1.129 1.237 1.129 1.237 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.158 1.208 mp_sum_d 4135 12.0 0.592 1.045 0.592 1.045 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.984 0.997 mp_waitany 5720 13.7 0.537 0.992 0.537 0.992 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.947 0.948 acc_transpose_blocks_kernels 27432 16.5 0.181 0.268 0.687 0.946 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.552 0.946 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.653000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=461.181818, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 517.484544E+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 2130764. 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 0.027 0.052 32.270 32.271 qs_mol_dyn_low 1 2.0 0.003 0.003 31.889 31.896 qs_forces 11 3.9 0.002 0.002 31.790 31.792 qs_energies 11 4.9 0.002 0.009 30.220 30.222 scf_env_do_scf 11 5.9 0.001 0.001 25.200 25.200 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 22.680 22.681 dbcsr_multiply_generic 2286 12.5 0.095 0.097 16.416 16.496 velocity_verlet 10 3.0 0.001 0.001 15.017 15.019 qs_scf_new_mos 108 7.5 0.001 0.001 14.610 14.629 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.610 14.628 multiply_cannon 2286 13.5 0.193 0.202 13.282 14.130 ot_scf_mini 108 9.5 0.002 0.003 13.899 13.915 multiply_cannon_loop 2286 14.5 0.629 0.663 12.515 13.356 ot_mini 108 10.5 0.001 0.001 8.567 8.585 multiply_cannon_multrec 18288 15.5 1.957 2.917 7.027 7.266 qs_ot_get_derivative 108 11.5 0.001 0.001 7.091 7.108 rebuild_ks_matrix 119 8.3 0.000 0.000 6.306 6.322 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.015 6.305 6.321 dbcsr_mm_accdrv_process 38222 16.0 4.118 5.362 4.987 5.851 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.572 5.588 sum_up_and_integrate 119 10.3 0.030 0.031 3.908 3.914 integrate_v_rspace 119 11.3 0.003 0.003 3.878 3.887 init_scf_run 11 5.9 0.000 0.001 3.849 3.849 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.848 3.849 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.385 3.396 calculate_rho_elec 119 8.7 0.031 0.031 3.384 3.396 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.697 3.326 mp_waitall_1 158411 16.6 2.434 3.279 2.434 3.279 qs_ot_get_p 119 10.4 0.001 0.001 3.198 3.224 calculate_first_density_matrix 1 7.0 0.000 0.002 2.661 2.661 init_scf_loop 11 6.9 0.001 0.005 2.504 2.504 rs_pw_transfer 974 11.9 0.009 0.010 2.192 2.484 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.032 2.339 apply_single 119 13.6 0.000 0.000 2.032 2.338 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.184 2.189 density_rs2pw 119 9.7 0.004 0.004 1.905 2.181 multiply_cannon_metrocomm3 18288 15.5 0.043 0.044 1.405 2.166 jit_kernel_multiply 10 16.0 0.818 2.080 0.818 2.080 calculate_dm_sparse 119 9.5 0.000 0.001 1.935 1.951 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.927 1.928 grid_integrate_task_list 119 12.3 1.799 1.890 1.799 1.890 pw_transfer 1439 11.6 0.065 0.070 1.867 1.875 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.863 1.865 make_m2s 4572 13.5 0.044 0.046 1.701 1.856 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.775 1.785 make_images 4572 14.5 0.190 0.201 1.616 1.769 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.712 1.717 prepare_preconditioner 11 7.9 0.000 0.001 1.713 1.715 make_preconditioner 11 8.9 0.000 0.002 1.713 1.715 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.688 1.700 cp_fm_diag_elpa_base 50 14.0 1.665 1.678 1.686 1.699 potential_pw2rs 119 12.3 0.007 0.009 1.676 1.681 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.564 1.651 multiply_cannon_sync_h2d 18288 15.5 1.400 1.575 1.400 1.575 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.499 1.509 ot_diis_step 108 11.5 0.011 0.011 1.456 1.456 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.435 1.445 acc_transpose_blocks 18288 15.5 0.074 0.076 1.349 1.429 mp_sum_l 7287 12.8 1.061 1.411 1.061 1.411 grid_collocate_task_list 119 9.7 1.211 1.360 1.211 1.360 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.314 1.317 fft3d_ps 1201 14.6 0.529 0.546 1.263 1.272 wfi_extrapolate 11 7.9 0.001 0.001 1.141 1.141 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.910 0.992 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.947 0.948 make_images_data 4572 15.5 0.045 0.049 0.768 0.918 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.855 0.879 mp_waitany 9880 13.7 0.518 0.823 0.518 0.823 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.649 0.822 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.805 0.806 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.123 0.510 0.797 jit_kernel_transpose 5 15.6 0.701 0.783 0.701 0.783 mp_alltoall_d11v 2130 13.8 0.638 0.753 0.638 0.753 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.651 0.746 cp_fm_cholesky_invert 11 10.9 0.698 0.702 0.698 0.702 mp_alltoall_z22v 1201 16.6 0.594 0.659 0.594 0.659 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.271000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=491.454545, yerr=1.827250 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 551.514112E+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 0.012 0.029 34.829 34.830 qs_mol_dyn_low 1 2.0 0.003 0.003 34.639 34.646 qs_forces 11 3.9 0.002 0.002 34.573 34.576 qs_energies 11 4.9 0.001 0.001 32.860 32.865 scf_env_do_scf 11 5.9 0.000 0.001 27.928 27.929 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.586 24.587 dbcsr_multiply_generic 2286 12.5 0.098 0.103 18.012 18.205 velocity_verlet 10 3.0 0.001 0.001 17.591 17.592 qs_scf_new_mos 108 7.5 0.001 0.001 16.189 16.234 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.188 16.234 ot_scf_mini 108 9.5 0.002 0.003 15.264 15.315 multiply_cannon 2286 13.5 0.231 0.275 14.554 14.998 multiply_cannon_loop 2286 14.5 0.931 0.962 13.628 13.969 ot_mini 108 10.5 0.001 0.001 9.315 9.382 multiply_cannon_multrec 27432 15.5 2.339 2.988 8.592 8.973 qs_ot_get_derivative 108 11.5 0.001 0.001 7.516 7.570 dbcsr_mm_accdrv_process 47916 15.9 5.352 6.757 6.160 7.324 rebuild_ks_matrix 119 8.3 0.000 0.000 6.536 6.590 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.535 6.590 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.800 5.847 sum_up_and_integrate 119 10.3 0.035 0.038 3.785 3.793 integrate_v_rspace 119 11.3 0.003 0.003 3.750 3.758 init_scf_run 11 5.9 0.000 0.001 3.615 3.616 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.615 3.615 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.506 3.534 calculate_rho_elec 119 8.7 0.040 0.046 3.505 3.533 qs_ot_get_p 119 10.4 0.001 0.001 3.356 3.430 init_scf_loop 11 6.9 0.000 0.000 3.323 3.323 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.730 3.111 prepare_preconditioner 11 7.9 0.000 0.000 2.489 2.497 make_preconditioner 11 8.9 0.000 0.000 2.489 2.497 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.098 2.427 calculate_first_density_matrix 1 7.0 0.000 0.000 2.262 2.263 mp_waitall_1 137007 16.6 1.883 2.256 1.883 2.256 make_m2s 4572 13.5 0.054 0.057 2.098 2.210 rs_pw_transfer 974 11.9 0.009 0.009 1.970 2.157 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.020 2.150 apply_single 119 13.6 0.000 0.000 2.020 2.149 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.112 2.120 density_rs2pw 119 9.7 0.004 0.004 1.927 2.106 calculate_dm_sparse 119 9.5 0.000 0.000 2.053 2.102 make_images 4572 14.5 0.269 0.331 1.990 2.100 pw_transfer 1439 11.6 0.065 0.070 1.985 2.020 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.892 1.931 grid_integrate_task_list 119 12.3 1.827 1.921 1.827 1.921 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.875 1.901 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.805 1.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.785 1.786 ot_diis_step 108 11.5 0.012 0.012 1.756 1.756 jit_kernel_multiply 9 16.1 0.748 1.667 0.748 1.667 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.623 1.636 fft_wrap_pw1pw2_140 487 13.2 0.288 0.299 1.561 1.603 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.559 1.569 cp_fm_diag_elpa_base 50 14.0 1.525 1.543 1.557 1.567 potential_pw2rs 119 12.3 0.009 0.010 1.538 1.541 acc_transpose_blocks 27432 15.5 0.109 0.110 1.480 1.518 fft3d_ps 1201 14.6 0.555 0.607 1.332 1.361 grid_collocate_task_list 119 9.7 1.221 1.341 1.221 1.341 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.883 1.311 wfi_extrapolate 11 7.9 0.001 0.001 1.305 1.305 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.253 mp_sum_l 7287 12.8 0.930 1.178 0.930 1.178 cp_fm_upper_to_full 72 14.2 0.814 1.161 0.814 1.161 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.088 1.110 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.069 1.071 multiply_cannon_sync_h2d 27432 15.5 0.995 1.051 0.995 1.051 dbcsr_complete_redistribute 329 12.2 0.119 0.144 0.743 1.017 make_images_data 4572 15.5 0.044 0.048 0.811 0.926 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.859 0.890 hybrid_alltoall_any 4725 16.4 0.062 0.153 0.697 0.881 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.870 mp_alltoall_d11v 2130 13.8 0.698 0.863 0.698 0.863 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.830 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.558 0.827 cp_fm_cholesky_invert 11 10.9 0.725 0.728 0.725 0.728 mp_alltoall_i22 627 13.8 0.424 0.710 0.424 0.710 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=34.830000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.181818, yerr=2.516064 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 600.809472E+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 0.028 0.045 27.706 27.706 qs_mol_dyn_low 1 2.0 0.003 0.003 27.439 27.447 qs_forces 11 3.9 0.002 0.002 27.374 27.375 qs_energies 11 4.9 0.001 0.001 25.668 25.671 scf_env_do_scf 11 5.9 0.000 0.001 20.894 20.894 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.470 18.470 velocity_verlet 10 3.0 0.001 0.001 14.142 14.145 dbcsr_multiply_generic 2286 12.5 0.092 0.096 11.914 11.983 qs_scf_new_mos 108 7.5 0.001 0.001 10.603 10.629 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.602 10.628 multiply_cannon 2286 13.5 0.231 0.237 9.506 9.967 ot_scf_mini 108 9.5 0.002 0.002 9.936 9.964 multiply_cannon_loop 2286 14.5 0.327 0.338 8.613 8.820 multiply_cannon_multrec 9144 15.5 1.632 1.900 5.813 6.040 rebuild_ks_matrix 119 8.3 0.000 0.000 5.888 5.911 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.888 5.911 ot_mini 108 10.5 0.001 0.001 5.560 5.592 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.243 5.263 qs_ot_get_derivative 108 11.5 0.001 0.001 4.282 4.309 dbcsr_mm_accdrv_process 12550 15.8 3.231 4.111 4.081 4.152 sum_up_and_integrate 119 10.3 0.038 0.041 3.606 3.610 integrate_v_rspace 119 11.3 0.003 0.003 3.568 3.574 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.478 3.484 calculate_rho_elec 119 8.7 0.060 0.061 3.477 3.484 init_scf_run 11 5.9 0.000 0.001 3.357 3.357 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.357 3.357 qs_ot_get_p 119 10.4 0.001 0.001 2.637 2.677 init_scf_loop 11 6.9 0.000 0.000 2.405 2.407 calculate_first_density_matrix 1 7.0 0.000 0.000 2.233 2.234 pw_transfer 1439 11.6 0.066 0.068 2.002 2.014 grid_integrate_task_list 119 12.3 1.849 1.924 1.849 1.924 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.909 1.922 density_rs2pw 119 9.7 0.004 0.004 1.788 1.921 make_m2s 4572 13.5 0.035 0.036 1.647 1.807 jit_kernel_multiply 10 15.9 0.812 1.785 0.812 1.785 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.754 1.757 rs_pw_transfer 974 11.9 0.008 0.008 1.613 1.749 calculate_dm_sparse 119 9.5 0.000 0.000 1.729 1.746 mp_waitall_1 115863 16.7 1.271 1.738 1.271 1.738 make_images 4572 14.5 0.267 0.302 1.558 1.716 prepare_preconditioner 11 7.9 0.000 0.000 1.681 1.686 make_preconditioner 11 8.9 0.000 0.000 1.681 1.686 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.573 1.598 fft_wrap_pw1pw2_140 487 13.2 0.365 0.374 1.578 1.590 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.565 1.566 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.510 1.511 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.474 1.489 grid_collocate_task_list 119 9.7 1.271 1.387 1.271 1.387 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.368 1.375 potential_pw2rs 119 12.3 0.010 0.011 1.348 1.351 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.280 1.288 cp_fm_diag_elpa_base 50 14.0 1.253 1.270 1.278 1.286 ot_diis_step 108 11.5 0.012 0.013 1.265 1.265 fft3d_ps 1201 14.6 0.558 0.569 1.243 1.253 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.212 1.213 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.143 1.161 apply_single 119 13.6 0.000 0.000 1.143 1.161 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.154 1.159 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.078 1.090 wfi_extrapolate 11 7.9 0.001 0.001 1.076 1.076 hybrid_alltoall_any 4725 16.4 0.063 0.175 0.760 0.991 make_images_data 4572 15.5 0.039 0.042 0.763 0.938 acc_transpose_blocks 9144 15.5 0.037 0.039 0.925 0.932 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.926 mp_alltoall_d11v 2130 13.8 0.760 0.838 0.760 0.838 cp_fm_cholesky_invert 11 10.9 0.826 0.829 0.826 0.829 multiply_cannon_sync_h2d 9144 15.5 0.711 0.791 0.711 0.