=== 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: b0f1689a086be006f33214c134fa9a07bd867c23 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.0.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (26.10.2022) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.0.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/01 job id: 42443204 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/02 job id: 42443205 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/03 job id: 42443206 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/04 job id: 42443208 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/05 job id: 42443209 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/06 job id: 42443211 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/07 job id: 42443212 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/08 job id: 42443214 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/09 job id: 42443215 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/10 job id: 42443216 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/11 job id: 42443218 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/12 job id: 42443220 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/13 job id: 42443221 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/14 job id: 42443224 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/15 job id: 42443225 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/16 job id: 42443226 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/17 job id: 42443229 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/18 job id: 42443230 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/19 job id: 42443233 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/20 job id: 42443235 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/21 job id: 42443236 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/22 job id: 42443238 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/23 job id: 42443240 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/24 job id: 42443241 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/25 job id: 42443243 --- 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/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/26 job id: 42443244 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 139.251 139.252 farming_run 1 2.0 138.697 138.698 139.222 139.227 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.460462E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.029 118.492 118.493 qs_energies 1 2.0 0.000 0.000 118.249 118.251 mp2_main 1 3.0 0.000 0.000 114.383 114.384 mp2_gpw_main 1 4.0 0.018 0.021 112.255 112.257 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 93.075 93.542 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.403 55.874 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.169 41.711 47.132 get_2c_integrals 1 6.0 0.000 0.000 37.159 37.499 integrate_v_rspace 273 8.0 0.435 0.449 25.136 30.317 pw_transfer 6555 10.6 0.376 0.399 27.316 27.839 grid_integrate_task_list 273 9.0 20.950 26.606 20.950 26.606 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.955 26.516 fft_wrap_pw1pw2_100 2178 12.4 1.174 1.395 23.487 24.063 compute_2c_integrals 1 7.0 0.002 0.002 19.785 19.787 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.864 19.322 mp2_eri_2c_integrate_gpw 1 9.0 2.387 2.442 18.862 19.321 rpa_ri_compute_en 1 5.0 0.000 0.000 19.066 19.285 cp_fm_cholesky_decompose 12 8.2 17.564 17.931 17.564 17.931 fft3d_s 5443 13.4 16.077 16.569 16.098 16.591 cholesky_decomp 1 7.0 0.000 0.000 16.222 16.550 ao_to_mo_and_store_B_mult_1 272 7.0 10.853 15.570 10.853 15.570 calculate_wavefunction 272 8.0 5.397 5.554 12.496 13.138 rpa_num_int 1 6.0 0.000 0.001 10.762 10.772 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.666 10.749 calc_mat_Q 8 8.0 0.000 0.000 9.359 9.487 contract_S_to_Q 8 9.0 0.000 0.000 8.779 8.904 calc_potential_gpw 544 9.5 0.005 0.006 8.281 8.660 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.212 8.512 parallel_gemm_fm 14 9.1 0.000 0.000 8.387 8.485 parallel_gemm_fm_cosma 14 10.1 8.387 8.485 8.387 8.485 potential_pw2rs 545 10.0 0.108 0.111 7.666 8.288 create_integ_mat 1 6.0 0.014 0.027 7.739 7.749 collocate_single_gaussian 272 10.0 0.041 0.043 7.441 7.714 array2fm 1 7.0 0.000 0.000 6.742 7.158 pw_scatter_s 2720 13.7 4.410 4.606 4.410 4.606 pw_gather_s 2722 13.2 3.894 4.288 3.894 4.288 cp_fm_syevd 4 7.5 0.000 0.000 3.655 3.659 scf_env_do_scf 1 3.0 0.000 0.000 3.546 3.546 cp_fm_syevd_base 4 7.9 3.359 3.416 3.359 3.416 init_scf_loop 1 4.0 0.000 0.000 3.349 3.349 array2fm_buffer_send 1 8.0 2.981 3.172 2.981 3.172 prepare_preconditioner 1 5.0 0.000 0.000 3.119 3.120 make_preconditioner 1 6.0 0.000 0.000 3.055 3.055 make_full_all 1 7.0 0.000 0.000 2.958 2.958 pw_poisson_solve 545 10.5 1.117 1.150 2.184 2.383 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.257179, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.045 398.963 398.965 farming_run 1 2.0 398.154 398.183 398.912 398.919 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224204E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 6852540. MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.041 209.652 209.653 qs_energies 1 2.0 0.003 0.021 209.410 209.434 scf_env_do_scf 1 3.0 0.000 0.000 106.341 106.341 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.488 105.496 rebuild_ks_matrix 4 6.0 0.000 0.000 105.487 105.495 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 105.487 105.495 hfx_ks_matrix 4 8.0 0.001 0.001 105.118 105.123 integrate_four_center 4 9.0 0.144 0.464 105.117 105.122 mp2_main 1 3.0 0.002 0.020 102.774 102.781 mp2_gpw_main 1 4.0 0.034 0.058 101.828 101.838 integrate_four_center_main 4 10.0 0.105 0.444 96.521 99.656 integrate_four_center_bin 269 11.0 96.416 99.637 96.416 99.637 init_scf_loop 1 4.0 0.000 0.000 92.013 92.013 mp2_ri_gpw_compute_in 1 5.0 0.067 0.104 74.969 76.060 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.518 55.609 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.163 42.265 47.394 integrate_v_rspace 95 8.0 0.398 0.563 28.619 33.478 pw_transfer 2240 10.6 0.145 0.165 29.882 30.329 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.897 29.359 grid_integrate_task_list 95 9.0 23.922 28.977 23.922 28.977 ao_to_mo_and_store_B_mult_1 91 7.0 10.556 28.520 10.556 28.520 mp2_ri_gpw_compute_en 1 5.0 0.058 0.113 26.688 28.477 fft_wrap_pw1pw2_100 730 12.4 1.267 1.466 26.651 27.006 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.911 24.771 24.780 get_2c_integrals 1 6.0 0.006 0.050 20.332 20.383 compute_2c_integrals 1 7.0 0.005 0.024 19.298 19.311 compute_2c_integrals_loop_lm 1 8.0 0.002 0.009 18.832 19.152 mp2_eri_2c_integrate_gpw 1 9.0 1.742 1.872 18.830 19.152 fft3d_s 1823 13.4 18.430 18.813 18.443 18.826 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.326 14.326 calculate_wavefunction 91 8.0 2.015 2.051 9.748 9.993 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.578 8.771 9.520 potential_pw2rs 186 10.0 0.033 0.035 8.621 9.182 local_gemm 172 8.0 8.214 8.948 8.214 8.948 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.220 8.562 calc_potential_gpw 182 9.5 0.002 0.002 7.899 8.180 collocate_single_gaussian 91 10.0 0.017 0.027 7.868 8.140 mp2_ri_gpw_compute_en_comm 22 7.0 0.518 0.536 7.766 8.120 mp_sync 38 10.4 3.255 6.483 3.255 6.483 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.429 6.349 6.429 mp_sendrecv_dm3 2068 8.0 5.759 6.113 5.759 6.113 pw_gather_s 912 13.2 4.926 5.325 4.926 5.325 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.819128, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.432448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 592243. MP_ISendRecv 45220 5520. MP_Wait 60486 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.048 60.575 60.576 qs_mol_dyn_low 1 2.0 0.003 0.003 60.272 60.279 qs_forces 11 3.9 0.002 0.007 60.191 60.193 qs_energies 11 4.9 0.003 0.005 58.480 58.489 scf_env_do_scf 11 5.9 0.000 0.001 51.580 51.580 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 48.798 48.798 dbcsr_multiply_generic 2286 12.5 0.094 0.099 35.511 36.104 qs_scf_new_mos 108 7.5 0.000 0.001 35.210 35.514 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.210 35.514 ot_scf_mini 108 9.5 0.002 0.002 33.517 33.693 velocity_verlet 10 3.0 0.001 0.003 28.985 28.986 multiply_cannon 2286 13.5 0.189 0.197 26.308 28.194 multiply_cannon_loop 2286 14.5 1.451 1.540 24.959 26.514 ot_mini 108 10.5 0.001 0.001 21.090 21.332 qs_ot_get_derivative 108 11.5 0.001 0.001 17.931 18.168 mp_waitall_1 267858 16.1 10.595 16.531 10.595 16.531 multiply_cannon_metrocomm3 54864 15.5 0.068 0.071 5.604 13.227 multiply_cannon_multrec 54864 15.5 4.286 6.492 7.456 10.760 rebuild_ks_matrix 119 8.3 0.000 0.000 9.943 10.075 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 9.943 10.074 qs_ks_update_qs_env 119 7.6 0.001 0.001 8.821 8.944 qs_ot_get_p 119 10.4 0.001 0.001 7.549 7.853 multiply_cannon_sync_h2d 54864 15.5 5.795 7.716 5.795 7.716 mp_sum_l 7207 12.9 5.362 7.272 5.362 7.272 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 6.628 6.758 sum_up_and_integrate 119 10.3 0.013 0.015 6.203 6.210 integrate_v_rspace 119 11.3 0.002 0.002 6.190 6.199 qs_rho_update_rho_low 119 7.7 0.000 0.001 6.105 6.197 calculate_rho_elec 119 8.7 0.011 0.016 6.105 6.197 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.647 6.086 init_scf_run 11 5.9 0.000 0.001 5.514 5.514 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.514 5.514 rs_pw_transfer 974 11.9 0.011 0.012 4.501 4.654 dbcsr_mm_accdrv_process 76910 16.1 1.078 1.776 3.093 4.386 multiply_cannon_metrocomm1 54864 15.5 0.053 0.057 2.282 4.254 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.019 4.082 density_rs2pw 119 9.7 0.004 0.004 3.881 3.990 make_m2s 4572 13.5 0.054 0.056 3.530 3.650 make_images 4572 14.5 0.133 0.140 3.447 3.567 pw_transfer 1439 11.6 0.051 0.055 3.220 3.345 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 3.146 3.273 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.047 3.266 apply_single 119 13.6 0.000 0.000 3.046 3.266 potential_pw2rs 119 12.3 0.004 0.004 3.206 3.252 cp_dbcsr_syevd 50 12.0 0.002 0.003 3.094 3.096 calculate_dm_sparse 119 9.5 0.000 0.000 2.940 3.090 fft3d_ps 1201 14.6 0.362 0.467 2.930 3.051 ot_diis_step 108 11.5 0.006 0.006 2.905 2.907 calculate_first_density_matrix 1 7.0 0.000 0.000 2.855 2.860 init_scf_loop 11 6.9 0.000 0.000 2.757 2.758 mp_alltoall_d11v 2130 13.8 2.582 2.737 2.582 2.737 fft_wrap_pw1pw2_140 487 13.2 0.075 0.089 2.531 2.660 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.539 2.595 wfi_extrapolate 11 7.9 0.001 0.001 2.580 2.580 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.568 2.570 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.564 2.568 jit_kernel_multiply 13 15.8 1.957 2.563 1.957 2.563 cp_fm_redistribute_end 50 14.0 2.297 2.501 2.305 2.507 make_images_sizes 4572 15.5 0.004 0.005 1.978 2.390 mp_alltoall_i44 4572 16.5 1.973 2.386 1.973 2.386 cp_fm_diag_elpa_base 50 14.0 0.197 2.362 0.198 2.371 mp_sum_d 4129 12.0 1.494 2.162 1.494 2.162 acc_transpose_blocks 54864 15.5 0.210 0.241 1.656 2.136 grid_integrate_task_list 119 12.3 1.998 2.108 1.998 2.108 mp_alltoall_z22v 1201 16.6 1.988 2.092 1.988 2.092 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.004 2.051 mp_waitany 12084 13.8 1.538 1.653 1.538 1.653 prepare_preconditioner 11 7.9 0.000 0.000 1.461 1.492 make_preconditioner 11 8.9 0.000 0.000 1.461 1.492 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.445 1.463 yz_to_x 249 15.8 0.025 0.033 1.336 1.432 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.341 1.380 grid_collocate_task_list 119 9.7 1.290 1.355 1.290 1.355 make_images_data 4572 15.5 0.038 0.042 0.894 1.329 parallel_gemm_fm 81 9.0 0.000 0.000 1.288 1.318 parallel_gemm_fm_cosma 81 10.0 1.287 1.318 1.287 1.318 arnoldi_extremal 119 11.4 0.001 0.001 1.150 1.284 arnoldi_normal_ev 119 12.4 0.002 0.003 1.150 1.283 mp_allgather_i34 2286 14.5 0.879 1.243 0.879 1.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=60.576000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.181818, yerr=1.192262 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.510016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1421490. MP_ISendRecv 33558 37093. MP_Wait 40318 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.102 43.292 43.295 qs_mol_dyn_low 1 2.0 0.003 0.007 42.659 42.667 qs_forces 11 3.9 0.186 0.199 42.578 42.581 qs_energies 11 4.9 0.002 0.002 40.646 40.674 scf_env_do_scf 11 5.9 0.000 0.001 35.048 35.049 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 32.051 32.051 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.544 24.093 qs_scf_new_mos 108 7.5 0.001 0.001 22.328 22.567 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.327 22.566 ot_scf_mini 108 9.5 0.002 0.003 21.327 21.511 velocity_verlet 10 3.0 0.002 0.004 20.308 20.310 multiply_cannon 2286 13.5 0.209 0.217 17.377 19.388 multiply_cannon_loop 2286 14.5 0.900 0.976 16.054 18.085 ot_mini 108 10.5 0.001 0.001 13.365 13.636 mp_waitall_1 217478 16.2 7.589 12.700 7.589 12.700 qs_ot_get_derivative 108 11.5 0.001 0.001 10.656 10.842 multiply_cannon_metrocomm3 27432 15.5 0.068 0.071 4.990 10.486 multiply_cannon_multrec 27432 15.5 1.905 4.221 5.876 8.605 rebuild_ks_matrix 119 8.3 0.000 0.000 7.761 7.945 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.760 7.945 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.875 7.045 dbcsr_mm_accdrv_process 47894 16.0 2.957 5.219 3.902 5.761 qs_ot_get_p 119 10.4 0.001 0.001 4.781 5.030 mp_sum_l 7207 12.9 2.686 4.864 2.686 4.864 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.934 4.788 sum_up_and_integrate 119 10.3 0.024 0.026 4.628 4.633 integrate_v_rspace 119 11.3 0.002 0.002 4.604 4.610 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.179 4.328 apply_single 119 13.6 0.000 0.000 3.179 4.328 init_scf_run 11 5.9 0.000 0.001 4.302 4.303 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.302 4.303 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.926 3.979 calculate_rho_elec 119 8.7 0.021 0.024 3.926 3.979 make_m2s 4572 13.5 0.052 0.053 3.046 3.307 make_images 4572 14.5 0.201 0.239 2.957 3.217 rs_pw_transfer 974 11.9 0.010 0.011 3.048 3.178 qs_ot_p2m_diag 50 11.0 0.008 0.012 3.028 3.053 init_scf_loop 11 6.9 0.000 0.000 2.972 2.973 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.670 2.765 multiply_cannon_sync_h2d 27432 15.5 2.103 2.700 2.103 2.700 ot_diis_step 108 11.5 0.011 0.011 2.626 2.627 calculate_first_density_matrix 1 7.0 0.000 0.001 2.584 2.588 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.547 2.548 calculate_dm_sparse 119 9.5 0.000 0.000 2.234 2.327 potential_pw2rs 119 12.3 0.006 0.006 2.313 2.326 density_rs2pw 119 9.7 0.004 0.004 2.206 2.306 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.166 2.167 jit_kernel_multiply 11 16.1 0.893 2.143 0.893 2.143 cp_fm_redistribute_end 50 14.0 1.781 2.137 1.787 2.139 cp_fm_diag_elpa_base 50 14.0 0.341 2.067 0.350 2.098 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.004 2.095 2.098 pw_transfer 1439 11.6 0.063 0.066 1.987 2.027 prepare_preconditioner 11 7.9 0.000 0.000 1.940 1.966 make_preconditioner 11 8.9 0.000 0.000 1.940 1.966 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.897 1.940 grid_integrate_task_list 119 12.3 1.826 1.938 1.826 1.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.827 1.885 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.783 1.833 make_images_data 4572 15.5 0.045 0.052 1.343 1.830 mp_alltoall_d11v 2130 13.8 1.450 1.666 1.450 1.666 wfi_extrapolate 11 7.9 0.001 0.001 1.656 1.656 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.161 1.654 fft3d_ps 1201 14.6 0.502 0.552 1.604 1.644 mp_allgather_i34 2286 14.5 0.764 1.598 0.764 1.598 acc_transpose_blocks 27432 15.5 0.107 0.111 1.204 1.513 fft_wrap_pw1pw2_140 487 13.2 0.076 0.082 1.410 1.449 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.358 1.403 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.357 1.373 mp_sum_d 4129 12.0 0.785 1.313 0.785 1.313 grid_collocate_task_list 119 9.7 1.237 1.309 1.237 1.309 make_images_sizes 4572 15.5 0.005 0.005 0.891 1.214 mp_alltoall_i44 4572 16.5 0.886 1.209 0.886 1.209 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 0.985 1.010 rs_pw_transfer_PW2RS_50 119 14.3 0.586 0.603 0.916 0.969 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.947 0.961 dbcsr_dot_sd 1205 11.9 0.070 0.087 0.528 0.961 cp_fm_cholesky_invert 11 10.9 0.954 0.959 0.954 0.959 mp_alltoall_z22v 1201 16.6 0.829 0.902 0.829 0.902 acc_transpose_blocks_kernels 27432 16.5 0.184 0.271 0.674 0.895 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.295000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.636364, yerr=1.553455 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 523.366400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.100 0.187 38.598 38.600 qs_mol_dyn_low 1 2.0 0.003 0.004 37.885 37.894 qs_forces 11 3.9 0.003 0.007 37.786 37.790 qs_energies 11 4.9 0.005 0.018 36.073 36.081 scf_env_do_scf 11 5.9 0.001 0.002 30.642 30.642 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.494 27.495 dbcsr_multiply_generic 2286 12.5 0.097 0.110 18.915 19.184 velocity_verlet 10 3.0 0.016 0.035 18.171 18.172 qs_scf_new_mos 108 7.5 0.001 0.001 17.904 17.927 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.904 17.926 ot_scf_mini 108 9.5 0.003 0.004 17.032 17.054 multiply_cannon 2286 13.5 0.201 0.245 14.098 15.428 multiply_cannon_loop 2286 14.5 0.640 0.673 13.039 14.457 ot_mini 108 10.5 0.001 0.001 10.384 10.420 qs_ot_get_derivative 108 11.5 0.001 0.001 8.625 8.649 rebuild_ks_matrix 119 8.3 0.000 0.000 7.370 7.442 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.370 7.441 multiply_cannon_multrec 18288 15.5 1.876 2.712 6.727 6.955 mp_waitall_1 169478 16.3 4.984 6.597 4.984 6.597 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.514 6.581 dbcsr_mm_accdrv_process 38222 16.0 3.998 5.184 4.767 5.591 sum_up_and_integrate 119 10.3 0.030 0.031 4.714 4.727 integrate_v_rspace 119 11.3 0.002 0.003 4.683 4.696 qs_ot_get_p 119 10.4 0.001 0.001 4.126 4.173 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.040 4.055 calculate_rho_elec 119 8.7 0.030 0.031 4.039 4.055 init_scf_run 11 5.9 0.000 0.001 3.972 3.972 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.971 3.972 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.281 3.903 multiply_cannon_metrocomm3 18288 15.5 0.045 0.047 2.085 3.638 rs_pw_transfer 974 11.9 0.009 0.010 3.269 3.434 init_scf_loop 11 6.9 0.001 0.004 3.119 3.120 make_m2s 4572 13.5 0.044 0.047 2.550 2.733 make_images 4572 14.5 0.199 0.222 2.464 2.646 density_rs2pw 119 9.7 0.004 0.004 2.538 2.638 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.618 2.632 calculate_first_density_matrix 1 7.0 0.001 0.002 2.601 2.603 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.156 2.497 apply_single 119 13.6 0.000 0.000 2.156 2.497 potential_pw2rs 119 12.3 0.007 0.008 2.418 2.436 mp_sum_l 7207 12.9 1.915 2.399 1.915 2.399 pw_transfer 1439 11.6 0.063 0.066 2.272 2.305 prepare_preconditioner 11 7.9 0.000 0.000 2.222 2.231 make_preconditioner 11 8.9 0.001 0.003 2.222 2.231 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.222 2.223 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.181 2.217 make_full_inverse_cholesky 11 9.9 0.000 0.001 2.049 2.135 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.060 2.073 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.004 