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.765 0.768 qs_env_update_s_mstruct 11 6.9 0.001 0.002 0.667 0.716 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.699 0.709 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.686 0.689 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.342 0.642 mp_allgather_i34 2286 14.5 0.226 0.615 0.226 0.615 mp_alltoall_z22v 1201 16.6 0.559 0.593 0.559 0.593 jit_kernel_transpose 5 15.6 0.570 0.572 0.570 0.572 mp_waitany 5200 13.7 0.443 0.569 0.443 0.569 qs_create_task_list 11 7.9 0.000 0.000 0.540 0.565 generate_qs_task_list 11 8.9 0.188 0.209 0.540 0.565 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=27.706000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.636364, yerr=5.725108 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 743.010304E+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 0.050 0.066 45.259 45.260 qs_mol_dyn_low 1 2.0 0.003 0.003 44.971 44.978 qs_forces 11 3.9 0.002 0.002 44.908 44.909 qs_energies 11 4.9 0.001 0.001 42.834 42.838 scf_env_do_scf 11 5.9 0.001 0.001 36.956 36.956 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 28.299 28.300 velocity_verlet 10 3.0 0.001 0.001 25.670 25.676 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.737 19.038 qs_scf_new_mos 108 7.5 0.001 0.001 17.693 17.788 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.693 17.788 ot_scf_mini 108 9.5 0.002 0.002 16.522 16.614 multiply_cannon 2286 13.5 0.303 0.317 14.052 15.294 multiply_cannon_loop 2286 14.5 0.344 0.352 12.655 13.531 ot_mini 108 10.5 0.001 0.001 9.685 9.815 multiply_cannon_multrec 9144 15.5 3.416 4.768 8.555 8.640 init_scf_loop 11 6.9 0.000 0.000 8.627 8.628 rebuild_ks_matrix 119 8.3 0.000 0.000 7.817 7.978 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.816 7.977 qs_ot_get_derivative 108 11.5 0.001 0.001 7.530 7.623 prepare_preconditioner 11 7.9 0.000 0.000 7.594 7.610 make_preconditioner 11 8.9 0.000 0.000 7.594 7.610 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.062 7.479 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.064 7.208 dbcsr_mm_accdrv_process 12550 15.8 4.193 5.646 5.017 6.371 cp_fm_upper_to_full 72 14.2 3.257 4.727 3.257 4.727 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.695 4.701 calculate_rho_elec 119 8.7 0.118 0.120 4.695 4.701 sum_up_and_integrate 119 10.3 0.064 0.066 4.395 4.401 integrate_v_rspace 119 11.3 0.003 0.003 4.331 4.336 mp_waitall_1 94719 16.7 3.282 4.200 3.282 4.200 qs_ot_get_p 119 10.4 0.001 0.001 3.865 4.018 init_scf_run 11 5.9 0.000 0.001 3.823 3.823 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.823 3.823 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.650 3.128 make_m2s 4572 13.5 0.038 0.038 2.807 3.083 dbcsr_complete_redistribute 329 12.2 0.284 0.290 2.148 3.014 pw_transfer 1439 11.6 0.069 0.069 2.974 2.981 make_images 4572 14.5 0.351 0.381 2.687 2.962 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.875 2.883 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.788 2.640 density_rs2pw 119 9.7 0.004 0.004 2.595 2.625 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.326 2.557 apply_single 119 13.6 0.000 0.000 2.326 2.556 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.656 2.493 mp_alltoall_i22 627 13.8 1.539 2.392 1.539 2.392 fft_wrap_pw1pw2_140 487 13.2 0.616 0.621 2.368 2.378 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.527 2.372 qs_ot_p2m_diag 50 11.0 0.042 0.043 2.361 2.364 calculate_dm_sparse 119 9.5 0.000 0.000 2.252 2.274 mp_sum_l 7287 12.8 1.334 2.165 1.334 2.165 calculate_first_density_matrix 1 7.0 0.000 0.000 2.152 2.154 ot_diis_step 108 11.5 0.014 0.014 2.103 2.103 grid_integrate_task_list 119 12.3 2.037 2.094 2.037 2.094 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.056 2.057 rs_pw_transfer 974 11.9 0.009 0.009 1.961 1.992 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.937 1.986 fft3d_ps 1201 14.6 0.595 0.606 1.893 1.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.896 1.897 cp_fm_cholesky_invert 11 10.9 1.893 1.897 1.893 1.897 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.452 1.862 make_images_data 4572 15.5 0.043 0.045 1.455 1.861 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.767 1.768 potential_pw2rs 119 12.3 0.014 0.014 1.717 1.722 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.704 1.704 cp_fm_diag_elpa_base 50 14.0 1.549 1.601 1.702 1.702 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.584 1.610 wfi_extrapolate 11 7.9 0.001 0.001 1.604 1.604 grid_collocate_task_list 119 9.7 1.471 1.508 1.471 1.508 mp_alltoall_d11v 2130 13.8 1.304 1.334 1.304 1.334 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.275 1.316 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.237 1.247 mp_alltoall_z22v 1201 16.6 1.162 1.184 1.162 1.184 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.101 1.112 multiply_cannon_sync_h2d 9144 15.5 1.042 1.047 1.042 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.989 1.043 jit_kernel_multiply 6 15.5 0.796 1.011 0.796 1.011 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.965 0.980 qs_create_task_list 11 7.9 0.000 0.000 0.939 0.954 generate_qs_task_list 11 8.9 0.368 0.387 0.939 0.954 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.260000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=699.272727, yerr=12.884740 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 497.926144E+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 1187921. 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 0.035 0.072 82.549 82.553 qs_mol_dyn_low 1 2.0 0.006 0.026 82.121 82.130 qs_forces 11 3.9 0.003 0.003 81.983 81.986 qs_energies 11 4.9 0.002 0.014 79.085 79.096 scf_env_do_scf 11 5.9 0.001 0.002 70.179 70.182 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 64.717 64.718 dbcsr_multiply_generic 2055 12.4 0.107 0.110 51.085 51.301 qs_scf_new_mos 99 7.5 0.001 0.001 47.356 47.469 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.356 47.468 ot_scf_mini 99 9.5 0.002 0.003 44.902 44.980 multiply_cannon 2055 13.4 0.176 0.181 42.428 43.161 velocity_verlet 10 3.0 0.003 0.014 42.908 42.910 multiply_cannon_loop 2055 14.4 1.503 1.529 41.485 42.146 ot_mini 99 10.5 0.001 0.001 26.538 26.628 qs_ot_get_derivative 99 11.5 0.001 0.001 19.772 19.852 multiply_cannon_multrec 49320 15.4 12.445 13.116 17.305 17.957 rebuild_ks_matrix 110 8.3 0.000 0.001 14.560 14.658 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.559 14.658 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.722 12.808 mp_waitall_1 220248 16.4 11.026 11.827 11.026 11.827 multiply_cannon_sync_h2d 49320 15.4 10.282 10.928 10.282 10.928 qs_ot_get_p 110 10.4 0.001 0.002 9.785 9.874 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.431 7.971 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.189 7.659 apply_single 110 13.6 0.000 0.001 7.188 7.659 multiply_cannon_metrocomm3 49320 15.4 0.079 0.081 6.734 7.652 sum_up_and_integrate 110 10.3 0.036 0.043 7.081 7.108 integrate_v_rspace 110 11.3 0.003 0.004 7.044 7.080 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.624 6.774 calculate_rho_elec 110 8.6 0.021 0.026 6.624 6.773 init_scf_run 11 5.9 0.000 0.001 6.760 6.760 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.760 6.760 qs_ot_p2m_diag 48 11.0 0.012 0.021 6.676 6.720 ot_diis_step 99 11.5 0.005 0.006 6.551 6.552 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.723 5.723 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.603 5.655 init_scf_loop 11 6.9 0.001 0.004 5.436 5.438 mp_sum_l 6594 12.7 4.449 5.201 4.449 5.201 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.150 5.171 cp_fm_diag_elpa_base 48 14.0 5.137 5.159 5.149 5.170 dbcsr_mm_accdrv_process 87628 16.1 1.907 1.971 4.737 4.973 rs_pw_transfer 902 11.9 0.013 0.023 3.657 4.163 wfi_extrapolate 11 7.9 0.001 0.001 3.992 3.992 calculate_dm_sparse 110 9.5 0.001 0.001 3.825 3.962 density_rs2pw 110 9.6 0.004 0.005 3.351 3.913 make_m2s 4110 13.4 0.061 0.066 3.784 3.910 make_images 4110 14.4 0.176 0.189 3.688 3.818 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.492 3.599 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.549 3.553 grid_integrate_task_list 110 12.3 3.251 3.408 3.251 3.408 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.327 3.397 prepare_preconditioner 11 7.9 0.000 0.000 3.262 3.286 make_preconditioner 11 8.9 0.000 0.001 3.262 3.286 pw_transfer 1331 11.6 0.055 0.062 3.151 3.238 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.197 3.227 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.062 3.152 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.046 3.093 fft_wrap_pw1pw2_140 451 13.1 0.451 0.497 2.621 2.713 calculate_first_density_matrix 1 7.0 0.000 0.003 2.680 2.684 potential_pw2rs 110 12.3 0.006 0.007 2.622 2.648 jit_kernel_multiply 13 15.9 2.555 2.596 2.555 2.596 mp_alltoall_d11v 2046 13.8 2.109 2.549 2.109 2.549 mp_waitany 14300 13.8 1.832 2.479 1.832 2.479 grid_collocate_task_list 110 9.6 2.081 2.366 2.081 2.366 fft3d_ps 1111 14.6 0.792 0.875 2.258 2.327 acc_transpose_blocks 49320 15.4 0.206 0.215 2.064 2.110 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.968 1.992 mp_sum_d 3889 11.9 1.391 1.943 1.391 1.943 make_images_data 4110 15.4 0.042 0.046 1.747 1.884 cp_fm_cholesky_invert 11 10.9 1.796 1.800 1.796 1.800 hybrid_alltoall_any 4261 16.3 0.082 0.481 1.519 1.791 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.643 1.668 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.553000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.818182, yerr=2.442733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.378112E+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 1900912. 