2.006 calculate_dm_sparse 119 9.5 0.000 0.000 1.989 2.005 fft3d_ps 1201 14.6 0.511 0.531 1.870 1.905 grid_integrate_task_list 119 12.3 1.807 1.905 1.807 1.905 jit_kernel_multiply 10 16.3 0.719 1.886 0.719 1.886 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.857 1.858 cp_fm_redistribute_end 50 14.0 1.363 1.807 1.365 1.807 cp_fm_diag_elpa_base 50 14.0 0.423 1.718 0.439 1.758 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 1.726 1.749 ot_diis_step 108 11.5 0.011 0.011 1.701 1.702 fft_wrap_pw1pw2_140 487 13.2 0.085 0.088 1.568 1.602 multiply_cannon_sync_h2d 18288 15.5 1.320 1.516 1.320 1.516 make_images_data 4572 15.5 0.045 0.049 1.157 1.376 grid_collocate_task_list 119 9.7 1.203 1.322 1.203 1.322 multiply_cannon_metrocomm1 18288 15.5 0.029 0.030 0.597 1.321 wfi_extrapolate 11 7.9 0.001 0.001 1.314 1.315 acc_transpose_blocks 18288 15.5 0.075 0.077 1.266 1.289 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.229 1.238 mp_alltoall_z22v 1201 16.6 1.133 1.232 1.133 1.232 mp_alltoall_d11v 2130 13.8 1.022 1.208 1.022 1.208 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.163 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.982 1.159 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.085 1.089 mp_allgather_i34 2286 14.5 0.510 1.067 0.510 1.067 cp_fm_cholesky_invert 11 10.9 1.001 1.006 1.001 1.006 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.944 0.950 make_images_sizes 4572 15.5 0.005 0.005 0.692 0.926 mp_alltoall_i44 4572 16.5 0.686 0.921 0.686 0.921 dbcsr_complete_redistribute 329 12.2 0.097 0.160 0.717 0.845 acc_transpose_blocks_kernels 18288 16.5 0.212 0.220 0.825 0.840 mp_waitany 9880 13.7 0.690 0.840 0.690 0.840 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.124 0.636 0.783 rs_pw_transfer_PW2RS_50 119 14.3 0.403 0.411 0.738 0.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.600000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.909091, yerr=2.778310 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.913792E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.052 39.235 39.236 qs_mol_dyn_low 1 2.0 0.003 0.004 38.900 38.908 qs_forces 11 3.9 0.002 0.003 38.832 38.833 qs_energies 11 4.9 0.002 0.006 37.049 37.053 scf_env_do_scf 11 5.9 0.001 0.001 31.411 31.413 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.475 27.476 velocity_verlet 10 3.0 0.001 0.002 20.149 20.151 dbcsr_multiply_generic 2286 12.5 0.101 0.104 19.977 20.136 qs_scf_new_mos 108 7.5 0.001 0.001 18.163 18.225 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.162 18.224 ot_scf_mini 108 9.5 0.003 0.004 17.136 17.188 multiply_cannon 2286 13.5 0.228 0.270 15.178 15.677 multiply_cannon_loop 2286 14.5 0.936 0.968 13.893 14.637 ot_mini 108 10.5 0.001 0.001 10.654 10.726 multiply_cannon_multrec 27432 15.5 2.335 3.005 8.633 9.007 qs_ot_get_derivative 108 11.5 0.001 0.001 8.679 8.734 dbcsr_mm_accdrv_process 47916 15.9 5.287 6.798 6.205 7.381 rebuild_ks_matrix 119 8.3 0.000 0.000 7.240 7.314 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.240 7.314 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.478 6.544 mp_waitall_1 145218 16.4 3.408 4.479 3.408 4.479 sum_up_and_integrate 119 10.3 0.034 0.038 4.315 4.323 integrate_v_rspace 119 11.3 0.002 0.003 4.280 4.289 init_scf_run 11 5.9 0.000 0.001 4.187 4.188 scf_env_initial_rho_setup 11 6.9 0.011 0.028 4.187 4.187 init_scf_loop 11 6.9 0.001 0.003 3.908 3.909 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.833 3.875 calculate_rho_elec 119 8.7 0.040 0.046 3.832 3.874 qs_ot_get_p 119 10.4 0.001 0.001 3.659 3.737 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.263 3.670 make_m2s 4572 13.5 0.053 0.055 2.891 3.162 make_images 4572 14.5 0.274 0.336 2.784 3.054 prepare_preconditioner 11 7.9 0.000 0.000 2.960 2.969 make_preconditioner 11 8.9 0.000 0.002 2.960 2.968 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.528 2.870 rs_pw_transfer 974 11.9 0.009 0.009 2.681 2.745 calculate_first_density_matrix 1 7.0 0.000 0.002 2.525 2.537 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.369 2.396 density_rs2pw 119 9.7 0.003 0.004 2.264 2.330 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.138 2.277 apply_single 119 13.6 0.000 0.000 2.137 2.276 pw_transfer 1439 11.6 0.063 0.066 2.159 2.192 calculate_dm_sparse 119 9.5 0.000 0.000 2.135 2.192 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.069 2.106 multiply_cannon_metrocomm3 27432 15.5 0.038 0.038 0.908 2.076 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.035 2.048 potential_pw2rs 119 12.3 0.008 0.009 2.024 2.036 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.989 1.991 ot_diis_step 108 11.5 0.012 0.012 1.919 1.919 grid_integrate_task_list 119 12.3 1.821 1.898 1.821 1.898 mp_sum_l 7207 12.9 1.492 1.852 1.492 1.852 fft3d_ps 1201 14.6 0.540 0.590 1.751 1.781 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.720 1.748 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.677 1.678 fft_wrap_pw1pw2_140 487 13.2 0.083 0.091 1.601 1.639 wfi_extrapolate 11 7.9 0.001 0.001 1.595 1.595 acc_transpose_blocks 27432 15.5 0.111 0.113 1.479 1.508 make_images_data 4572 15.5 0.045 0.049 1.170 1.481 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.390 1.394 cp_fm_redistribute_end 50 14.0 0.912 1.360 0.914 1.362 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.994 1.338 cp_fm_diag_elpa_base 50 14.0 0.423 1.289 0.443 1.331 grid_collocate_task_list 119 9.7 1.221 1.319 1.221 1.319 jit_kernel_multiply 7 15.9 0.857 1.305 0.857 1.305 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.227 1.249 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.249 dbcsr_complete_redistribute 329 12.2 0.117 0.149 0.953 1.240 mp_alltoall_d11v 2130 13.8 1.037 1.230 1.037 1.230 cp_fm_upper_to_full 72 13.5 0.830 1.157 0.830 1.157 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.152 1.154 mp_alltoall_z22v 1201 16.6 1.092 1.123 1.092 1.123 multiply_cannon_sync_h2d 27432 15.5 0.975 1.050 0.975 1.050 multiply_cannon_metrocomm1 27432 15.5 0.033 0.034 0.443 0.978 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.685 0.966 cp_fm_cholesky_invert 11 10.9 0.950 0.953 0.950 0.953 make_images_sizes 4572 15.5 0.005 0.005 0.692 0.896 mp_alltoall_i44 4572 16.5 0.687 0.891 0.687 0.891 acc_transpose_blocks_kernels 27432 16.5 0.270 0.277 0.857 0.878 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.863 0.873 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.869 mp_alltoall_i22 627 13.8 0.516 0.829 0.516 0.829 mp_allgather_i34 2286 14.5 0.577 0.823 0.577 0.823 mp_sum_d 4127 12.0 0.664 0.812 0.664 0.812 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.236000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=527.545455, yerr=2.535696 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 607.674368E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.029 28.135 28.135 qs_mol_dyn_low 1 2.0 0.003 0.003 27.942 27.949 qs_forces 11 3.9 0.002 0.003 27.727 27.729 qs_energies 11 4.9 0.001 0.001 26.050 26.053 scf_env_do_scf 11 5.9 0.001 0.006 21.144 21.144 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 18.629 18.629 velocity_verlet 10 3.0 0.006 0.008 14.303 14.315 dbcsr_multiply_generic 2286 12.5 0.092 0.094 12.269 12.406 qs_scf_new_mos 108 7.5 0.001 0.001 11.067 11.089 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.066 11.088 ot_scf_mini 108 9.5 0.002 0.002 10.383 10.410 multiply_cannon 2286 13.5 0.229 0.236 9.692 10.241 multiply_cannon_loop 2286 14.5 0.330 0.342 8.786 9.038 multiply_cannon_multrec 9144 15.5 1.610 1.832 5.777 5.997 ot_mini 108 10.5 0.001 0.001 5.842 5.875 rebuild_ks_matrix 119 8.3 0.000 0.000 5.802 5.824 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.801 5.823 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.184 5.204 qs_ot_get_derivative 108 11.5 0.001 0.001 4.531 4.558 dbcsr_mm_accdrv_process 12550 15.8 3.253 3.755 4.066 4.153 sum_up_and_integrate 119 10.3 0.038 0.042 3.561 3.567 integrate_v_rspace 119 11.3 0.003 0.003 3.523 3.530 init_scf_run 11 5.9 0.000 0.001 3.484 3.484 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.484 3.484 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.202 3.208 calculate_rho_elec 119 8.7 0.059 0.061 3.201 3.208 qs_ot_get_p 119 10.4 0.001 0.001 2.765 2.809 init_scf_loop 11 6.9 0.000 0.000 2.487 2.488 calculate_first_density_matrix 1 7.0 0.000 0.000 2.365 2.366 mp_waitall_1 121218 16.5 1.823 2.317 1.823 2.317 make_m2s 4572 13.5 0.034 0.035 1.795 1.934 grid_integrate_task_list 119 12.3 1.849 1.908 1.849 1.908 make_images 4572 14.5 0.268 0.304 1.706 1.844 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.820 1.822 prepare_preconditioner 11 7.9 0.000 0.000 1.751 1.756 make_preconditioner 11 8.9 0.000 0.000 1.751 1.756 pw_transfer 1439 11.6 0.063 0.065 1.726 1.733 calculate_dm_sparse 119 9.5 0.000 0.000 1.710 1.732 rs_pw_transfer 974 11.9 0.008 0.008 1.633 1.725 density_rs2pw 119 9.7 0.004 0.004 1.603 1.697 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.637 1.661 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.646 1.647 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.635 1.643 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.621 1.622 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.520 1.534 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.469 1.475 jit_kernel_multiply 8 15.9 0.775 1.417 0.775 1.417 potential_pw2rs 119 12.3 0.010 0.011 1.341 1.344 grid_collocate_task_list 119 9.7 1.271 1.341 1.271 1.341 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.339 1.340 fft_wrap_pw1pw2_140 487 13.2 0.082 0.087 1.325 1.334 cp_fm_redistribute_end 50 14.0 0.665 1.317 0.665 1.318 fft3d_ps 1201 14.6 0.542 0.552 1.302 1.310 cp_fm_diag_elpa_base 50 14.0 0.608 1.242 0.651 1.302 ot_diis_step 108 11.5 0.012 0.013 1.297 1.297 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.209 1.211 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.171 1.193 apply_single 119 13.6 0.000 0.000 1.171 1.193 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.130 1.145 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.120 1.125 wfi_extrapolate 11 7.9 0.001 0.001 1.074 1.074 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.823 1.052 make_images_data 4572 15.5 0.039 0.042 0.846 1.031 acc_transpose_blocks 9144 15.5 0.039 0.039 0.916 0.927 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.921 mp_alltoall_d11v 2130 13.8 0.821 0.893 0.821 0.893 cp_fm_cholesky_invert 11 10.9 0.873 0.876 0.873 0.876 multiply_cannon_sync_h2d 9144 15.5 0.717 0.813 0.717 0.813 multiply_cannon_metrocomm3 9144 15.5 0.018 0.019 0.433 0.741 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.739 0.741 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.730 0.740 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.666 0.714 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.676 0.685 mp_allgather_i34 2286 14.5 0.239 0.684 0.239 0.684 mp_alltoall_z22v 1201 16.6 0.634 0.666 0.634 0.666 qs_create_task_list 11 7.9 0.001 0.001 0.546 0.570 generate_qs_task_list 11 8.9 0.191 0.213 0.545 0.569 jit_kernel_transpose 5 15.6 0.559 0.566 0.559 0.566 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.135000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=572.545455, yerr=8.958121 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 754.913280E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.085 41.412 41.412 qs_mol_dyn_low 1 2.0 0.003 0.003 40.629 40.636 qs_forces 11 3.9 0.015 0.017 40.553 40.555 qs_energies 11 4.9 0.001 0.001 38.601 38.604 scf_env_do_scf 11 5.9 0.001 0.002 32.607 32.608 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 24.778 24.779 velocity_verlet 10 3.0 0.002 0.003 22.856 22.861 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.660 17.833 qs_scf_new_mos 108 7.5 0.001 0.001 15.872 15.967 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.871 15.966 ot_scf_mini 108 9.5 0.002 0.002 14.795 14.892 multiply_cannon 2286 13.5 0.301 0.304 13.711 14.642 multiply_cannon_loop 2286 14.5 0.343 0.349 12.424 13.337 ot_mini 108 10.5 0.001 0.001 8.889 9.004 multiply_cannon_multrec 9144 15.5 3.390 4.715 8.614 8.680 init_scf_loop 11 6.9 0.000 0.000 7.801 7.802 rebuild_ks_matrix 119 8.3 0.000 0.000 6.833 6.972 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.832 6.971 qs_ot_get_derivative 108 11.5 0.001 0.001 6.852 6.949 prepare_preconditioner 11 7.9 0.000 0.000 6.867 6.880 make_preconditioner 11 8.9 0.000 0.000 6.867 6.880 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.425 6.756 dbcsr_mm_accdrv_process 12550 15.8 4.074 5.634 5.101 6.410 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.177 6.302 cp_fm_upper_to_full 72 14.2 3.158 4.540 3.158 4.540 init_scf_run 11 5.9 0.000 0.001 3.957 3.957 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.957 3.957 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.766 3.821 calculate_rho_elec 119 8.7 0.118 0.121 3.766 3.821 sum_up_and_integrate 119 10.3 0.064 0.065 3.674 3.682 integrate_v_rspace 119 11.3 0.003 0.003 3.610 3.619 mp_waitall_1 97218 16.6 2.638 3.560 2.638 3.560 qs_ot_get_p 119 10.4 0.001 0.001 3.180 3.315 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.482 2.916 dbcsr_complete_redistribute 329 12.2 0.288 0.293 1.994 2.798 calculate_first_density_matrix 1 7.0 0.000 0.000 2.484 2.545 make_m2s 4572 13.5 0.037 0.038 2.356 2.534 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.677 2.476 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.204 2.466 apply_single 119 13.6 0.000 0.000 2.204 2.466 make_images 4572 14.5 0.352 0.383 2.236 2.412 calculate_dm_sparse 119 9.5 0.000 0.000 2.252 2.305 mp_alltoall_i22 627 13.8 1.418 2.251 1.418 2.251 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.436 2.229 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.314 2.196 pw_transfer 1439 11.6 0.066 0.067 2.051 2.056 grid_integrate_task_list 119 12.3 2.006 2.024 2.006 2.024 ot_diis_step 108 11.5 0.014 0.014 2.012 2.013 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.957 1.960 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.922 1.923 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.906 1.908 density_rs2pw 119 9.7 0.003 0.003 1.809 1.842 mp_sum_l 7207 12.9 1.058 1.790 1.058 1.790 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.735 1.785 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.772 1.772 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.692 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.634 1.638 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.628 1.629 fft3d_ps 1201 14.6 0.571 0.580 1.591 1.594 jit_kernel_multiply 8 15.3 1.000 1.503 1.000 1.503 cp_fm_cholesky_invert 11 10.9 1.478 1.481 1.478 1.481 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.208 1.470 grid_collocate_task_list 119 9.7 1.447 1.461 1.447 1.461 rs_pw_transfer 974 11.9 0.009 0.009 1.399 1.428 make_images_data 4572 15.5 0.043 0.046 1.154 1.379 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.322 1.376 wfi_extrapolate 11 7.9 0.001 0.001 1.355 1.355 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.337 1.337 cp_fm_diag_elpa_base 50 14.0 1.193 1.245 1.335 1.335 mp_alltoall_d11v 2130 13.8 1.231 1.298 1.231 1.298 potential_pw2rs 119 12.3 0.014 0.014 1.207 1.209 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.137 1.156 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.110 1.119 multiply_cannon_sync_h2d 9144 15.5 1.042 1.046 1.042 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.987 1.030 qs_create_task_list 11 7.9 0.006 0.014 0.957 0.970 generate_qs_task_list 11 8.9 0.375 0.395 0.950 0.964 acc_transpose_blocks 9144 15.5 0.038 0.039 0.957 0.962 mp_alltoall_z22v 1201 16.6 0.885 0.905 0.885 0.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.881 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.412000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=708.818182, yerr=13.421951 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.398400E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1256205. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.071 82.677 82.678 qs_mol_dyn_low 1 2.0 0.003 0.003 82.297 82.307 qs_forces 11 3.9 0.002 0.003 82.220 82.223 qs_energies 11 4.9 0.002 0.002 79.406 79.422 scf_env_do_scf 11 5.9 0.000 0.001 70.391 70.393 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.656 64.656 dbcsr_multiply_generic 2055 12.4 0.105 0.107 51.746 51.945 qs_scf_new_mos 99 7.5 0.000 0.001 47.588 47.709 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.587 47.709 ot_scf_mini 99 9.5 0.002 0.002 45.211 45.309 multiply_cannon 2055 13.4 0.181 0.189 42.188 43.200 multiply_cannon_loop 2055 14.4 1.539 1.592 41.175 42.109 velocity_verlet 10 3.0 0.004 0.015 41.956 41.957 ot_mini 99 10.5 0.001 0.001 27.377 27.469 qs_ot_get_derivative 99 11.5 0.001 0.001 20.593 20.681 multiply_cannon_multrec 49320 15.4 12.240 12.786 17.190 17.912 rebuild_ks_matrix 110 8.3 0.000 0.001 14.484 14.592 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.016 14.484 14.592 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.736 12.837 mp_waitall_1 241148 16.1 11.405 12.350 11.405 12.350 multiply_cannon_sync_h2d 49320 15.4 10.094 10.673 10.094 10.673 qs_ot_get_p 110 10.4 0.001 0.001 9.292 9.390 multiply_cannon_metrocomm3 49320 15.4 0.077 0.080 6.643 7.729 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.164 7.628 apply_single 110 13.6 0.000 0.000 7.164 7.627 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.946 7.504 sum_up_and_integrate 110 10.3 0.037 0.043 7.064 7.083 integrate_v_rspace 110 11.3 0.003 0.003 7.027 7.053 init_scf_run 11 5.9 0.000 0.001 6.964 6.964 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.964 6.964 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.774 6.829 ot_diis_step 99 11.5 0.006 0.006 6.529 6.529 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.310 6.429 calculate_rho_elec 110 8.6 0.020 0.024 6.309 6.428 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.078 6.127 mp_sum_l 6514 12.8 5.070 5.903 5.070 5.903 init_scf_loop 11 6.9 0.000 0.000 5.707 5.707 dbcsr_mm_accdrv_process 87628 16.1 2.007 2.181 4.829 5.189 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.109 5.109 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.604 4.604 cp_fm_redistribute_end 48 14.0 4.006 4.577 4.010 4.578 cp_fm_diag_elpa_base 48 14.0 0.562 4.505 0.566 4.529 make_m2s 4110 13.4 0.060 0.064 4.031 4.128 rs_pw_transfer 902 11.9 0.012 0.013 3.877 4.090 wfi_extrapolate 11 7.9 0.001 0.001 4.042 4.042 make_images 4110 14.4 0.178 0.191 3.936 4.036 calculate_dm_sparse 110 9.5 0.000 0.001 3.752 3.859 density_rs2pw 110 9.6 0.004 0.004 3.395 3.597 prepare_preconditioner 11 7.9 0.000 0.000 3.553 3.566 make_preconditioner 11 8.9 0.000 0.000 3.553 3.566 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.483 3.559 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.553 3.556 grid_integrate_task_list 110 12.3 3.254 3.430 3.254 3.430 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.329 3.369 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.273 3.324 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.229 3.273 pw_transfer 1331 11.6 0.053 0.065 2.937 3.005 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.851 2.920 calculate_first_density_matrix 1 7.0 0.000 0.000 2.826 2.837 potential_pw2rs 110 12.3 0.006 0.007 2.666 2.690 mp_alltoall_d11v 2046 13.8 2.159 2.599 2.159 2.599 jit_kernel_multiply 13 15.9 2.546 2.561 2.546 2.561 fft3d_ps 1111 14.6 0.756 0.845 2.390 2.449 fft_wrap_pw1pw2_140 451 13.1 0.168 0.188 2.361 2.429 acc_transpose_blocks 49320 15.4 0.206 0.223 2.141 2.253 grid_collocate_task_list 110 9.6 2.083 2.190 2.083 2.190 mp_waitany 14300 13.8 1.844 2.126 1.844 2.126 make_images_data 4110 15.4 0.043 0.046 1.824 1.965 cp_fm_cholesky_invert 11 10.9 1.912 1.916 1.912 1.