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 0.012 0.030 68.773 68.774 qs_mol_dyn_low 1 2.0 0.003 0.003 68.474 68.484 qs_forces 11 3.9 0.003 0.003 68.340 68.341 qs_energies 11 4.9 0.001 0.001 65.002 65.005 scf_env_do_scf 11 5.9 0.001 0.004 56.422 56.426 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.887 48.888 dbcsr_multiply_generic 2055 12.4 0.115 0.119 37.571 37.735 velocity_verlet 10 3.0 0.010 0.074 35.945 35.946 qs_scf_new_mos 99 7.5 0.001 0.001 32.650 32.793 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.649 32.792 multiply_cannon 2055 13.4 0.222 0.246 31.146 32.153 ot_scf_mini 99 9.5 0.003 0.003 31.002 31.131 multiply_cannon_loop 2055 14.4 0.915 0.932 29.916 30.726 ot_mini 99 10.5 0.001 0.001 18.083 18.212 multiply_cannon_multrec 24660 15.4 7.648 9.353 14.040 15.677 rebuild_ks_matrix 110 8.3 0.000 0.000 13.732 13.823 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.731 13.823 qs_ot_get_derivative 99 11.5 0.001 0.001 12.317 12.444 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.070 12.142 mp_waitall_1 176588 16.5 7.506 9.909 7.506 9.909 multiply_cannon_sync_h2d 24660 15.4 7.045 7.907 7.045 7.907 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.131 7.745 init_scf_loop 11 6.9 0.000 0.000 7.501 7.502 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.479 7.247 apply_single 110 13.6 0.000 0.001 6.479 7.247 sum_up_and_integrate 110 10.3 0.053 0.060 6.602 6.615 integrate_v_rspace 110 11.3 0.003 0.003 6.549 6.564 dbcsr_mm_accdrv_process 52282 16.1 4.554 5.619 6.232 6.555 qs_ot_get_p 110 10.4 0.001 0.001 6.317 6.509 init_scf_run 11 5.9 0.000 0.001 6.215 6.216 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.215 6.216 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.121 6.129 calculate_rho_elec 110 8.6 0.040 0.048 6.121 6.128 ot_diis_step 99 11.5 0.010 0.010 5.718 5.718 prepare_preconditioner 11 7.9 0.000 0.000 5.486 5.509 make_preconditioner 11 8.9 0.000 0.000 5.486 5.509 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.658 5.384 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.089 5.250 qs_ot_p2m_diag 48 11.0 0.028 0.043 4.483 4.503 make_m2s 4110 13.4 0.057 0.060 4.046 4.472 make_images 4110 14.4 0.397 0.442 3.938 4.361 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.057 4.057 density_rs2pw 110 9.6 0.005 0.005 3.258 3.802 pw_transfer 1331 11.6 0.066 0.074 3.599 3.731 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.492 3.627 rs_pw_transfer 902 11.9 0.014 0.028 2.953 3.527 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.500 3.512 cp_fm_diag_elpa_base 48 14.0 3.453 3.474 3.497 3.508 wfi_extrapolate 11 7.9 0.001 0.001 3.479 3.479 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.320 3.321 grid_integrate_task_list 110 12.3 3.158 3.313 3.158 3.313 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.123 3.194 fft_wrap_pw1pw2_140 451 13.1 0.520 0.534 3.001 3.135 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.996 3.032 calculate_dm_sparse 110 9.5 0.001 0.001 2.988 3.020 hybrid_alltoall_any 4261 16.3 0.103 0.447 2.013 2.769 make_images_data 4110 15.4 0.047 0.050 2.291 2.744 calculate_first_density_matrix 1 7.0 0.000 0.000 2.650 2.651 fft3d_ps 1111 14.6 1.106 1.330 2.466 2.615 cp_fm_cholesky_invert 11 10.9 2.514 2.521 2.514 2.521 grid_collocate_task_list 110 9.6 2.054 2.485 2.054 2.485 potential_pw2rs 110 12.3 0.008 0.009 2.398 2.413 mp_sum_l 6594 12.7 1.751 2.387 1.751 2.387 jit_kernel_multiply 11 16.2 1.327 2.287 1.327 2.287 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.923 1.940 mp_alltoall_d11v 2046 13.8 1.686 1.939 1.686 1.939 qs_energies_init_hamiltonians 11 5.9 0.002 0.008 1.847 1.847 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.817 1.837 mp_waitany 10164 13.8 1.249 1.757 1.249 1.757 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.788 1.684 acc_transpose_blocks 24660 15.4 0.110 0.112 1.576 1.616 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.571 1.583 mp_irecv_dv 57340 16.2 0.660 1.564 0.660 1.564 mp_allgather_i34 2055 14.4 0.562 1.547 0.562 1.547 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.217 0.964 1.537 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.367 1.473 dbcsr_complete_redistribute 325 12.2 0.241 0.311 1.119 1.384 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.774000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=553.909091, yerr=8.173733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 656.474112E+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 0.056 0.137 65.749 65.751 qs_mol_dyn_low 1 2.0 0.003 0.003 65.236 65.246 qs_forces 11 3.9 0.010 0.037 65.142 65.143 qs_energies 11 4.9 0.012 0.045 61.787 61.795 scf_env_do_scf 11 5.9 0.001 0.002 53.338 53.339 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 43.619 43.620 velocity_verlet 10 3.0 0.001 0.001 35.346 35.348 dbcsr_multiply_generic 2055 12.4 0.108 0.112 29.957 30.410 qs_scf_new_mos 99 7.5 0.001 0.001 27.794 27.889 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.793 27.889 ot_scf_mini 99 9.5 0.003 0.003 26.488 26.593 multiply_cannon 2055 13.4 0.210 0.220 22.759 24.472 multiply_cannon_loop 2055 14.4 0.614 0.631 21.290 22.533 ot_mini 99 10.5 0.001 0.001 14.752 14.880 rebuild_ks_matrix 110 8.3 0.000 0.000 12.918 13.108 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.917 13.108 mp_waitall_1 139946 16.5 8.077 11.634 8.077 11.634 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.385 11.562 multiply_cannon_multrec 16440 15.4 3.770 4.778 9.501 10.509 qs_ot_get_derivative 99 11.5 0.001 0.001 10.104 10.213 init_scf_loop 11 6.9 0.002 0.010 9.677 9.678 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.866 8.104 prepare_preconditioner 11 7.9 0.000 0.000 7.824 7.845 make_preconditioner 11 8.9 0.000 0.002 7.824 7.845 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.122 7.489 sum_up_and_integrate 110 10.3 0.061 0.068 6.973 6.987 integrate_v_rspace 110 11.3 0.005 0.022 6.911 6.926 qs_ot_get_p 110 10.4 0.001 0.001 6.721 6.882 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.350 6.361 calculate_rho_elec 110 8.6 0.059 0.060 6.349 6.360 init_scf_run 11 5.9 0.000 0.001 5.847 5.847 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.847 5.847 dbcsr_mm_accdrv_process 34862 16.1 4.600 5.280 5.592 5.724 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.137 5.612 apply_single 110 13.6 0.000 0.000 5.136 5.612 make_m2s 4110 13.4 0.050 0.051 4.596 5.118 make_images 4110 14.4 0.389 0.505 4.481 5.004 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.782 4.789 density_rs2pw 110 9.6 0.005 0.005 3.431 4.635 ot_diis_step 99 11.5 0.010 0.011 4.583 4.583 rs_pw_transfer 902 11.9 0.010 0.012 3.321 4.508 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.399 4.400 multiply_cannon_sync_h2d 16440 15.4 3.604 4.201 3.604 4.201 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.383 4.056 pw_transfer 1331 11.6 0.065 0.074 3.762 3.774 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.741 3.751 cp_fm_diag_elpa_base 48 14.0 3.668 3.700 3.739 3.750 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.655 3.667 grid_integrate_task_list 110 12.3 3.167 3.428 3.167 3.428 cp_fm_cholesky_invert 11 10.9 3.390 3.398 3.390 3.398 make_images_data 4110 15.4 0.043 0.047 2.643 3.247 wfi_extrapolate 11 7.9 0.001 0.001 3.218 3.218 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.302 3.072 fft_wrap_pw1pw2_140 451 13.1 0.636 0.645 3.046 3.062 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.874 2.876 mp_sum_l 6594 12.7 2.040 2.860 2.040 2.860 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.721 2.781 potential_pw2rs 110 12.3 0.011 0.012 2.683 2.699 calculate_dm_sparse 110 9.5 0.001 0.001 2.590 2.624 mp_waitany 17072 13.8 1.332 2.584 1.332 2.584 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.465 2.529 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.915 2.525 calculate_first_density_matrix 1 7.0 0.001 0.003 2.521 2.522 fft3d_ps 1111 14.6 1.089 1.099 2.478 2.488 grid_collocate_task_list 110 9.6 2.082 2.472 2.082 2.472 mp_irecv_dv 48980 15.7 0.844 2.398 0.844 2.398 mp_alltoall_d11v 2046 13.8 1.882 2.277 1.882 2.277 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.182 0.943 2.127 dbcsr_complete_redistribute 325 12.2 0.327 0.365 1.588 2.076 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.031 2.032 mp_allgather_i34 2055 14.4 0.796 1.939 0.796 1.939 cp_fm_cholesky_decompose 22 10.9 1.919 1.935 1.919 1.935 jit_kernel_multiply 10 16.4 0.604 1.895 0.604 1.895 cp_fm_upper_to_full 70 14.2 1.419 1.814 1.419 1.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.731 1.754 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.081 1.551 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.373 1.503 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.478 1.495 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.410 1.444 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.751000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.363636, yerr=8.824272 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 718.987264E+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 0.059 0.080 65.037 65.038 qs_mol_dyn_low 1 2.0 0.003 0.003 64.570 64.581 qs_forces 11 3.9 0.002 0.003 64.432 64.433 qs_energies 11 4.9 0.003 0.016 61.030 61.034 scf_env_do_scf 11 5.9 0.001 0.005 52.561 52.564 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.068 41.069 velocity_verlet 10 3.0 0.001 0.001 36.739 36.744 dbcsr_multiply_generic 2055 12.4 0.113 0.120 29.400 29.652 qs_scf_new_mos 99 7.5 0.001 0.001 26.066 26.185 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.066 26.184 ot_scf_mini 99 9.5 0.003 0.004 24.452 24.567 multiply_cannon 2055 13.4 0.243 0.260 22.563 23.863 multiply_cannon_loop 2055 14.4 0.880 0.907 21.147 21.795 multiply_cannon_multrec 24660 15.4 4.222 6.835 12.821 14.066 ot_mini 99 10.5 0.001 0.001 13.917 14.036 rebuild_ks_matrix 110 8.3 0.000 0.000 12.202 12.322 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.202 12.322 init_scf_loop 11 6.9 0.001 0.007 11.453 11.454 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.775 10.881 qs_ot_get_derivative 99 11.5 0.001 0.001 9.768 9.880 prepare_preconditioner 11 7.9 0.000 0.001 9.702 9.719 make_preconditioner 11 8.9 0.000 0.001 9.702 9.719 dbcsr_mm_accdrv_process 52304 16.0 6.826 8.730 8.453 9.424 make_full_inverse_cholesky 11 9.9 0.000 0.001 7.966 9.401 sum_up_and_integrate 110 10.3 0.068 0.071 6.498 6.512 integrate_v_rspace 110 11.3 0.003 0.003 6.431 6.444 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.064 6.074 calculate_rho_elec 110 8.6 0.078 0.081 6.064 6.074 mp_waitall_1 121746 16.5 4.134 5.906 4.134 5.906 qs_ot_get_p 110 10.4 0.001 0.001 5.546 5.717 init_scf_run 11 5.9 0.000 0.001 5.656 5.657 scf_env_initial_rho_setup 11 6.9 0.002 0.005 5.656 5.656 make_m2s 4110 13.4 0.060 0.062 5.287 5.643 make_images 4110 14.4 0.574 0.695 5.146 5.498 cp_fm_upper_to_full 70 14.2 3.309 4.791 3.309 4.791 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.002 4.134 apply_single 110 13.6 0.000 0.000 4.001 4.134 ot_diis_step 99 11.5 0.011 0.011 4.114 4.114 qs_ot_p2m_diag 48 11.0 0.054 0.063 3.816 3.832 pw_transfer 1331 11.6 0.065 0.076 3.731 3.762 dbcsr_complete_redistribute 325 12.2 0.421 0.469 2.571 3.688 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.625 3.659 density_rs2pw 110 9.6 0.004 0.005 3.001 3.515 multiply_cannon_sync_h2d 24660 15.4 3.196 3.454 3.196 3.454 grid_integrate_task_list 110 12.3 3.260 3.415 3.260 3.415 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.350 3.351 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.268 3.325 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.069 3.173 fft_wrap_pw1pw2_140 451 13.1 0.668 0.691 3.116 3.152 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.405 3.104 hybrid_alltoall_any 4261 16.3 0.120 0.455 2.275 3.056 make_images_data 4110 15.4 0.046 0.049 2.691 3.020 calculate_dm_sparse 110 9.5 0.001 0.001 2.979 3.011 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.944 3.004 wfi_extrapolate 11 7.9 0.001 0.001 2.970 2.970 rs_pw_transfer 902 11.9 0.010 0.011 2.371 2.895 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.861 2.863 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.823 2.833 cp_fm_diag_elpa_base 48 14.0 2.677 2.739 2.820 2.831 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.727 2.815 mp_alltoall_i22 605 13.7 1.613 2.763 1.613 2.763 cp_fm_cholesky_invert 11 10.9 2.628 2.636 2.628 2.636 calculate_first_density_matrix 1 7.0 0.000 0.002 2.593 2.595 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.517 2.557 grid_collocate_task_list 110 9.6 2.181 2.441 2.181 2.441 fft3d_ps 1111 14.6 1.086 1.115 2.385 2.404 jit_kernel_multiply 12 15.8 1.298 2.394 1.298 2.394 qs_energies_init_hamiltonians 11 5.9 0.007 0.023 2.292 2.294 potential_pw2rs 110 12.3 0.013 0.013 2.147 2.155 mp_alltoall_d11v 2046 13.8 1.724 1.944 1.724 1.944 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.673 1.705 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.594 1.690 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.671 1.683 cp_fm_cholesky_decompose 22 10.9 1.604 1.655 1.604 1.655 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.599 1.615 acc_transpose_blocks 24660 15.4 0.102 0.104 1.566 1.603 mp_allgather_i34 2055 14.4 0.510 1.560 0.510 1.560 mp_waitany 13376 13.8 1.072 1.546 1.072 1.546 multiply_cannon_metrocomm4 20550 15.4 0.060 0.062 0.857 1.532 mp_irecv_dv 62702 16.1 0.756 1.446 0.756 1.446 mp_sum_l 6594 12.7 0.884 1.417 0.884 1.417 rs_pw_transfer_RS2PW_140 121 11.5 0.170 0.178 0.830 1.360 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.221 1.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.038000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=683.363636, yerr=6.357140 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 831.275008E+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 0.031 54.467 54.468 qs_mol_dyn_low 1 2.0 0.003 0.003 54.197 54.208 qs_forces 11 3.9 0.003 0.003 54.121 54.122 qs_energies 11 4.9 0.001 0.001 50.444 50.448 scf_env_do_scf 11 5.9 0.001 0.001 42.184 42.184 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 34.621 34.621 velocity_verlet 10 3.0 0.001 0.001 30.741 30.745 dbcsr_multiply_generic 2055 12.4 0.104 0.109 22.554 22.679 qs_scf_new_mos 99 7.5 0.001 0.001 19.677 19.718 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.677 19.718 multiply_cannon 2055 13.4 0.245 0.256 17.176 18.574 ot_scf_mini 99 9.5 0.002 0.002 18.451 18.483 multiply_cannon_loop 2055 14.4 0.319 0.331 15.844 16.139 rebuild_ks_matrix 110 8.3 0.000 0.000 11.869 11.896 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.868 11.896 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.533 10.557 ot_mini 99 10.5 0.001 0.001 10.106 10.135 multiply_cannon_multrec 8220 15.4 3.215 4.469 7.435 8.347 mp_waitall_1 103326 16.6 5.896 7.610 5.896 7.610 init_scf_loop 11 6.9 0.000 0.000 7.518 7.518 sum_up_and_integrate 110 10.3 0.080 0.081 6.476 6.490 integrate_v_rspace 110 11.3 0.003 0.003 6.396 6.411 qs_ot_get_derivative 99 11.5 0.001 0.001 6.350 6.382 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.211 6.225 calculate_rho_elec 110 8.6 0.114 0.115 6.211 6.225 prepare_preconditioner 11 7.9 0.000 0.000 5.871 5.877 make_preconditioner 11 8.9 0.000 0.000 5.871 5.877 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.442 5.510 init_scf_run 11 5.9 0.000 0.001 5.114 5.114 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.114 5.114 dbcsr_mm_accdrv_process 17442 15.9 2.741 3.600 4.091 4.996 qs_ot_get_p 110 10.4 0.001 0.001 4.452 4.478 make_m2s 4110 13.4 0.039 0.040 4.195 4.456 make_images 4110 14.4 0.636 0.693 4.066 4.326 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.942 4.270 pw_transfer 1331 11.6 0.066 0.071 3.878 3.886 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.770 3.783 ot_diis_step 99 11.5 0.012 0.012 3.737 3.737 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.660 3.680 apply_single 110 13.6 0.000 0.000 3.660 3.680 grid_integrate_task_list 110 12.3 3.364 3.512 3.364 3.512 density_rs2pw 110 9.6 0.004 0.004 2.968 3.342 fft_wrap_pw1pw2_140 451 13.1 0.835 0.849 3.272 3.290 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.169 3.172 multiply_cannon_sync_h2d 8220 15.4 2.925 3.026 2.925 3.026 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.875 2.875 make_images_data 4110 15.4 0.038 0.043 2.389 2.827 cp_fm_cholesky_invert 11 10.9 2.790 2.793 2.790 2.793 hybrid_alltoall_any 4261 16.3 0.201 0.864 2.313 2.732 wfi_extrapolate 11 7.9 0.001 0.001 2.699 2.699 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.684 2.686 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.557 2.558 grid_collocate_task_list 110 9.6 2.277 2.541 2.277 2.541 rs_pw_transfer 902 11.9 0.010 0.011 2.131 2.533 calculate_dm_sparse 110 9.5 0.001 0.001 2.457 2.503 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.377 2.385 cp_fm_diag_elpa_base 48 14.0 2.323 2.348 2.376 2.383 fft3d_ps 1111 14.6 1.126 1.135 2.314 2.330 calculate_first_density_matrix 1 7.0 0.000 0.000 2.322 2.323 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.228 2.238 potential_pw2rs 110 12.3 0.015 0.016 2.031 2.034 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.770 1.984 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.842 1.855 mp_alltoall_d11v 2046 13.8 1.572 1.827 1.572 1.827 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.728 1.745 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.664 1.666 cp_fm_cholesky_decompose 22 10.9 1.628 1.644 1.628 1.644 qs_env_update_s_mstruct 11 6.9 0.006 0.023 1.507 1.634 mp_allgather_i34 2055 14.4 0.512 1.609 0.512 1.609 dbcsr_complete_redistribute 325 12.2 0.562 0.587 1.415 1.509 mp_waitany 9240 13.8 1.057 1.494 1.057 1.494 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.429 1.440 jit_kernel_multiply 8 15.8 1.043 1.425 1.043 1.425 qs_create_task_list 11 7.9 0.000 0.000 1.219 1.327 generate_qs_task_list 11 8.9 0.374 0.442 1.219 1.327 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.748 1.222 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.791 1.202 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.104 1.129 rs_gather_matrices 110 12.3 0.325 0.369 0.913 1.107 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.484 1.107 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.468000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.727273, yerr=14.585740 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 1.366712E+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 0.019 0.036 92.373 92.373 qs_mol_dyn_low 1 2.0 0.003 0.003 92.067 92.075 qs_forces 11 3.9 0.002 0.003 91.997 91.998 qs_energies 11 4.9 0.001 0.001 87.745 87.747 scf_env_do_scf 11 5.9 0.001 0.001 77.479 77.479 velocity_verlet 10 3.0 0.001 0.001 59.186 59.192 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 47.137 47.139 init_scf_loop 11 6.9 0.000 0.000 30.259 30.262 dbcsr_multiply_generic 2055 12.4 0.119 0.124 29.714 29.839 prepare_preconditioner 11 7.9 0.000 0.000 28.145 28.153 make_preconditioner 11 8.9 0.000 0.000 28.145 28.153 qs_scf_new_mos 99 7.5 0.001 0.001 28.098 28.142 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.097 28.142 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.272 27.601 ot_scf_mini 99 9.5 0.002 0.002 26.277 26.308 multiply_cannon 2055 13.4 0.344 0.372 22.013 23.031 multiply_cannon_loop 2055 14.4 0.341 0.343 20.091 20.450 cp_fm_upper_to_full 70 14.2 12.720 18.303 12.720 18.303 rebuild_ks_matrix 110 8.3 0.000 0.001 14.633 14.680 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.633 14.680 ot_mini 99 10.5 0.001 0.001 14.306 14.354 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.250 13.292 dbcsr_complete_redistribute 325 12.2 1.019 1.051 7.551 10.737 qs_ot_get_derivative 99 11.5 0.001 0.001 9.656 9.693 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.472 9.662 multiply_cannon_multrec 8220 15.4 4.397 4.641 9.558 9.652 mp_waitall_1 84994 16.7 8.565 9.599 8.565 9.599 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.858 9.021 mp_alltoall_i22 605 13.7 5.481 8.663 5.481 8.663 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.151 8.186 calculate_rho_elec 110 8.6 0.225 0.225 8.150 8.185 sum_up_and_integrate 110 10.3 0.151 0.154 7.706 7.720 integrate_v_rspace 110 11.3 0.004 0.004 7.555 7.569 cp_fm_cholesky_invert 11 10.9 6.590 6.595 6.590 6.595 qs_ot_get_p 110 10.4 0.001 0.001 6.499 6.539 make_m2s 4110 13.4 0.043 0.044 5.824 6.467 make_images 4110 14.4 0.875 0.921 5.635 6.277 init_scf_run 11 5.9 0.000 0.001 5.933 5.934 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.933 5.933 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.188 5.529 pw_transfer 1331 11.6 0.075 0.075 5.459 5.473 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.342 5.355 dbcsr_mm_accdrv_process 11614 15.7 3.300 3.617 5.017 5.296 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.752 5.237 apply_single 110 13.6 0.000 0.000 4.752 5.237 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.747 4.754 ot_diis_step 99 11.5 0.015 0.015 4.616 4.617 fft_wrap_pw1pw2_140 451 13.1 1.336 1.341 4.593 4.607 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.252 4.252 density_rs2pw 110 9.6 0.004 0.004 4.142 4.176 make_images_data 4110 15.4 0.042 0.044 3.297 4.128 hybrid_alltoall_any 4261 16.3 0.257 0.552 3.250 4.101 multiply_cannon_sync_h2d 8220 15.4 3.952 3.958 3.952 3.958 grid_integrate_task_list 110 12.3 3.675 3.735 3.675 3.735 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.676 3.677 wfi_extrapolate 11 7.9 0.001 0.001 3.568 3.568 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.551 3.551 cp_fm_diag_elpa_base 48 14.0 2.999 3.195 3.549 3.549 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.042 3.501 fft3d_ps 1111 14.6 1.304 1.316 3.271 3.282 calculate_dm_sparse 110 9.5 0.001 0.001 3.186 3.215 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.896 2.898 potential_pw2rs 110 12.3 0.022 0.022 2.748 2.754 rs_pw_transfer 902 11.9 0.010 0.011 2.696 2.731 grid_collocate_task_list 110 9.6 2.643 2.671 2.643 2.671 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.612 2.634 cp_fm_cholesky_decompose 22 10.9 2.447 2.472 2.447 2.472 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.367 2.383 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.205 2.266 mp_alltoall_d11v 2046 13.8 2.160 2.257 2.160 2.257 calculate_first_density_matrix 1 7.0 0.000 0.000 2.230 2.231 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.193 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.984 1.989 qs_create_task_list 11 7.9 0.000 0.000 1.890 1.939 generate_qs_task_list 11 8.9 0.731 0.784 1.890 1.938 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.838 1.869 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=92.373000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1216.090909, yerr=65.469569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/15/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 1.094965E+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 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 625.954816E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2252618. 