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.887 1.899 mp_sum_d 3883 11.9 1.434 1.859 1.434 1.859 hybrid_alltoall_any 4261 16.3 0.081 0.479 1.596 1.837 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.653 1.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.678000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.727273, yerr=2.987578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.665408E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2909509. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.053 69.094 69.097 qs_mol_dyn_low 1 2.0 0.003 0.007 68.727 68.737 qs_forces 11 3.9 0.002 0.003 68.645 68.647 qs_energies 11 4.9 0.027 0.044 65.312 65.318 scf_env_do_scf 11 5.9 0.001 0.002 56.696 56.699 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 48.688 48.696 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.826 38.009 velocity_verlet 10 3.0 0.001 0.002 36.375 36.378 qs_scf_new_mos 99 7.5 0.001 0.001 32.683 32.851 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.682 32.851 multiply_cannon 2055 13.4 0.221 0.243 31.051 32.490 ot_scf_mini 99 9.5 0.003 0.005 31.020 31.176 multiply_cannon_loop 2055 14.4 0.928 0.947 29.770 30.804 ot_mini 99 10.5 0.001 0.001 18.280 18.447 multiply_cannon_multrec 24660 15.4 7.653 9.647 13.897 16.017 rebuild_ks_matrix 110 8.3 0.000 0.000 13.742 13.923 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.022 13.741 13.922 qs_ot_get_derivative 99 11.5 0.001 0.001 12.492 12.654 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.104 12.257 mp_waitall_1 186928 16.3 7.964 10.159 7.964 10.159 multiply_cannon_sync_h2d 24660 15.4 7.056 8.537 7.056 8.537 init_scf_loop 11 6.9 0.001 0.008 7.972 8.018 multiply_cannon_metrocomm3 24660 15.4 0.071 0.073 5.131 7.851 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.512 7.225 apply_single 110 13.6 0.000 0.001 6.512 7.225 sum_up_and_integrate 110 10.3 0.052 0.060 6.604 6.619 integrate_v_rspace 110 11.3 0.002 0.002 6.551 6.570 dbcsr_mm_accdrv_process 52282 16.1 4.674 5.433 6.082 6.504 qs_ot_get_p 110 10.4 0.001 0.003 6.163 6.339 init_scf_run 11 5.9 0.000 0.001 6.203 6.205 scf_env_initial_rho_setup 11 6.9 0.002 0.005 6.203 6.205 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.873 5.882 calculate_rho_elec 110 8.6 0.039 0.047 5.872 5.881 prepare_preconditioner 11 7.9 0.000 0.000 5.834 5.856 make_preconditioner 11 8.9 0.002 0.014 5.834 5.856 ot_diis_step 99 11.5 0.010 0.010 5.740 5.740 make_full_inverse_cholesky 11 9.9 0.001 0.003 5.412 5.586 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.745 5.448 make_m2s 4110 13.4 0.056 0.059 4.298 4.744 make_images 4110 14.4 0.400 0.450 4.190 4.632 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.158 4.177 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.713 3.714 wfi_extrapolate 11 7.9 0.001 0.001 3.514 3.514 pw_transfer 1331 11.6 0.065 0.070 3.345 3.482 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.238 3.375 density_rs2pw 110 9.6 0.004 0.004 3.146 3.328 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.316 3.318 grid_integrate_task_list 110 12.3 3.147 3.306 3.147 3.306 rs_pw_transfer 902 11.9 0.012 0.013 3.069 3.276 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.179 3.263 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.149 3.151 cp_fm_redistribute_end 48 14.0 2.344 3.121 2.346 3.123 cp_fm_diag_elpa_base 48 14.0 0.742 3.009 0.773 3.096 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.973 3.021 calculate_dm_sparse 110 9.5 0.001 0.001 2.983 3.014 make_images_data 4110 15.4 0.047 0.051 2.420 2.874 fft_wrap_pw1pw2_140 451 13.1 0.200 0.218 2.714 2.853 hybrid_alltoall_any 4261 16.3 0.102 0.440 2.084 2.764 fft3d_ps 1111 14.6 1.077 1.270 2.596 2.721 cp_fm_cholesky_invert 11 10.9 2.692 2.699 2.692 2.699 calculate_first_density_matrix 1 7.0 0.001 0.008 2.601 2.605 mp_sum_l 6514 12.8 1.860 2.523 1.860 2.523 jit_kernel_multiply 13 16.3 1.061 2.513 1.061 2.513 potential_pw2rs 110 12.3 0.008 0.009 2.452 2.470 grid_collocate_task_list 110 9.6 2.121 2.248 2.121 2.248 mp_alltoall_d11v 2046 13.8 1.799 2.156 1.799 2.156 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.948 1.965 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.857 1.859 multiply_cannon_metrocomm4 22605 15.4 0.074 0.077 0.776 1.859 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.802 1.828 mp_irecv_dv 57340 16.2 0.654 1.739 0.654 1.739 mp_allgather_i34 2055 14.4 0.610 1.640 0.610 1.640 acc_transpose_blocks 24660 15.4 0.106 0.110 1.538 1.564 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.553 1.564 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.371 1.479 cp_fm_cholesky_decompose 22 10.9 1.466 1.473 1.466 1.473 dbcsr_complete_redistribute 325 12.2 0.239 0.300 1.170 1.446 mp_waitany 10164 13.8 1.231 1.382 1.231 1.382 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.097000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.363636, yerr=7.351842 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 656.285696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.190 60.516 60.519 qs_mol_dyn_low 1 2.0 0.003 0.003 59.668 59.678 qs_forces 11 3.9 0.013 0.017 59.603 59.604 qs_energies 11 4.9 0.005 0.028 56.389 56.401 scf_env_do_scf 11 5.9 0.001 0.003 48.286 48.289 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 39.495 39.497 velocity_verlet 10 3.0 0.001 0.002 32.595 32.597 dbcsr_multiply_generic 2055 12.4 0.108 0.114 28.349 28.564 qs_scf_new_mos 99 7.5 0.001 0.001 24.942 25.043 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.941 25.042 ot_scf_mini 99 9.5 0.003 0.003 23.710 23.845 multiply_cannon 2055 13.4 0.212 0.222 22.003 23.245 multiply_cannon_loop 2055 14.4 0.619 0.632 20.840 21.838 ot_mini 99 10.5 0.001 0.001 13.478 13.617 rebuild_ks_matrix 110 8.3 0.000 0.000 12.237 12.352 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 12.237 12.352 multiply_cannon_multrec 16440 15.4 3.998 5.403 9.775 11.113 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.780 10.889 mp_waitall_1 146766 16.3 7.067 10.303 7.067 10.303 qs_ot_get_derivative 99 11.5 0.001 0.001 9.088 9.224 init_scf_loop 11 6.9 0.001 0.005 8.755 8.757 multiply_cannon_metrocomm3 16440 15.4 0.043 0.043 4.073 7.158 prepare_preconditioner 11 7.9 0.000 0.000 6.943 6.962 make_preconditioner 11 8.9 0.000 0.002 6.943 6.962 make_full_inverse_cholesky 11 9.9 0.000 0.001 6.278 6.643 sum_up_and_integrate 110 10.3 0.060 0.061 6.482 6.513 integrate_v_rspace 110 11.3 0.002 0.003 6.422 6.454 dbcsr_mm_accdrv_process 34862 16.1 4.657 5.126 5.630 5.779 init_scf_run 11 5.9 0.000 0.001 5.634 5.634 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.634 5.634 qs_ot_get_p 110 10.4 0.001 0.001 5.443 5.593 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.578 5.587 calculate_rho_elec 110 8.6 0.058 0.058 5.577 5.586 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.927 5.328 apply_single 110 13.6 0.000 0.000 4.926 5.328 make_m2s 4110 13.4 0.049 0.051 4.255 4.627 make_images 4110 14.4 0.391 0.509 4.140 4.510 ot_diis_step 99 11.5 0.011 0.011 4.361 4.361 multiply_cannon_sync_h2d 16440 15.4 3.758 4.350 3.758 4.350 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.827 3.830 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.104 3.710 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.486 3.486 grid_integrate_task_list 110 12.3 3.185 3.382 3.185 3.382 pw_transfer 1331 11.6 0.064 0.070 3.243 3.263 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.137 3.159 density_rs2pw 110 9.6 0.004 0.004 2.857 3.090 wfi_extrapolate 11 7.9 0.001 0.001 3.003 3.004 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.956 2.958 make_images_data 4110 15.4 0.044 0.048 2.492 2.952 rs_pw_transfer 902 11.9 0.010 0.011 2.735 2.950 cp_fm_redistribute_end 48 14.0 1.835 2.931 1.838 2.931 cp_fm_diag_elpa_base 48 14.0 1.031 2.801 1.089 2.908 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.173 2.839 cp_fm_cholesky_invert 11 10.9 2.803 2.809 2.803 2.809 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.808 2.809 fft_wrap_pw1pw2_140 451 13.1 0.210 0.213 2.684 2.709 calculate_dm_sparse 110 9.5 0.001 0.001 2.540 2.574 calculate_first_density_matrix 1 7.0 0.001 0.003 2.544 2.545 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.464 2.503 fft3d_ps 1111 14.6 1.067 1.079 2.451 2.475 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.327 2.404 grid_collocate_task_list 110 9.6 2.171 2.372 2.171 2.372 multiply_cannon_metrocomm4 14385 15.4 0.045 0.049 0.835 2.326 potential_pw2rs 110 12.3 0.011 0.011 2.292 2.320 mp_irecv_dv 48980 15.7 0.766 2.204 0.766 2.204 mp_alltoall_d11v 2046 13.8 1.805 2.192 1.805 2.192 mp_sum_l 6514 12.8 1.529 2.146 1.529 2.146 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.002 2.012 dbcsr_complete_redistribute 325 12.2 0.329 0.361 1.477 1.943 cp_fm_upper_to_full 70 13.6 1.391 1.882 1.391 1.882 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.658 1.671 cp_fm_cholesky_decompose 22 10.9 1.630 1.655 1.630 1.655 mp_allgather_i34 2055 14.4 0.487 1.604 0.487 1.604 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.499 1.510 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.504 jit_kernel_multiply 8 16.7 0.588 1.485 0.588 1.485 mp_waitany 17072 13.8 1.181 1.467 1.181 1.467 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.003 1.466 rs_gather_matrices 110 12.3 0.138 0.151 0.892 1.274 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.257 1.264 acc_transpose_blocks 16440 15.4 0.072 0.075 1.227 1.260 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.519000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.181818, yerr=6.350636 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 727.240704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.107 66.600 66.601 qs_mol_dyn_low 1 2.0 0.006 0.024 66.169 66.178 qs_forces 11 3.9 0.005 0.023 66.071 66.072 qs_energies 11 4.9 0.008 0.021 62.676 62.682 scf_env_do_scf 11 5.9 0.001 0.002 54.264 54.267 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 42.187 42.187 velocity_verlet 10 3.0 0.002 0.002 36.721 36.723 dbcsr_multiply_generic 2055 12.4 0.116 0.118 30.107 30.316 qs_scf_new_mos 99 7.5 0.001 0.001 27.637 27.741 qs_scf_loop_do_ot 99 8.5 0.001 0.002 27.636 27.740 ot_scf_mini 99 9.5 0.003 0.005 25.981 26.094 multiply_cannon 2055 13.4 0.242 0.260 22.800 24.165 multiply_cannon_loop 2055 14.4 0.879 0.910 21.382 21.898 ot_mini 99 10.5 0.001 0.001 14.662 14.801 multiply_cannon_multrec 24660 15.4 4.203 6.621 12.630 13.765 rebuild_ks_matrix 110 8.3 0.000 0.000 12.015 12.148 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.029 12.015 12.147 init_scf_loop 11 6.9 0.001 0.005 12.035 12.036 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.652 10.767 qs_ot_get_derivative 99 11.5 0.001 0.001 10.409 10.526 prepare_preconditioner 11 7.9 0.000 0.000 10.295 10.311 make_preconditioner 11 8.9 0.001 0.004 10.295 10.311 make_full_inverse_cholesky 11 9.9 0.000 0.001 8.459 9.984 dbcsr_mm_accdrv_process 52304 16.0 6.712 7.957 8.281 9.191 mp_waitall_1 126806 16.4 4.882 6.980 4.882 6.980 sum_up_and_integrate 110 10.3 0.068 0.072 6.402 6.413 qs_ot_get_p 110 10.4 0.001 0.001 6.252 6.385 integrate_v_rspace 110 11.3 0.003 0.003 6.334 6.347 make_m2s 4110 13.4 0.059 0.060 5.538 5.816 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.665 5.676 calculate_rho_elec 110 8.6 0.077 0.081 5.664 5.675 make_images 4110 14.4 0.575 0.694 5.397 5.671 init_scf_run 11 5.9 0.000 0.002 5.627 5.627 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.626 5.627 cp_fm_upper_to_full 70 13.8 3.390 4.838 3.390 4.838 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.372 4.387 ot_diis_step 99 11.5 0.011 0.011 4.202 4.203 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.074 4.191 apply_single 110 13.6 0.000 0.000 4.073 4.191 dbcsr_complete_redistribute 325 12.2 0.418 0.460 2.730 3.912 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.846 3.846 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.710 3.566 grid_integrate_task_list 110 12.3 3.280 3.421 3.280 3.421 multiply_cannon_sync_h2d 24660 15.4 3.196 3.375 3.196 3.375 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.187 3.363 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.289 3.352 pw_transfer 1331 11.6 0.064 0.072 3.316 3.342 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.225 3.287 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.270 3.272 cp_fm_redistribute_end 48 14.0 1.628 3.238 1.629 3.240 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.211 3.238 cp_fm_diag_elpa_base 48 14.0 1.524 3.100 1.607 3.215 make_images_data 4110 15.4 0.046 0.050 2.759 3.072 calculate_dm_sparse 110 9.5 0.001 0.001 2.969 2.996 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.827 2.988 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.417 2.964 density_rs2pw 110 9.6 0.004 0.004 2.793 2.937 wfi_extrapolate 11 7.9 0.001 0.001 2.924 2.924 mp_alltoall_i22 605 13.7 1.724 2.923 1.724 2.923 cp_fm_cholesky_invert 11 10.9 2.865 2.874 2.865 2.874 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.855 2.858 fft_wrap_pw1pw2_140 451 13.1 0.202 0.214 2.712 2.741 rs_pw_transfer 902 11.9 0.010 0.011 2.496 2.653 calculate_first_density_matrix 1 7.0 0.001 0.004 2.612 2.615 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.496 2.543 fft3d_ps 1111 14.6 1.068 1.104 2.514 2.531 grid_collocate_task_list 110 9.6 2.222 2.387 2.222 2.387 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.274 2.276 jit_kernel_multiply 12 15.8 1.240 2.250 1.240 2.250 potential_pw2rs 110 12.3 0.012 0.013 2.159 2.167 mp_alltoall_d11v 2046 13.8 1.856 2.102 1.856 2.102 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.769 1.798 cp_fm_cholesky_decompose 22 10.9 1.738 1.797 1.738 1.797 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.705 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.615 1.631 acc_transpose_blocks 24660 15.4 0.104 0.106 1.533 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.550 1.565 mp_allgather_i34 2055 14.4 0.515 1.499 0.515 1.499 mp_sum_l 6514 12.8 1.094 1.453 1.094 1.453 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.848 1.406 mp_waitany 13376 13.8 1.118 1.362 1.118 1.362 mp_irecv_dv 62702 16.1 0.751 1.332 0.751 1.332 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.601000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=689.909091, yerr=8.016512 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 827.944960E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.089 0.107 56.621 56.622 qs_mol_dyn_low 1 2.0 0.003 0.003 56.045 56.054 qs_forces 11 3.9 0.035 0.036 55.971 55.973 qs_energies 11 4.9 0.001 0.002 52.301 52.307 scf_env_do_scf 11 5.9 0.000 0.001 43.820 43.820 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.714 35.715 velocity_verlet 10 3.0 0.002 0.003 31.121 31.124 dbcsr_multiply_generic 2055 12.4 0.106 0.107 23.375 23.579 qs_scf_new_mos 99 7.5 0.001 0.001 21.415 21.478 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.414 21.478 ot_scf_mini 99 9.5 0.002 0.002 20.142 20.199 multiply_cannon 2055 13.4 0.247 0.262 17.716 19.202 multiply_cannon_loop 2055 14.4 0.323 0.336 16.322 16.783 rebuild_ks_matrix 110 8.3 0.000 0.000 11.568 11.622 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 11.567 11.621 ot_mini 99 10.5 0.001 0.001 10.782 10.836 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.314 10.361 multiply_cannon_multrec 8220 15.4 3.223 4.741 7.497 8.600 mp_waitall_1 106626 16.5 6.547 8.188 6.547 8.188 init_scf_loop 11 6.9 0.000 0.000 8.059 8.060 qs_ot_get_derivative 99 11.5 0.001 0.001 6.888 6.946 prepare_preconditioner 11 7.9 0.000 0.000 6.422 6.431 make_preconditioner 11 8.9 0.000 0.000 6.422 6.431 sum_up_and_integrate 110 10.3 0.079 0.081 6.245 6.258 integrate_v_rspace 110 11.3 0.003 0.003 6.165 6.178 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.992 6.075 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.699 5.715 calculate_rho_elec 110 8.6 0.115 0.115 5.698 5.714 qs_ot_get_p 110 10.4 0.001 0.001 5.366 5.428 init_scf_run 11 5.9 0.000 0.001 5.351 5.352 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.351 5.351 dbcsr_mm_accdrv_process 17442 15.9 2.921 4.037 4.144 5.101 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.213 4.524 make_m2s 4110 13.4 0.038 0.040 4.297 4.524 make_images 4110 14.4 0.644 0.707 4.167 4.391 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.945 3.949 ot_diis_step 99 11.5 0.012 0.012 3.863 3.864 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.765 3.823 apply_single 110 13.6 0.000 0.000 3.765 3.823 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.596 3.597 grid_integrate_task_list 110 12.3 3.359 3.535 3.359 3.535 pw_transfer 1331 11.6 0.065 0.069 3.348 3.359 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.241 3.253 cp_fm_cholesky_invert 11 10.9 3.180 3.185 3.180 3.185 multiply_cannon_sync_h2d 8220 15.4 2.909 3.065 2.909 3.065 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.057 3.058 cp_fm_redistribute_end 48 14.0 0.777 3.033 0.783 3.034 cp_fm_diag_elpa_base 48 14.0 2.061 2.827 2.241 2.994 density_rs2pw 110 9.6 0.004 0.004 2.738 2.964 make_images_data 4110 15.4 0.038 0.043 2.461 2.854 fft_wrap_pw1pw2_140 451 13.1 0.212 0.217 2.760 2.778 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.373 2.738 wfi_extrapolate 11 7.9 0.001 0.001 2.714 2.714 qs_energies_init_hamiltonians 11 5.9 0.004 0.007 2.662 2.665 calculate_dm_sparse 110 9.5 0.001 0.001 2.520 2.564 calculate_first_density_matrix 1 7.0 0.000 0.000 2.548 2.550 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.536 2.537 fft3d_ps 1111 14.6 1.119 1.146 2.500 2.513 grid_collocate_task_list 110 9.6 2.324 2.498 2.324 2.498 rs_pw_transfer 902 11.9 0.010 0.010 2.217 2.460 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.181 2.197 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.776 2.014 mp_alltoall_d11v 2046 13.8 1.710 1.981 1.710 1.981 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.946 1.972 potential_pw2rs 110 12.3 0.015 0.015 1.967 1.972 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.922 1.954 cp_fm_cholesky_decompose 22 10.9 1.762 1.773 1.762 1.773 mp_allgather_i34 2055 14.4 0.567 1.691 0.567 1.691 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.504 1.618 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.589 1.597 dbcsr_complete_redistribute 325 12.2 0.558 0.584 1.475 1.561 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.453 1.468 mp_waitany 9240 13.8 1.093 1.338 1.093 1.338 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.479 1.329 qs_create_task_list 11 7.9 0.000 0.001 1.227 1.324 generate_qs_task_list 11 8.9 0.380 0.448 1.227 1.324 mp_irecv_dv 24056 15.7 0.455 1.291 0.455 1.291 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.165 1.190 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.905 1.188 mp_alltoall_z22v 1111 16.6 1.148 1.169 1.148 1.