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 0.039 0.095 205.742 205.744 qs_mol_dyn_low 1 2.0 0.004 0.006 205.106 205.151 qs_forces 11 3.9 0.007 0.024 205.014 205.015 qs_energies 11 4.9 0.002 0.010 199.418 199.437 scf_env_do_scf 11 5.9 0.001 0.001 182.850 182.854 scf_env_do_scf_inner_loop 117 6.6 0.004 0.008 162.441 162.442 dbcsr_multiply_generic 2507 12.6 0.178 0.181 123.703 124.053 velocity_verlet 10 3.0 0.001 0.005 122.976 122.979 qs_scf_new_mos 117 7.6 0.001 0.001 122.669 122.871 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.668 122.870 ot_scf_mini 117 9.6 0.003 0.003 116.072 116.255 multiply_cannon 2507 13.6 0.239 0.248 100.863 102.461 multiply_cannon_loop 2507 14.6 2.063 2.097 98.625 99.669 ot_mini 117 10.6 0.001 0.001 65.257 65.457 multiply_cannon_multrec 60168 15.6 33.412 35.231 41.735 43.730 qs_ot_get_derivative 117 11.6 0.001 0.001 40.590 40.788 rebuild_ks_matrix 128 8.3 0.001 0.001 33.561 33.868 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.561 33.868 mp_waitall_1 267128 16.5 27.811 30.781 27.811 30.781 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.135 30.416 qs_ot_get_p 128 10.4 0.001 0.002 29.068 29.308 multiply_cannon_sync_h2d 60168 15.6 27.630 29.011 27.630 29.011 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.300 24.902 apply_single 128 13.6 0.001 0.001 24.299 24.902 ot_diis_step 117 11.6 0.008 0.008 24.445 24.446 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.602 22.649 init_scf_loop 11 6.9 0.000 0.001 20.337 20.338 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.118 20.119 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.313 19.524 multiply_cannon_metrocomm3 60168 15.6 0.113 0.117 15.458 17.358 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.182 17.205 cp_fm_diag_elpa_base 83 14.4 17.115 17.152 17.178 17.203 prepare_preconditioner 11 7.9 0.000 0.000 15.721 15.769 make_preconditioner 11 8.9 0.000 0.000 15.721 15.769 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.951 15.113 sum_up_and_integrate 128 10.3 0.089 0.108 14.124 14.140 integrate_v_rspace 128 11.3 0.004 0.018 14.034 14.054 make_m2s 5014 13.6 0.104 0.111 13.678 13.969 make_images 5014 14.6 0.396 0.413 13.497 13.799 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.573 13.699 calculate_rho_elec 128 8.7 0.046 0.065 13.572 13.698 init_scf_run 11 5.9 0.000 0.001 12.380 12.381 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.380 12.381 density_rs2pw 128 9.7 0.006 0.007 6.989 10.147 wfi_extrapolate 11 7.9 0.001 0.001 9.086 9.086 mp_sum_l 7950 12.9 7.918 8.995 7.918 8.995 cp_fm_cholesky_invert 11 10.9 8.913 8.920 8.913 8.920 rs_pw_transfer 1046 11.9 0.017 0.021 5.620 8.780 calculate_dm_sparse 128 9.5 0.001 0.001 8.478 8.575 dbcsr_mm_accdrv_process 124484 16.2 3.126 3.223 7.877 8.367 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.899 8.073 pw_transfer 1547 11.6 0.074 0.094 7.740 7.928 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.743 7.831 make_images_data 5014 15.6 0.067 0.073 6.789 7.735 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.538 7.727 multiply_cannon_metrocomm1 60168 15.6 0.087 0.090 5.880 7.554 grid_integrate_task_list 128 12.3 7.076 7.548 7.076 7.548 hybrid_alltoall_any 5200 16.5 0.291 2.264 5.943 7.350 fft_wrap_pw1pw2_140 523 13.2 1.273 1.320 6.611 6.798 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.004 6.637 6.648 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.765 5.856 mp_waitany 16020 13.9 2.613 5.794 2.613 5.794 grid_collocate_task_list 128 9.7 4.558 5.787 4.558 5.787 fft3d_ps 1291 14.7 2.142 2.789 5.305 5.636 mp_alltoall_d11v 2415 14.1 4.145 5.569 4.145 5.569 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.301 2.101 5.278 cp_fm_cholesky_decompose 22 10.9 4.631 4.646 4.631 4.646 potential_pw2rs 128 12.3 0.009 0.011 4.521 4.544 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.744000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=592.545455, yerr=5.331956 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 828.837888E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132587E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702166E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473827979536 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57641. MP_Allreduce 11185 956. MP_Sync 88 MP_Alltoall 1983 4911494. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.114 191.672 191.675 qs_mol_dyn_low 1 2.0 0.004 0.008 191.142 191.161 qs_forces 11 3.9 0.017 0.087 191.011 191.013 qs_energies 11 4.9 0.013 0.047 184.064 184.078 scf_env_do_scf 11 5.9 0.003 0.020 167.417 167.427 scf_env_do_scf_inner_loop 118 6.6 0.007 0.025 134.476 134.485 velocity_verlet 10 3.0 0.005 0.020 120.675 120.677 dbcsr_multiply_generic 2529 12.6 0.189 0.196 97.913 99.110 qs_scf_new_mos 118 7.6 0.001 0.001 94.094 94.641 qs_scf_loop_do_ot 118 8.6 0.001 0.002 94.093 94.640 ot_scf_mini 118 9.6 0.004 0.007 89.261 89.932 multiply_cannon 2529 13.6 0.482 0.535 77.867 82.619 multiply_cannon_loop 2529 14.6 1.249 1.284 74.561 76.834 ot_mini 118 10.6 0.002 0.008 49.906 50.501 mp_waitall_1 216598 16.6 24.571 38.107 24.571 38.107 multiply_cannon_multrec 30348 15.6 22.158 26.211 31.805 36.380 rebuild_ks_matrix 129 8.3 0.001 0.001 33.674 34.250 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.022 33.674 34.249 init_scf_loop 11 6.9 0.001 0.005 32.851 32.870 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.248 30.764 multiply_cannon_metrocomm3 30348 15.6 0.091 0.095 15.835 29.048 qs_ot_get_derivative 118 11.6 0.001 0.003 27.872 28.504 prepare_preconditioner 11 7.9 0.000 0.001 28.282 28.361 make_preconditioner 11 8.9 0.000 0.002 28.282 28.361 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.017 27.557 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.120 23.273 apply_single 129 13.6 0.001 0.001 22.120 23.273 ot_diis_step 118 11.6 0.015 0.020 21.868 21.870 qs_ot_get_p 129 10.4 0.002 0.009 21.123 21.830 multiply_cannon_sync_h2d 30348 15.6 19.400 21.523 19.400 21.523 cp_fm_cholesky_invert 11 10.9 16.539 16.551 16.539 16.551 qs_ot_p2m_diag 84 11.4 0.190 0.219 16.163 16.196 make_m2s 5058 13.6 0.089 0.096 14.227 15.766 make_images 5058 14.6 1.155 1.346 14.018 15.557 cp_dbcsr_syevd 84 12.4 0.006 0.006 15.026 15.026 sum_up_and_integrate 129 10.3 0.119 0.141 14.823 14.868 integrate_v_rspace 129 11.3 0.004 0.006 14.704 14.756 qs_rho_update_rho_low 129 7.7 0.001 0.002 14.126 14.155 calculate_rho_elec 129 8.7 0.088 0.105 14.126 14.154 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.868 11.900 cp_fm_diag_elpa_base 84 14.4 11.611 11.723 11.863 11.897 init_scf_run 11 5.9 0.000 0.001 11.690 11.692 scf_env_initial_rho_setup 11 6.9 0.001 0.005 11.690 11.691 qs_ot_get_derivative_diag 78 12.4 0.003 0.005 10.981 11.471 multiply_cannon_metrocomm4 27819 15.6 0.105 0.119 3.809 10.559 make_images_data 5058 15.6 0.065 0.074 8.679 10.527 mp_irecv_dv 70084 16.3 3.606 10.169 3.606 10.169 hybrid_alltoall_any 5245 16.5 0.346 1.530 7.376 9.955 density_rs2pw 129 9.7 0.007 0.007 7.437 9.865 dbcsr_mm_accdrv_process 62758 16.2 4.528 5.354 9.100 9.754 pw_transfer 1559 11.6 0.086 0.095 8.825 8.884 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.599 8.661 wfi_extrapolate 11 7.9 0.001 0.001 8.385 8.385 rs_pw_transfer 1054 12.0 0.017 0.035 5.826 8.305 fft_wrap_pw1pw2_140 527 13.2 1.341 1.358 7.628 7.710 grid_integrate_task_list 129 12.3 7.225 7.579 7.225 7.579 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.198 6.967 cp_fm_cholesky_decompose 22 10.9 6.859 6.947 6.859 6.947 calculate_dm_sparse 129 9.5 0.001 0.001 6.482 6.626 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.158 6.168 mp_sum_l 8016 12.9 4.104 6.122 4.104 6.122 fft3d_ps 1301 14.7 2.837 3.010 5.890 5.935 grid_collocate_task_list 129 9.7 4.739 5.885 4.739 5.885 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.332 5.460 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.322 5.394 mp_allgather_i34 2529 14.6 1.873 5.131 1.873 5.131 mp_waitany 11836 13.9 2.617 5.130 2.617 5.130 potential_pw2rs 129 12.3 0.016 0.019 5.013 5.046 mp_alltoall_d11v 2429 14.1 4.055 4.953 4.055 4.953 rs_pw_transfer_RS2PW_140 140 11.5 0.356 0.380 2.188 4.641 mp_sum_d 4509 12.1 2.712 4.072 2.712 4.072 dbcsr_complete_redistribute 397 12.7 0.766 0.863 3.084 3.950 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.818 3.873 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.675000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.818182, yerr=3.128118 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 959.971328E+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 0.046 0.118 175.981 175.981 qs_mol_dyn_low 1 2.0 0.003 0.003 175.394 175.418 qs_forces 11 3.9 0.004 0.004 175.237 175.242 qs_energies 11 4.9 0.008 0.059 168.606 168.619 scf_env_do_scf 11 5.9 0.001 0.002 152.990 152.991 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.888 117.889 velocity_verlet 10 3.0 0.001 0.001 113.126 113.130 dbcsr_multiply_generic 2507 12.6 0.179 0.185 81.398 82.535 qs_scf_new_mos 117 7.6 0.001 0.001 79.726 80.047 qs_scf_loop_do_ot 117 8.6 0.001 0.002 79.726 80.047 ot_scf_mini 117 9.6 0.004 0.005 75.588 75.974 multiply_cannon 2507 13.6 0.504 0.519 61.834 65.892 multiply_cannon_loop 2507 14.6 0.853 0.882 58.507 61.245 ot_mini 117 10.6 0.001 0.001 42.374 42.747 init_scf_loop 11 6.9 0.001 0.006 35.002 35.004 mp_waitall_1 170520 16.6 24.702 33.961 24.702 33.961 rebuild_ks_matrix 128 8.3 0.001 0.001 30.785 31.247 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.021 30.784 31.246 prepare_preconditioner 11 7.9 0.000 0.001 30.966 31.012 make_preconditioner 11 8.9 0.000 0.001 30.966 31.012 make_full_inverse_cholesky 11 9.9 0.000 0.001 28.540 30.039 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.721 28.155 multiply_cannon_multrec 20056 15.6 13.572 16.618 22.189 25.133 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.094 24.519 qs_ot_get_derivative 117 11.6 0.001 0.002 22.612 22.993 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.867 20.799 apply_single 128 13.6 0.001 0.001 19.866 20.798 ot_diis_step 117 11.6 0.018 0.019 19.659 19.659 qs_ot_get_p 128 10.4 0.001 0.001 18.610 19.048 multiply_cannon_sync_h2d 20056 15.6 14.308 15.716 14.308 15.716 make_m2s 5014 13.6 0.080 0.085 14.598 15.711 make_images 5014 14.6 1.175 1.268 14.367 15.474 sum_up_and_integrate 128 10.3 0.133 0.145 14.562 14.588 integrate_v_rspace 128 11.3 0.004 0.004 14.428 14.459 qs_ot_p2m_diag 83 11.4 0.266 0.273 14.351 14.358 cp_fm_cholesky_invert 11 10.9 14.332 14.341 14.332 14.341 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.254 14.287 calculate_rho_elec 128 8.7 0.130 0.145 14.254 14.286 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.356 13.357 init_scf_run 11 5.9 0.001 0.005 10.535 10.536 scf_env_initial_rho_setup 11 6.9 0.003 0.013 10.534 10.535 make_images_data 5014 15.6 0.059 0.067 8.967 10.358 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.202 10.223 cp_fm_diag_elpa_base 83 14.4 9.793 9.967 10.199 10.221 hybrid_alltoall_any 5200 16.5 0.433 1.979 7.955 9.594 density_rs2pw 128 9.7 0.006 0.008 7.134 9.166 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.876 9.148 multiply_cannon_metrocomm4 17549 15.6 0.062 0.071 3.421 8.985 pw_transfer 1547 11.6 0.086 0.102 8.862 8.965 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.638 8.747 mp_irecv_dv 50230 16.2 3.298 8.739 3.298 8.739 dbcsr_mm_accdrv_process 41502 16.2 4.456 5.219 8.064 8.185 grid_integrate_task_list 128 12.3 7.289 7.794 7.289 7.794 fft_wrap_pw1pw2_140 523 13.2 1.415 1.445 7.663 7.785 wfi_extrapolate 11 7.9 0.001 0.001 7.468 7.468 cp_fm_upper_to_full 105 14.8 5.830 7.411 5.830 7.411 cp_fm_cholesky_decompose 22 10.9 7.339 7.395 7.339 7.395 rs_pw_transfer 1046 11.9 0.014 0.015 5.082 7.117 dbcsr_complete_redistribute 395 12.7 1.168 1.194 4.619 6.436 calculate_dm_sparse 128 9.5 0.001 0.001 5.729 5.840 fft3d_ps 1291 14.7 2.739 2.966 5.760 5.839 grid_collocate_task_list 128 9.7 4.896 5.825 4.896 5.825 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.422 5.428 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.468 5.286 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.522 5.195 mp_alltoall_d11v 2415 14.1 4.282 5.100 4.282 5.100 mp_allgather_i34 2507 14.6 1.852 4.776 1.852 4.776 mp_sum_l 7950 12.9 3.206 4.687 3.206 4.687 potential_pw2rs 128 12.3 0.021 0.023 4.672 4.684 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.540 4.654 mp_waitany 11748 13.9 2.310 4.390 2.310 4.390 transfer_fm_to_dbcsr 11 9.9 0.018 0.028 2.407 4.147 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.976 4.002 mp_alltoall_i22 716 14.1 2.010 3.952 2.010 3.952 rs_pw_transfer_RS2PW_140 139 11.5 0.335 0.355 1.910 3.931 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.745 3.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.565 3.604 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.981000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=910.272727, yerr=8.475536 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 1.166033E+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 0.061 0.112 195.512 195.513 qs_mol_dyn_low 1 2.0 0.003 0.004 194.980 194.993 qs_forces 11 3.9 0.014 0.070 194.866 194.873 qs_energies 11 4.9 0.002 0.007 187.548 187.560 scf_env_do_scf 11 5.9 0.001 0.001 170.462 170.469 velocity_verlet 10 3.0 0.001 0.001 129.022 129.025 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.874 121.881 qs_scf_new_mos 117 7.6 0.001 0.001 83.291 83.579 qs_scf_loop_do_ot 117 8.6 0.001 0.001 83.290 83.578 dbcsr_multiply_generic 2507 12.6 0.186 0.192 79.635 80.282 ot_scf_mini 117 9.6 0.006 0.023 78.794 79.135 multiply_cannon 2507 13.6 0.555 0.588 54.577 58.007 multiply_cannon_loop 2507 14.6 1.174 1.201 50.841 52.480 init_scf_loop 11 6.9 0.001 0.004 48.461 48.500 prepare_preconditioner 11 7.9 0.000 0.000 44.185 44.213 make_preconditioner 11 8.9 0.000 0.002 44.185 44.213 ot_mini 117 10.6 0.001 0.006 42.624 42.940 make_full_inverse_cholesky 11 9.9 0.012 0.022 37.681 42.715 multiply_cannon_multrec 30084 15.6 14.069 19.886 25.901 31.205 rebuild_ks_matrix 128 8.3 0.001 0.001 30.668 30.995 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.022 30.668 30.994 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.640 27.928 mp_waitall_1 147882 16.7 17.946 27.311 17.946 27.311 qs_ot_get_derivative 117 11.6 0.002 0.003 22.990 23.321 qs_ot_get_p 128 10.4 0.003 0.021 22.383 22.681 make_m2s 5014 13.6 0.096 0.099 20.662 21.714 make_images 5014 14.6 1.929 2.254 20.357 21.410 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.080 19.566 apply_single 128 13.6 0.001 0.001 19.080 19.566 ot_diis_step 117 11.6 0.019 0.034 19.518 19.520 qs_ot_p2m_diag 83 11.4 0.344 0.390 18.158 18.208 cp_fm_cholesky_invert 11 10.9 17.269 17.279 17.269 17.279 cp_fm_upper_to_full 105 14.8 11.498 16.949 11.498 16.949 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.842 16.846 sum_up_and_integrate 128 10.3 0.140 0.151 15.369 15.394 integrate_v_rspace 128 11.3 0.004 0.006 15.228 15.257 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.488 15.220 qs_rho_update_rho_low 128 7.7 0.001 0.002 15.090 15.124 calculate_rho_elec 128 8.7 0.173 0.189 15.089 15.123 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.483 13.492 cp_fm_diag_elpa_base 83 14.4 12.486 12.811 13.476 13.486 make_images_data 5014 15.6 0.063 0.067 11.138 13.133 dbcsr_complete_redistribute 395 12.7 1.509 1.640 9.220 13.092 multiply_cannon_sync_h2d 30084 15.6 11.717 12.822 11.717 12.822 hybrid_alltoall_any 5200 16.5 0.527 2.222 9.871 12.192 dbcsr_mm_accdrv_process 62264 16.2 7.281 8.138 11.407 11.899 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.778 11.634 init_scf_run 11 5.9 0.000 0.001 11.073 11.074 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.072 11.074 transfer_fm_to_dbcsr 11 9.9 0.002 0.006 6.481 10.268 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.578 9.824 mp_alltoall_i22 716 14.1 5.645 9.621 5.645 9.621 pw_transfer 1547 11.6 0.087 0.102 9.448 9.532 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.221 9.308 density_rs2pw 128 9.7 0.006 0.006 7.481 8.944 cp_fm_cholesky_decompose 22 10.9 8.205 8.302 8.205 8.302 fft_wrap_pw1pw2_140 523 13.2 1.559 1.592 8.105 8.200 grid_integrate_task_list 128 12.3 7.549 8.030 7.549 8.030 wfi_extrapolate 11 7.9 0.001 0.001 7.893 7.893 multiply_cannon_metrocomm4 25070 15.6 0.080 0.091 2.812 7.205 rs_pw_transfer 1046 11.9 0.013 0.015 5.458 7.041 mp_irecv_dv 76098 16.2 2.664 6.934 2.664 6.934 calculate_dm_sparse 128 9.5 0.001 0.001 6.171 6.254 fft3d_ps 1291 14.7 2.835 2.925 6.149 6.206 grid_collocate_task_list 128 9.7 5.071 5.817 5.071 5.817 mp_alltoall_d11v 2415 14.1 5.191 5.804 5.191 5.804 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.492 5.546 potential_pw2rs 128 12.3 0.023 0.023 5.061 5.104 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.436 4.543 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.474 4.476 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.374 4.462 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.208 4.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=195.513000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1082.545455, yerr=22.640927 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 1.481540E+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 3995 58308. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828162. 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 0.027 0.058 180.089 180.090 qs_mol_dyn_low 1 2.0 0.003 0.003 179.568 179.581 qs_forces 11 3.9 0.004 0.004 179.461 179.465 qs_energies 11 4.9 0.001 0.001 171.806 171.815 scf_env_do_scf 11 5.9 0.001 0.001 153.948 153.956 velocity_verlet 10 3.0 0.001 0.001 117.385 117.388 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.275 116.276 qs_scf_new_mos 116 7.6 0.001 0.001 77.110 77.190 qs_scf_loop_do_ot 116 8.6 0.001 0.001 77.110 77.190 dbcsr_multiply_generic 2485 12.5 0.180 0.192 75.258 75.756 ot_scf_mini 116 9.6 0.003 0.004 72.666 72.712 multiply_cannon 2485 13.5 0.580 0.622 54.756 59.239 multiply_cannon_loop 2485 14.5 0.438 0.452 49.908 50.751 ot_mini 116 10.6 0.001 0.001 40.389 40.440 init_scf_loop 11 6.9 0.000 0.000 37.522 37.524 mp_waitall_1 124680 16.7 26.656 34.190 26.656 34.190 prepare_preconditioner 11 7.9 0.000 0.000 33.542 33.565 make_preconditioner 11 8.9 0.000 0.000 33.542 33.565 make_full_inverse_cholesky 11 9.9 0.017 0.025 31.381 31.696 rebuild_ks_matrix 127 8.3 0.001 0.001 30.169 30.220 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 30.169 30.220 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.325 27.375 multiply_cannon_multrec 9940 15.5 10.385 15.218 17.970 20.974 qs_ot_get_derivative 116 11.6 0.001 0.002 20.616 20.669 ot_diis_step 116 11.6 0.021 0.037 19.707 19.707 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.264 19.569 apply_single 127 13.6 0.001 0.001 19.263 19.569 make_m2s 4970 13.5 0.067 0.071 16.876 19.452 qs_ot_get_p 127 10.4 0.001 0.001 19.261 19.325 cp_fm_cholesky_invert 11 10.9 19.165 19.172 19.165 19.172 make_images 4970 14.5 2.294 2.749 16.573 19.152 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 11.921 18.953 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.757 15.802 calculate_rho_elec 127 8.7 0.254 0.265 15.756 15.801 qs_ot_p2m_diag 82 11.4 0.490 0.495 15.346 15.363 sum_up_and_integrate 127 10.3 0.181 0.190 15.275 15.321 integrate_v_rspace 127 11.3 0.004 0.005 15.094 15.147 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.152 14.153 make_images_data 4970 15.5 0.051 0.060 10.306 12.930 multiply_cannon_sync_h2d 9940 15.5 11.527 12.391 11.527 12.391 hybrid_alltoall_any 5155 16.4 0.808 3.792 10.061 12.318 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.987 10.998 cp_fm_diag_elpa_base 82 14.4 10.702 10.810 10.963 10.972 init_scf_run 11 5.9 0.000 0.001 10.905 10.905 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.905 10.905 pw_transfer 1535 11.6 0.085 0.095 9.550 9.572 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.327 9.357 cp_fm_cholesky_decompose 22 10.9 8.658 8.782 8.658 8.782 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.918 8.441 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.211 8.247 fft_wrap_pw1pw2_140 519 13.2 1.896 1.933 8.160 8.183 grid_integrate_task_list 127 12.3 7.705 8.177 7.705 8.177 density_rs2pw 127 9.7 0.006 0.006 7.248 8.140 dbcsr_mm_accdrv_process 20590 16.1 2.824 3.882 7.220 7.977 wfi_extrapolate 11 7.9 0.001 0.001 7.788 7.788 mp_allgather_i34 2485 14.5 3.072 7.639 3.072 7.639 calculate_dm_sparse 127 9.5 0.001 0.001 6.147 6.245 mp_alltoall_d11v 2401 14.1 5.216 6.165 5.216 6.165 grid_collocate_task_list 127 9.7 5.352 6.009 5.352 6.009 fft3d_ps 1281 14.7 2.742 2.810 5.834 5.874 rs_pw_transfer 1038 11.9 0.012 0.013 4.809 5.711 dbcsr_complete_redistribute 393 12.7 2.125 2.170 5.231 5.541 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.363 5.364 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.239 5.252 potential_pw2rs 127 12.3 0.026 0.027 4.733 4.755 multiply_cannon_metrocomm4 7455 15.5 0.023 0.027 1.832 4.278 mp_irecv_dv 28618 15.9 1.797 4.191 1.797 4.191 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.141 4.161 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.553 3.865 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.452 3.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.732 3.734 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.711 3.730 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.535 3.617 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=180.090000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1391.272727, yerr=32.395541 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_performance_tests/20/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 11.696233E+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 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.133821E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59285. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.043 311.975 311.976 qs_mol_dyn_low 1 2.0 0.003 0.003 311.412 311.425 qs_forces 11 3.9 0.004 0.004 311.036 311.040 qs_energies 11 4.9 0.001 0.001 301.596 301.609 scf_env_do_scf 11 5.9 0.001 0.001 278.726 278.742 velocity_verlet 10 3.0 0.001 0.001 224.107 224.150 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 147.756 147.759 init_scf_loop 11 6.9 0.000 0.000 130.701 130.705 prepare_preconditioner 11 7.9 0.000 0.000 125.519 125.551 make_preconditioner 11 8.9 0.000 0.000 125.519 125.551 make_full_inverse_cholesky 11 9.9 0.036 0.039 100.960 122.605 qs_scf_new_mos 117 7.6 0.001 0.001 95.604 95.786 qs_scf_loop_do_ot 117 8.6 0.001 0.001 95.603 95.785 ot_scf_mini 117 9.6 0.004 0.004 90.758 90.844 dbcsr_multiply_generic 2507 12.6 0.212 0.224 84.956 85.757 cp_fm_upper_to_full 105 14.8 54.118 77.670 54.118 77.670 multiply_cannon 2507 13.6 0.718 0.770 59.016 60.068 multiply_cannon_loop 2507 14.6 0.471 0.480 55.158 57.112 ot_mini 117 10.6 0.001 0.001 46.061 46.154 dbcsr_complete_redistribute 395 12.7 3.989 4.039 30.450 43.612 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.976 40.133 rebuild_ks_matrix 128 8.3 0.001 0.001 39.270 39.374 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.018 39.270 39.374 transfer_fm_to_dbcsr 11 9.9 0.032 0.035 24.512 37.536 cp_fm_cholesky_invert 11 10.9 36.574 36.581 36.574 36.581 qs_ks_update_qs_env 128 7.6 0.001 0.001 36.205 36.304 mp_alltoall_i22 716 14.1 22.320 35.759 22.320 35.759 mp_waitall_1 103674 16.8 29.372 33.968 29.372 33.968 qs_ot_get_p 128 10.4 0.001 0.001 29.035 29.119 qs_ot_get_derivative 117 11.6 0.002 0.002 25.535 25.622 qs_ot_p2m_diag 83 11.4 0.879 0.885 24.684 24.728 cp_dbcsr_syevd 83 12.4 0.006 0.006 22.815 22.817 make_m2s 5014 13.6 0.077 0.081 20.854 21.957 make_images 5014 14.6 3.774 3.992 20.376 21.483 qs_rho_update_rho_low 128 7.7 0.001 0.001 21.213 21.235 calculate_rho_elec 128 8.7 0.481 0.482 21.213 21.234 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 19.082 20.814 sum_up_and_integrate 128 10.3 0.321 0.323 20.428 20.515 ot_diis_step 117 11.6 0.022 0.023 20.483 20.484 integrate_v_rspace 128 11.3 0.004 0.004 20.106 20.192 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.568 19.920 apply_single 128 13.6 0.001 0.001 19.568 19.920 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.288 19.288 cp_fm_diag_elpa_base 83 14.4 14.768 16.454 19.283 19.283 multiply_cannon_multrec 10028 15.6 10.473 12.223 17.861 17.938 multiply_cannon_sync_h2d 10028 15.6 15.658 15.667 15.658 15.667 make_images_data 5014 15.6 0.059 0.064 11.314 13.416 hybrid_alltoall_any 5200 16.5 1.301 3.026 11.253 13.338 init_scf_run 11 5.9 0.000 0.001 12.671 12.672 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.671 12.671 pw_transfer 1547 11.6 0.093 0.094 12.201 12.207 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 11.964 11.971 fft_wrap_pw1pw2_140 523 13.2 3.075 3.123 10.544 10.549 cp_fm_cholesky_decompose 22 10.9 10.185 10.282 10.185 10.282 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.081 10.149 mp_alltoall_d11v 2415 14.1 8.903 9.809 8.903 9.809 wfi_extrapolate 11 7.9 0.001 0.001 9.530 9.530 dbcsr_mm_accdrv_process 20762 16.1 3.726 5.416 7.147 8.987 grid_integrate_task_list 128 12.3 8.594 8.753 8.594 8.753 density_rs2pw 128 9.7 0.005 0.005 8.581 8.708 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.053 8.056 fft3d_ps 1291 14.7 2.849 2.880 7.038 7.067 calculate_dm_sparse 128 9.5 0.001 0.001 6.713 6.825 rs_scatter_matrices 139 9.7 3.740 4.637 6.223 6.473 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.371 6.456 grid_collocate_task_list 128 9.7 6.380 6.403 6.380 6.403 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.184 6.293 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=311.976000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2837.818182, yerr=159.999897 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 1.255903E+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 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 0.044 0.073 84.839 84.847 qs_energies 1 2.0 0.000 0.000 84.289 84.316 ls_scf 1 3.0 0.002 0.013 83.355 83.385 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.342 72.522 multiply_cannon 111 7.7 0.017 0.020 55.862 57.015 multiply_cannon_loop 111 8.7 0.208 0.223 52.453 53.813 ls_scf_main 1 4.0 0.000 0.010 52.188 52.190 density_matrix_trs4 2 5.0 0.002 0.012 46.629 46.690 ls_scf_init_scf 1 4.0 0.000 0.002 28.107 28.108 ls_scf_init_matrix_S 1 5.0 0.000 0.003 27.033 27.082 mp_waitall_1 11031 10.9 22.199 25.174 22.199 25.174 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.932 24.955 multiply_cannon_multrec 2664 9.7 8.163 8.916 15.468 17.238 multiply_cannon_sync_h2d 2664 9.7 13.775 15.602 13.775 15.602 make_m2s 222 7.7 0.008 0.010 12.902 13.448 make_images 222 8.7 0.098 0.108 12.880 13.427 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.618 12.936 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.391 8.690 make_images_data 222 9.7 0.004 0.005 7.501 8.017 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.639 6.924 7.917 hybrid_alltoall_any 227 10.6 0.216 1.840 6.436 7.663 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.126 6.212 7.126 calculate_norms 4752 9.8 5.546 6.260 5.546 6.260 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.066 5.184 mp_sum_l 887 5.1 3.046 4.651 3.046 4.651 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.258 3.527 arnoldi_extremal 4 6.8 0.000 0.000 3.324 3.357 arnoldi_normal_ev 4 7.8 0.001 0.003 3.324 3.357 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.051 3.341 mp_irecv_dv 6231 10.9 2.034 3.309 2.034 3.309 build_subspace 16 8.4 0.009 0.012 3.229 3.231 make_images_sizes 222 9.7 0.000 0.000 0.700 3.213 mp_alltoall_i44 222 10.7 0.700 3.213 0.700 3.213 ls_scf_post 1 4.0 0.003 0.022 3.058 3.087 ls_scf_store_result 1 5.0 0.000 0.000 2.848 2.903 dbcsr_special_finalize 555 9.7 0.005 0.006 2.350 2.761 dbcsr_merge_single_wm 555 10.7 0.457 0.603 2.342 2.753 make_images_pack 222 9.7 2.204 2.618 2.206 2.619 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.355 2.584 dbcsr_sort_data 658 11.4 2.142 2.464 2.142 2.464 dbcsr_matrix_vector_mult_local 304 10.0 2.058 2.446 2.060 2.448 ls_scf_dm_to_ks 2 5.0 0.000 0.003 2.285 2.354 buffer_matrices_ensure_size 222 8.7 1.748 2.047 1.748 2.047 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.790 1.793 rebuild_ks_matrix 3 7.3 0.000 0.000 1.781 1.784 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.781 1.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.847000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 2.112422E+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 0.055 0.084 89.790 89.793 qs_energies 1 2.0 0.000 0.000 89.208 89.213 ls_scf 1 3.0 0.000 0.000 87.907 87.912 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.062 74.397 multiply_cannon 111 7.7 0.029 0.039 52.575 56.022 ls_scf_main 1 4.0 0.000 0.000 53.868 53.869 multiply_cannon_loop 111 8.7 0.116 0.124 49.286 52.446 density_matrix_trs4 2 5.0 0.002 0.003 48.266 48.447 ls_scf_init_scf 1 4.0 0.000 0.000 30.514 30.515 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.342 29.434 mp_waitall_1 9105 10.9 19.655 28.531 19.655 28.531 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.970 26.981 multiply_cannon_multrec 1332 9.7 13.261 16.532 22.471 26.877 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 10.923 19.805 make_m2s 222 7.7 0.006 0.008 14.771 15.426 make_images 222 8.7 1.573 1.942 14.740 15.396 dbcsr_mm_accdrv_process 4041 10.4 0.281 0.453 8.805 10.381 dbcsr_mm_accdrv_process_sort 4041 11.4 8.398 9.928 8.398 9.928 make_images_data 222 9.7 0.004 0.004 8.186 9.064 hybrid_alltoall_any 227 10.6 0.522 2.459 7.638 8.391 mp_sum_l 887 5.1 5.417 8.341 5.417 8.341 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.234 7.774 mp_irecv_dv 3311 11.0 3.214 7.719 3.214 7.719 calculate_norms 2376 9.8 6.041 6.901 6.041 6.901 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.216 6.679 multiply_cannon_sync_h2d 1332 9.7 4.853 6.178 4.853 6.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.982 5.187 arnoldi_extremal 4 6.8 0.000 0.000 4.646 4.664 arnoldi_normal_ev 4 7.8 0.001 0.005 4.646 4.664 build_subspace 16 8.4 0.014 0.021 4.388 4.392 ls_scf_post 1 4.0 0.000 0.000 3.525 3.529 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.132 3.360 ls_scf_store_result 1 5.0 0.000 0.000 3.234 3.349 dbcsr_matrix_vector_mult_local 304 10.0 2.719 3.202 2.721 3.203 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.544 2.641 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.181 2.555 make_images_pack 222 9.7 2.023 2.393 2.026 2.395 mp_allgather_i34 111 8.7 0.994 2.238 0.994 2.238 dbcsr_sort_data 436 11.2 1.841 2.108 1.841 2.108 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.852 1.854 dbcsr_data_new 4174 10.1 1.610 1.849 1.610 1.849 rebuild_ks_matrix 3 7.3 0.000 0.000 1.839 1.841 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.015 1.839 1.841 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.793000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 2.756829E+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 0.044 0.081 92.876 92.877 qs_energies 1 2.0 0.000 0.000 92.237 92.245 ls_scf 1 3.0 0.000 0.001 90.837 90.843 dbcsr_multiply_generic 111 6.7 0.016 0.018 75.492 75.806 ls_scf_main 1 4.0 0.000 0.001 56.767 56.771 multiply_cannon 111 7.7 0.041 0.095 52.130 56.661 multiply_cannon_loop 111 8.7 0.099 0.105 48.493 52.290 density_matrix_trs4 2 5.0 0.003 0.012 50.843 50.984 mp_waitall_1 7281 11.0 22.358 32.376 22.358 32.376 ls_scf_init_scf 1 4.0 0.000 0.001 30.645 30.648 ls_scf_init_matrix_S 1 5.0 0.000 0.002 29.381 29.453 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.006 27.026 multiply_cannon_multrec 888 9.7 12.681 15.086 21.182 24.166 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.537 22.542 make_m2s 222 7.7 0.006 0.007 16.558 17.856 make_images 222 8.7 1.970 2.296 16.520 17.817 hybrid_alltoall_any 227 10.6 0.622 2.869 8.737 10.375 make_images_data 222 9.7 0.003 0.004 9.075 10.281 dbcsr_mm_accdrv_process 3754 10.4 0.235 0.417 8.042 9.268 dbcsr_mm_accdrv_process_sort 3754 11.4 7.670 8.851 7.670 8.851 mp_sum_l 887 5.1 5.260 8.713 5.260 8.713 multiply_cannon_sync_h2d 888 9.7 6.065 7.460 6.065 7.460 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.453 6.958 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.118 6.921 mp_irecv_dv 2335 11.1 2.436 6.905 2.436 6.905 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.364 6.385 arnoldi_extremal 4 6.8 0.000 0.000 5.185 5.221 arnoldi_normal_ev 4 7.8 0.001 0.005 5.185 5.221 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.855 5.025 build_subspace 16 8.4 0.014 0.020 4.876 4.886 calculate_norms 1584 9.8 4.269 4.618 4.269 4.618 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.475 3.780 mp_allgather_i34 111 8.7 1.487 3.682 1.487 3.682 dbcsr_matrix_vector_mult_local 304 10.0 2.989 3.563 2.991 3.565 ls_scf_post 1 4.0 0.001 0.006 3.425 3.433 ls_scf_store_result 1 5.0 0.000 0.000 3.198 3.250 ls_scf_dm_to_ks 2 5.0 0.000 0.013 2.810 2.910 dbcsr_sort_data 325 11.1 1.846 2.166 1.846 2.166 make_images_sizes 222 9.7 0.000 0.000 1.214 2.114 mp_alltoall_i44 222 10.7 1.214 2.114 1.214 2.114 make_images_pack 222 9.7 1.816 2.111 1.819 2.114 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.998 2.000 rebuild_ks_matrix 3 7.3 0.000 0.000 1.980 1.982 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.980 1.982 dbcsr_data_release 9322 10.9 1.290 1.976 1.290 1.976 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.877000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2161.