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.622000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.272727, yerr=11.193268 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 1.350554E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.081 87.083 87.084 qs_mol_dyn_low 1 2.0 0.003 0.003 86.690 86.699 qs_forces 11 3.9 0.002 0.003 86.619 86.620 qs_energies 11 4.9 0.002 0.003 82.540 82.543 scf_env_do_scf 11 5.9 0.001 0.001 72.639 72.640 velocity_verlet 10 3.0 0.002 0.002 55.331 55.337 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.642 43.644 dbcsr_multiply_generic 2055 12.4 0.120 0.122 29.049 29.097 init_scf_loop 11 6.9 0.000 0.000 28.923 28.926 prepare_preconditioner 11 7.9 0.000 0.000 26.965 26.970 make_preconditioner 11 8.9 0.000 0.000 26.965 26.970 qs_scf_new_mos 99 7.5 0.001 0.001 26.907 26.939 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.906 26.938 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.007 26.434 ot_scf_mini 99 9.5 0.002 0.002 25.132 25.154 multiply_cannon 2055 13.4 0.349 0.364 21.688 22.467 multiply_cannon_loop 2055 14.4 0.343 0.349 19.847 20.214 cp_fm_upper_to_full 70 14.2 13.017 18.698 13.017 18.698 ot_mini 99 10.5 0.001 0.001 13.810 13.825 rebuild_ks_matrix 110 8.3 0.000 0.001 13.407 13.450 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.407 13.450 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.200 12.239 dbcsr_complete_redistribute 325 12.2 1.024 1.049 7.544 10.837 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.513 9.800 multiply_cannon_multrec 8220 15.4 4.381 4.595 9.568 9.645 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.943 9.191 mp_waitall_1 87304 16.6 8.268 9.166 8.268 9.166 qs_ot_get_derivative 99 11.5 0.001 0.001 9.138 9.160 mp_alltoall_i22 605 13.7 5.561 8.866 5.561 8.866 sum_up_and_integrate 110 10.3 0.151 0.152 6.770 6.785 integrate_v_rspace 110 11.3 0.003 0.004 6.619 6.635 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.576 6.617 calculate_rho_elec 110 8.6 0.227 0.227 6.576 6.616 make_m2s 4110 13.4 0.042 0.043 5.566 6.109 qs_ot_get_p 110 10.4 0.001 0.001 6.036 6.057 make_images 4110 14.4 0.874 0.916 5.379 5.920 init_scf_run 11 5.9 0.000 0.001 5.646 5.646 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.645 5.645 cp_fm_cholesky_invert 11 10.9 5.418 5.422 5.418 5.422 dbcsr_mm_accdrv_process 11614 15.7 3.259 3.603 5.044 5.273 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.914 5.235 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.771 5.223 apply_single 110 13.6 0.000 0.000 4.771 5.223 ot_diis_step 99 11.5 0.015 0.015 4.646 4.646 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.399 4.407 hybrid_alltoall_any 4261 16.3 0.257 0.553 3.145 3.963 multiply_cannon_sync_h2d 8220 15.4 3.950 3.955 3.950 3.955 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.922 3.923 pw_transfer 1331 11.6 0.073 0.074 3.890 3.896 make_images_data 4110 15.4 0.041 0.044 3.164 3.861 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.774 3.779 grid_integrate_task_list 110 12.3 3.677 3.744 3.677 3.744 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.659 3.661 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.859 3.311 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.296 3.297 cp_fm_diag_elpa_base 48 14.0 2.737 2.939 3.293 3.294 fft_wrap_pw1pw2_140 451 13.1 0.214 0.216 3.252 3.257 wfi_extrapolate 11 7.9 0.001 0.001 3.225 3.225 calculate_dm_sparse 110 9.5 0.001 0.001 3.150 3.177 density_rs2pw 110 9.6 0.004 0.004 3.039 3.057 fft3d_ps 1111 14.6 1.271 1.277 2.986 2.993 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.887 2.894 grid_collocate_task_list 110 9.6 2.643 2.675 2.643 2.675 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.526 2.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.384 2.399 calculate_first_density_matrix 1 7.0 0.000 0.000 2.314 2.317 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.189 2.254 rs_pw_transfer 902 11.9 0.010 0.011 2.205 2.242 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.207 potential_pw2rs 110 12.3 0.021 0.021 2.154 2.160 mp_alltoall_d11v 2046 13.8 2.109 2.159 2.109 2.159 cp_fm_cholesky_decompose 22 10.9 2.056 2.066 2.056 2.066 qs_create_task_list 11 7.9 0.001 0.001 1.903 1.946 generate_qs_task_list 11 8.9 0.742 0.796 1.902 1.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.810 1.815 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.783 1.798 jit_kernel_multiply 10 15.3 1.585 1.751 1.585 1.751 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.084000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1215.818182, yerr=60.046676 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.773440E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 1919874. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.090 205.359 205.360 qs_mol_dyn_low 1 2.0 0.003 0.004 204.795 204.810 qs_forces 11 3.9 0.003 0.004 204.680 204.681 qs_energies 11 4.9 0.005 0.032 199.176 199.193 scf_env_do_scf 11 5.9 0.001 0.009 182.406 182.410 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.181 161.183 dbcsr_multiply_generic 2507 12.6 0.176 0.179 124.466 125.186 velocity_verlet 10 3.0 0.006 0.009 123.446 123.447 qs_scf_new_mos 117 7.6 0.001 0.001 122.171 122.430 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.170 122.429 ot_scf_mini 117 9.6 0.003 0.005 115.598 115.870 multiply_cannon 2507 13.6 0.239 0.247 100.876 102.872 multiply_cannon_loop 2507 14.6 2.109 2.172 98.653 100.753 ot_mini 117 10.6 0.001 0.001 66.020 66.310 multiply_cannon_multrec 60168 15.6 33.212 35.638 41.577 43.517 qs_ot_get_derivative 117 11.6 0.001 0.001 41.084 41.363 rebuild_ks_matrix 128 8.3 0.001 0.001 33.524 33.884 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 33.523 33.884 mp_waitall_1 291448 16.2 28.917 31.847 28.917 31.847 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.201 30.505 multiply_cannon_sync_h2d 60168 15.6 27.199 28.981 27.199 28.981 qs_ot_get_p 128 10.4 0.001 0.001 27.662 27.938 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.303 25.477 apply_single 128 13.6 0.001 0.001 24.302 25.477 ot_diis_step 117 11.6 0.007 0.008 24.602 24.604 init_scf_loop 11 6.9 0.004 0.026 21.140 21.142 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.888 20.952 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.217 19.407 multiply_cannon_metrocomm3 60168 15.6 0.110 0.114 15.607 19.066 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.329 18.330 prepare_preconditioner 11 7.9 0.000 0.000 16.427 16.478 make_preconditioner 11 8.9 0.000 0.002 16.427 16.477 make_full_inverse_cholesky 11 9.9 0.000 0.001 15.649 15.833 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.485 15.492 cp_fm_redistribute_end 83 14.4 12.195 15.404 12.211 15.408 cp_fm_diag_elpa_base 83 14.4 3.149 15.171 3.184 15.298 make_m2s 5014 13.6 0.103 0.110 14.056 14.475 make_images 5014 14.6 0.405 0.424 13.876 14.304 sum_up_and_integrate 128 10.3 0.091 0.108 14.125 14.165 integrate_v_rspace 128 11.3 0.003 0.004 14.034 14.075 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.752 12.863 calculate_rho_elec 128 8.7 0.045 0.063 12.751 12.862 init_scf_run 11 5.9 0.000 0.001 12.454 12.454 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.453 12.454 cp_fm_cholesky_invert 11 10.9 9.498 9.505 9.498 9.505 mp_sum_l 7870 13.0 8.286 9.370 8.286 9.370 wfi_extrapolate 11 7.9 0.001 0.001 9.116 9.116 calculate_dm_sparse 128 9.5 0.001 0.001 8.382 8.516 dbcsr_mm_accdrv_process 124484 16.2 3.310 3.465 7.926 8.480 multiply_cannon_metrocomm1 60168 15.6 0.085 0.089 6.223 8.289 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.975 8.113 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.820 7.912 make_images_data 5014 15.6 0.069 0.076 6.927 7.822 grid_integrate_task_list 128 12.3 7.089 7.517 7.089 7.517 hybrid_alltoall_any 5200 16.5 0.291 2.267 6.058 7.276 pw_transfer 1547 11.6 0.074 0.107 6.959 7.260 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 6.755 7.028 density_rs2pw 128 9.7 0.006 0.007 6.493 6.957 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.753 6.759 rs_pw_transfer 1046 11.9 0.017 0.018 5.733 6.222 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.866 5.990 fft_wrap_pw1pw2_140 523 13.2 0.444 0.505 5.779 5.982 fft3d_ps 1291 14.7 2.099 2.576 5.547 5.763 mp_alltoall_d11v 2415 14.1 4.386 5.420 4.386 5.420 grid_collocate_task_list 128 9.7 4.729 5.051 4.729 5.051 cp_fm_cholesky_decompose 22 10.9 4.723 4.738 4.723 4.738 potential_pw2rs 128 12.3 0.009 0.010 4.547 4.604 mp_sum_d 4459 12.1 3.748 4.490 3.748 4.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.360000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.545455, yerr=6.372721 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.664128E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 5620363. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.110 0.312 190.053 190.056 qs_mol_dyn_low 1 2.0 0.003 0.004 189.288 189.300 qs_forces 11 3.9 0.007 0.024 189.192 189.196 qs_energies 11 4.9 0.005 0.012 182.464 182.477 scf_env_do_scf 11 5.9 0.001 0.004 165.924 165.934 scf_env_do_scf_inner_loop 117 6.6 0.007 0.020 133.008 133.013 velocity_verlet 10 3.0 0.005 0.008 118.836 118.837 dbcsr_multiply_generic 2507 12.6 0.189 0.202 97.110 98.509 qs_scf_new_mos 117 7.6 0.001 0.001 94.760 95.138 qs_scf_loop_do_ot 117 8.6 0.001 0.003 94.759 95.137 ot_scf_mini 117 9.6 0.004 0.006 89.954 90.418 multiply_cannon 2507 13.6 0.472 0.527 76.997 81.324 multiply_cannon_loop 2507 14.6 1.256 1.298 74.059 76.833 ot_mini 117 10.6 0.001 0.001 49.801 50.242 mp_waitall_1 226760 16.4 25.096 38.574 25.096 38.574 multiply_cannon_multrec 30084 15.6 21.798 26.407 31.454 36.229 rebuild_ks_matrix 128 8.3 0.001 0.001 32.483 33.025 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.023 32.482 33.024 init_scf_loop 11 6.9 0.001 0.005 32.827 32.841 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.217 29.718 prepare_preconditioner 11 7.9 0.000 0.000 28.421 28.516 make_preconditioner 11 8.9 0.000 0.003 28.421 28.516 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.981 28.326 qs_ot_get_derivative 117 11.6 0.001 0.002 27.812 28.275 make_full_inverse_cholesky 11 9.9 0.000 0.001 27.105 27.634 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.037 23.296 apply_single 128 13.6 0.001 0.001 22.036 23.296 qs_ot_get_p 128 10.4 0.001 0.002 22.139 22.891 multiply_cannon_sync_h2d 30084 15.6 19.042 21.895 19.042 21.895 ot_diis_step 117 11.6 0.014 0.015 21.820 21.822 qs_ot_p2m_diag 83 11.4 0.187 0.217 17.299 17.338 cp_fm_cholesky_invert 11 10.9 16.636 16.648 16.636 16.648 cp_dbcsr_syevd 83 12.4 0.005 0.005 16.152 16.153 make_m2s 5014 13.6 0.089 0.094 14.266 15.499 make_images 5014 14.6 1.154 1.343 14.057 15.292 sum_up_and_integrate 128 10.3 0.116 0.131 14.279 14.315 integrate_v_rspace 128 11.3 0.003 0.004 14.162 14.203 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.033 13.043 cp_fm_redistribute_end 83 14.4 7.581 12.976 7.594 12.979 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.855 12.889 calculate_rho_elec 128 8.7 0.088 0.105 12.855 12.888 cp_fm_diag_elpa_base 83 14.4 5.145 12.511 5.365 12.884 init_scf_run 11 5.9 0.000 0.002 11.649 11.