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 3.364340E+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 0.048 0.100 97.117 97.125 qs_energies 1 2.0 0.000 0.000 96.412 96.418 ls_scf 1 3.0 0.000 0.000 94.779 94.787 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.449 78.727 ls_scf_main 1 4.0 0.000 0.000 58.626 58.627 multiply_cannon 111 7.7 0.059 0.126 51.648 56.046 density_matrix_trs4 2 5.0 0.002 0.002 52.540 52.678 multiply_cannon_loop 111 8.7 0.114 0.125 46.597 50.210 ls_scf_init_scf 1 4.0 0.000 0.000 32.907 32.909 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.695 31.761 mp_waitall_1 6369 11.0 22.523 29.083 22.523 29.083 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.011 29.023 multiply_cannon_multrec 1332 9.7 14.257 17.747 22.115 24.641 make_m2s 222 7.7 0.007 0.008 20.938 22.369 make_images 222 8.7 3.137 3.610 20.888 22.321 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.012 16.862 make_images_data 222 9.7 0.004 0.004 11.599 13.209 hybrid_alltoall_any 227 10.6 0.797 3.763 10.930 12.655 dbcsr_mm_accdrv_process 3641 10.4 0.219 0.409 7.496 9.008 dbcsr_mm_accdrv_process_sort 3641 11.4 7.097 8.586 7.097 8.586 mp_sum_l 887 5.1 4.293 7.956 4.293 7.956 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.319 6.310 multiply_cannon_sync_h2d 1332 9.7 5.510 6.036 5.510 6.036 multiply_cannon_metrocomm4 1110 9.7 0.004 0.007 2.077 6.019 mp_irecv_dv 3229 10.9 2.055 5.945 2.055 5.945 arnoldi_extremal 4 6.8 0.000 0.000 5.326 5.336 arnoldi_normal_ev 4 7.8 0.001 0.005 5.325 5.336 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.733 5.283 build_subspace 16 8.4 0.014 0.021 4.984 4.993 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.752 4.933 calculate_norms 2376 9.8 4.202 4.557 4.202 4.557 mp_allgather_i34 111 8.7 2.169 4.498 2.169 4.498 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.604 3.899 dbcsr_matrix_vector_mult_local 304 10.0 3.142 3.647 3.144 3.649 dbcsr_sort_data 658 11.4 3.079 3.382 3.079 3.382 ls_scf_post 1 4.0 0.000 0.000 3.247 3.255 dbcsr_special_finalize 555 9.7 0.006 0.007 2.814 3.206 dbcsr_merge_single_wm 555 10.7 0.535 0.662 2.806 3.198 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.988 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.965 3.026 dbcsr_data_release 10477 10.7 1.591 2.433 1.591 2.433 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.967 1.968 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.967 rebuild_ks_matrix 3 7.3 0.000 0.000 1.944 1.946 qs_ks_build_kohn_sham_matrix 3 8.3 0.014 0.027 1.944 1.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.125000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 4.670415E+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 0.042 0.058 97.156 97.157 qs_energies 1 2.0 0.000 0.000 96.308 96.312 ls_scf 1 3.0 0.000 0.000 94.395 94.405 dbcsr_multiply_generic 111 6.7 0.018 0.018 76.017 76.201 ls_scf_main 1 4.0 0.000 0.000 60.620 60.621 multiply_cannon 111 7.7 0.099 0.168 54.109 59.269 density_matrix_trs4 2 5.0 0.002 0.003 53.551 53.634 multiply_cannon_loop 111 8.7 0.070 0.079 49.477 50.971 mp_waitall_1 5436 11.0 24.688 30.256 24.688 30.256 ls_scf_init_scf 1 4.0 0.000 0.000 30.203 30.208 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.959 28.999 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.763 26.775 multiply_cannon_multrec 444 9.7 13.994 16.801 21.024 23.890 make_m2s 222 7.7 0.004 0.005 17.274 19.866 make_images 222 8.7 3.718 4.417 17.212 19.806 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.033 14.621 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.966 13.928 make_images_data 222 9.7 0.003 0.004 9.491 12.022 hybrid_alltoall_any 227 10.6 0.789 3.760 9.219 11.805 multiply_cannon_sync_h2d 444 9.7 6.632 8.361 6.632 8.361 dbcsr_mm_accdrv_process 3003 10.4 0.188 0.344 6.731 7.863 dbcsr_mm_accdrv_process_sort 3003 11.4 6.403 7.513 6.403 7.513 mp_allgather_i34 111 8.7 2.866 7.015 2.866 7.015 arnoldi_extremal 4 6.8 0.000 0.000 5.790 5.802 arnoldi_normal_ev 4 7.8 0.002 0.005 5.790 5.801 build_subspace 16 8.4 0.015 0.019 5.399 5.406 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.720 mp_sum_l 887 5.1 2.679 4.460 2.679 4.460 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.111 4.322 dbcsr_matrix_vector_mult_local 304 10.0 3.613 4.082 3.615 4.084 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.616 3.834 mp_irecv_dv 1241 11.2 1.597 3.787 1.597 3.787 calculate_norms 792 9.8 3.542 3.727 3.542 3.727 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.513 3.592 ls_scf_post 1 4.0 0.000 0.000 3.572 3.576 ls_scf_store_result 1 5.0 0.000 0.000 3.341 3.397 make_images_sizes 222 9.7 0.000 0.000 1.139 3.164 mp_alltoall_i44 222 10.7 1.138 3.163 1.138 3.163 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.754 3.158 dbcsr_finalize 304 7.8 0.062 0.077 2.193 2.265 dbcsr_merge_all 275 8.9 0.473 0.528 2.045 2.103 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.053 2.055 rebuild_ks_matrix 3 7.3 0.000 0.000 2.019 2.021 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.019 2.021 dbcsr_data_release 10123 10.8 1.326 1.981 1.326 1.981 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=97.157000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3674.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 8.793334E+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 0.092 0.143 106.541 106.553 qs_energies 1 2.0 0.000 0.000 105.066 105.078 ls_scf 1 3.0 0.000 0.000 102.113 102.124 dbcsr_multiply_generic 111 6.7 0.023 0.028 75.586 75.689 ls_scf_main 1 4.0 0.000 0.000 64.772 64.773 density_matrix_trs4 2 5.0 0.002 0.003 55.550 55.617 multiply_cannon 111 7.7 0.183 0.368 48.247 50.010 multiply_cannon_loop 111 8.7 0.067 0.071 44.890 45.802 ls_scf_init_scf 1 4.0 0.000 0.000 33.717 33.718 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.160 32.171 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.366 29.376 make_m2s 222 7.7 0.005 0.005 23.680 24.529 make_images 222 8.7 4.590 4.966 23.574 24.420 mp_waitall_1 4527 11.1 20.438 24.050 20.438 24.050 multiply_cannon_multrec 444 9.7 17.866 18.680 22.461 23.063 hybrid_alltoall_any 227 10.6 1.664 3.643 12.629 15.330 make_images_data 222 9.7 0.003 0.004 12.868 15.183 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 9.007 9.677 multiply_cannon_sync_h2d 444 9.7 8.842 8.893 8.842 8.893 arnoldi_extremal 4 6.8 0.000 0.000 7.364 7.376 arnoldi_normal_ev 4 7.8 0.003 0.009 7.364 7.376 build_subspace 16 8.4 0.026 0.036 6.816 6.827 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.419 5.546 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.228 5.316 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.953 5.210 dbcsr_matrix_vector_mult_local 304 10.0 4.861 5.155 4.864 5.158 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.323 4.418 4.549 dbcsr_mm_accdrv_process_sort 1814 11.4 4.119 4.253 4.119 4.253 make_images_sizes 222 9.7 0.000 0.000 1.480 3.650 mp_alltoall_i44 222 10.7 1.480 3.650 1.480 3.650 ls_scf_post 1 4.0 0.000 0.000 3.625 3.636 ls_scf_store_result 1 5.0 0.000 0.000 3.369 3.382 calculate_norms 792 9.8 3.249 3.296 3.249 3.296 mp_allgather_i34 111 8.7 1.011 3.152 1.011 3.152 dbcsr_finalize 304 7.8 0.082 0.089 3.082 3.139 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.922 2.922 dbcsr_complete_redistribute 5 7.6 1.434 1.477 2.785 2.920 dbcsr_merge_all 275 8.9 0.891 0.913 2.868 2.913 dbcsr_data_release 12724 10.6 2.320 2.867 2.320 2.867 matrix_ls_to_qs 2 6.0 0.000 0.000 2.452 2.600 dbcsr_sort_data 325 11.1 2.442 2.504 2.442 2.504 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.492 2.494 rebuild_ks_matrix 3 7.3 0.000 0.000 2.426 2.428 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 2.426 2.428 dbcsr_new_transposed 4 7.5 0.243 0.253 2.333 2.350 dbcsr_frobenius_norm 74 6.6 2.062 2.145 2.214 2.247 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.203 mp_sum_l 887 5.1 1.362 2.203 1.362 2.203 dbcsr_add_anytype 103 7.2 0.860 0.893 2.133 2.203 mp_alltoall_d11v 48 9.2 2.077 2.137 2.077 2.137 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.553000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6935.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0e0e430d9aee02af32c52f46e573f4f07a4ec1f_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 586.711040E+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 504054042. 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 0.088 0.156 256.716 256.718 qs_mol_dyn_low 1 2.0 0.004 0.011 255.297 255.335 qs_forces 5 3.8 0.007 0.058 255.166 255.170 qs_energies 5 4.8 0.008 0.083 251.851 251.879 scf_env_do_scf 5 5.8 0.002 0.036 234.090 234.096 scf_env_do_scf_inner_loop 105 6.6 0.005 0.022 203.209 203.219 qs_scf_new_mos 105 7.6 0.000 0.001 157.888 158.161 qs_scf_loop_do_ot 105 8.6 0.002 0.012 157.888 158.160 ot_scf_mini 105 9.6 0.003 0.016 148.127 148.276 dbcsr_multiply_generic 1445 12.2 0.125 0.134 135.304 135.927 multiply_cannon 1445 13.2 0.277 0.291 115.502 117.686 multiply_cannon_loop 1445 14.2 2.352 2.492 113.306 115.209 velocity_verlet 4 3.0 0.001 0.029 112.152 112.153 ot_mini 105 10.6 0.001 0.013 60.689 60.838 qs_ot_get_p 112 10.4 0.001 0.001 57.385 57.660 qs_ot_p2m_diag 40 11.0 0.020 0.032 45.943 46.061 multiply_cannon_multrec 69360 15.2 31.693 37.429 39.413 45.405 mp_waitall_1 488190 16.1 35.330 43.797 35.330 43.797 cp_dbcsr_syevd 40 12.0 0.002 0.004 42.804 42.806 qs_ot_get_derivative 55 11.6 0.001 0.017 38.947 39.079 cp_fm_syevd 40 13.0 0.000 0.002 36.985 37.139 multiply_cannon_metrocomm3 69360 15.2 0.193 0.205 25.018 35.148 multiply_cannon_sync_h2d 69360 15.2 30.630 35.117 30.630 35.117 cp_fm_redistribute_end 40 14.0 15.949 31.819 15.956 31.822 cp_fm_syevd_base 40 14.0 15.854 31.727 15.854 31.727 rebuild_ks_matrix 110 8.4 0.000 0.000 31.234 31.463 qs_ks_build_kohn_sham_matrix 110 9.4 0.017 0.087 31.233 31.462 init_scf_loop 7 6.6 0.003 0.034 30.831 30.837 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.629 28.831 prepare_preconditioner 7 7.6 0.000 0.001 25.338 25.373 make_preconditioner 7 8.6 0.002 0.028 25.338 25.373 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.011 23.278 apply_single 62 13.6 0.000 0.000 23.011 23.278 qs_rho_update_rho_low 110 7.6 0.001 0.010 20.971 21.350 calculate_rho_elec 110 8.6 0.030 0.034 20.970 21.349 ot_new_cg_direction 55 11.6 0.001 0.029 21.015 21.039 rs_pw_transfer 690 11.5 0.011 0.019 15.956 17.227 make_full_inverse_cholesky 7 9.6 0.001 0.010 17.015 17.083 density_rs2pw 110 9.6 0.007 0.020 14.744 15.939 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.801 14.966 init_scf_run 5 5.8 0.000 0.001 14.873 14.874 scf_env_initial_rho_setup 5 6.8 0.002 0.020 14.873 14.874 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.144 14.229 pw_transfer 1645 12.4 0.086 0.109 12.810 13.136 fft_wrap_pw1pw2 1425 13.5 0.014 0.017 12.665 12.994 mp_sum_l 4764 12.2 12.052 12.959 12.052 12.959 cp_fm_cholesky_invert 7 10.6 11.610 11.631 11.610 11.631 calculate_dm_sparse 110 9.5 0.000 0.001 11.429 11.622 fft_wrap_pw1pw2_240 915 15.0 1.227 1.317 10.581 10.862 qs_vxc_create 110 10.4 0.004 0.022 10.670 10.732 sum_up_and_integrate 60 10.3 0.028 0.046 10.514 10.531 integrate_v_rspace 60 11.3 0.003 0.020 10.485 10.504 check_diag 80 13.5 8.584 8.857 10.311 10.459 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.338 10.431 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.014 9.749 9.767 calculate_first_density_matrix 1 7.0 0.001 0.006 9.538 9.551 acc_transpose_blocks 69360 15.2 0.349 0.379 8.667 9.190 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 8.987 9.053 fft3d_pb 915 16.0 2.393 2.712 8.592 8.965 multiply_cannon_metrocomm1 69360 15.2 0.090 0.096 4.416 8.699 dbcsr_mm_accdrv_process 154766 15.8 3.959 4.194 7.589 8.477 make_full_single_inverse 7 9.6 0.003 0.036 7.981 8.016 make_m2s 2890 13.2 0.077 0.084 7.329 7.902 make_images 2890 14.2 0.241 0.260 7.225 7.797 xc_rho_set_and_dset_create 110 12.4 0.081 0.167 7.361 7.693 xc_vxc_pw_create 60 11.3 0.041 0.075 7.218 7.281 acc_transpose_blocks_kernels 69360 16.2 0.843 0.896 6.820 7.214 potential_pw2rs 60 12.3 0.003 0.004 6.508 6.570 xc_pw_derive 510 13.4 0.006 0.007 6.390 6.478 jit_kernel_transpose 5 15.0 5.976 6.340 5.976 6.340 mp_alltoall_z22v 2340 17.7 5.636 5.972 5.636 5.972 mp_waitany 7680 13.5 4.476 5.800 4.476 5.800 rs_pw_transfer_RS2PW_30 110 11.6 1.578 1.664 5.117 5.516 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=256.718000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b0e0e430d9aee02af32c52f46e573f4f07a4ec1f Summary: empty Status: OK