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.648 11.650 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.913 11.258 multiply_cannon_metrocomm4 27577 15.6 0.097 0.113 3.856 10.779 mp_irecv_dv 69486 16.3 3.660 10.389 3.660 10.389 make_images_data 5014 15.6 0.067 0.075 8.720 10.188 dbcsr_mm_accdrv_process 62242 16.2 4.644 5.329 9.122 9.750 hybrid_alltoall_any 5200 16.5 0.343 1.526 7.420 9.316 wfi_extrapolate 11 7.9 0.001 0.001 8.268 8.268 pw_transfer 1547 11.6 0.084 0.100 7.762 7.822 grid_integrate_task_list 128 12.3 7.167 7.624 7.167 7.624 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.539 7.598 density_rs2pw 128 9.7 0.006 0.007 6.655 7.241 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.273 7.034 cp_fm_cholesky_decompose 22 10.9 6.843 6.926 6.843 6.926 calculate_dm_sparse 128 9.5 0.001 0.001 6.534 6.692 fft_wrap_pw1pw2_140 523 13.2 0.468 0.518 6.619 6.686 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.212 6.220 mp_sum_l 7870 13.0 4.089 6.143 4.089 6.143 rs_pw_transfer 1046 11.9 0.014 0.016 5.515 6.040 fft3d_ps 1291 14.7 2.775 2.927 5.911 5.975 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 5.324 5.481 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.343 5.407 grid_collocate_task_list 128 9.7 4.909 5.316 4.909 5.316 mp_alltoall_d11v 2415 14.1 4.119 5.124 4.119 5.124 potential_pw2rs 128 12.3 0.015 0.017 4.740 4.767 mp_allgather_i34 2507 14.6 1.514 4.614 1.514 4.614 dbcsr_complete_redistribute 395 12.7 0.784 0.889 3.201 4.086 mp_sum_d 4464 12.1 2.673 4.056 2.673 4.056 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=190.056000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.090909, yerr=3.058655 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 938.655744E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+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 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4077 57304. MP_Allreduce 11231 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 175.566 175.567 qs_mol_dyn_low 1 2.0 0.003 0.003 175.181 175.194 qs_forces 11 3.9 0.003 0.003 175.080 175.089 qs_energies 11 4.9 0.002 0.002 168.599 168.612 scf_env_do_scf 11 5.9 0.001 0.001 152.874 152.875 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 117.465 117.465 velocity_verlet 10 3.0 0.004 0.009 111.295 111.297 dbcsr_multiply_generic 2485 12.5 0.180 0.184 81.094 82.130 qs_scf_new_mos 116 7.6 0.001 0.001 81.408 81.795 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.407 81.794 ot_scf_mini 116 9.6 0.003 0.004 77.331 77.777 multiply_cannon 2485 13.5 0.494 0.517 61.370 65.990 multiply_cannon_loop 2485 14.5 0.852 0.879 58.339 60.791 ot_mini 116 10.6 0.001 0.001 42.259 42.704 init_scf_loop 11 6.9 0.000 0.000 35.303 35.304 mp_waitall_1 176908 16.5 24.934 34.146 24.934 34.146 prepare_preconditioner 11 7.9 0.000 0.000 31.277 31.324 make_preconditioner 11 8.9 0.000 0.000 31.277 31.324 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.014 30.317 rebuild_ks_matrix 127 8.3 0.001 0.001 29.721 30.179 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.721 30.178 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.821 27.244 multiply_cannon_multrec 19880 15.5 13.445 16.839 22.333 25.747 multiply_cannon_metrocomm3 19880 15.5 0.058 0.061 14.946 24.837 qs_ot_get_derivative 116 11.6 0.001 0.002 22.757 23.207 qs_ot_get_p 127 10.4 0.001 0.001 20.645 21.185 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.639 20.655 apply_single 127 13.6 0.001 0.001 19.638 20.655 ot_diis_step 116 11.6 0.018 0.018 19.384 19.385 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.398 16.408 multiply_cannon_sync_h2d 19880 15.5 14.102 15.708 14.102 15.708 make_m2s 4970 13.5 0.078 0.086 14.506 15.420 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.365 15.366 make_images 4970 14.5 1.164 1.281 14.277 15.191 cp_fm_cholesky_invert 11 10.9 14.736 14.745 14.736 14.745 sum_up_and_integrate 127 10.3 0.133 0.145 14.030 14.055 integrate_v_rspace 127 11.3 0.004 0.004 13.897 13.926 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.928 12.961 calculate_rho_elec 127 8.7 0.131 0.145 12.927 12.961 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.221 12.222 cp_fm_redistribute_end 82 14.4 4.587 12.156 4.601 12.159 cp_fm_diag_elpa_base 82 14.4 7.133 11.596 7.537 12.073 init_scf_run 11 5.9 0.000 0.001 10.748 10.749 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.748 10.749 make_images_data 4970 15.5 0.061 0.068 8.904 10.228 hybrid_alltoall_any 5155 16.4 0.427 1.954 7.650 9.361 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.809 9.124 multiply_cannon_metrocomm4 17395 15.5 0.063 0.072 3.393 9.070 mp_irecv_dv 49801 16.2 3.270 8.823 3.270 8.823 dbcsr_mm_accdrv_process 41158 16.2 4.703 5.505 8.344 8.507 pw_transfer 1535 11.6 0.084 0.101 7.651 7.756 cp_fm_cholesky_decompose 22 10.9 7.610 7.647 7.610 7.647 grid_integrate_task_list 127 12.3 7.223 7.597 7.223 7.597 fft_wrap_pw1pw2 1281 12.7 0.010 0.010 7.430 7.541 wfi_extrapolate 11 7.9 0.001 0.001 7.315 7.316 cp_fm_upper_to_full 104 14.5 5.618 7.278 5.618 7.278 density_rs2pw 127 9.7 0.006 0.006 6.468 6.855 fft_wrap_pw1pw2_140 519 13.2 0.474 0.525 6.501 6.614 dbcsr_complete_redistribute 393 12.7 1.171 1.198 4.550 6.257 calculate_dm_sparse 127 9.5 0.001 0.001 5.794 5.885 fft3d_ps 1281 14.7 2.676 2.884 5.709 5.784 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.616 5.624 grid_collocate_task_list 127 9.7 5.036 5.548 5.036 5.548 rs_pw_transfer 1038 11.9 0.013 0.014 5.075 5.522 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.597 5.251 mp_sum_l 7804 13.0 3.408 5.085 3.408 5.085 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.332 5.034 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.694 4.820 mp_alltoall_d11v 2401 14.1 4.282 4.775 4.282 4.775 mp_allgather_i34 2485 14.5 1.569 4.615 1.569 4.615 potential_pw2rs 127 12.3 0.020 0.022 4.458 4.474 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.049 4.084 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.243 3.888 mp_alltoall_i22 712 14.1 1.870 3.723 1.870 3.723 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.712 3.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.567000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=889.363636, yerr=12.694113 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/18/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 4.353788E+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 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 1.128092E+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 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57332. MP_Allreduce 11226 1068. MP_Sync 170 MP_Alltoall 1712 12503084. MP_ISendRecv 11776 75008. MP_Wait 28330 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.061 0.086 222.978 222.979 qs_mol_dyn_low 1 2.0 0.003 0.004 222.430 222.441 qs_forces 11 3.9 0.010 0.065 222.316 222.323 qs_energies 11 4.9 0.028 0.051 214.757 214.770 scf_env_do_scf 11 5.9 0.008 0.023 196.547 196.560 velocity_verlet 10 3.0 0.010 0.011 146.181 146.185 scf_env_do_scf_inner_loop 117 6.6 0.020 0.044 143.488 143.506 qs_scf_new_mos 117 7.6 0.001 0.001 101.943 102.429 qs_scf_loop_do_ot 117 8.6 0.005 0.012 101.942 102.428 ot_scf_mini 117 9.6 0.011 0.019 96.915 97.401 dbcsr_multiply_generic 2507 12.6 0.191 0.196 90.759 92.660 multiply_cannon 2507 13.6 0.546 0.577 59.631 63.677 multiply_cannon_loop 2507 14.6 1.181 1.214 55.067 59.105 init_scf_loop 11 6.9 0.002 0.007 52.875 52.917 ot_mini 117 10.6 0.001 0.001 50.266 50.790 prepare_preconditioner 11 7.9 0.000 0.000 48.146 48.182 make_preconditioner 11 8.9 0.000 0.001 48.146 48.182 make_full_inverse_cholesky 11 9.9 0.000 0.000 41.200 46.415 mp_waitall_1 153770 16.5 26.967 37.950 26.967 37.950 rebuild_ks_matrix 128 8.3 0.001 0.001 32.990 33.692 qs_ks_build_kohn_sham_matrix 128 9.3 0.018 0.030 32.990 33.691 qs_ot_get_p 128 10.4 0.004 0.024 31.277 31.803 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.837 30.477 multiply_cannon_multrec 30084 15.6 13.305 17.895 25.299 30.116 qs_ot_get_derivative 117 11.6 0.001 0.002 28.156 28.642 make_m2s 5014 13.6 0.093 0.097 24.638 25.717 qs_ot_p2m_diag 83 11.4 0.343 0.390 25.600 25.663 make_images 5014 14.6 1.973 2.256 24.333 25.410 cp_dbcsr_syevd 83 12.4 0.005 0.005 24.019 24.020 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 20.689 22.139 apply_single 128 13.6 0.001 0.001 20.689 22.139 ot_diis_step 117 11.6 0.018 0.019 21.910 21.911 multiply_cannon_metrocomm3 30084 15.6 0.047 0.049 10.238 21.675 cp_fm_diag_elpa 83 13.4 0.000 0.001 20.288 20.293 cp_fm_redistribute_end 83 14.4 3.492 20.091 3.528 20.097 cp_fm_diag_elpa_base 83 14.4 15.855 19.296 16.512 19.827 cp_fm_cholesky_invert 11 10.9 19.485 19.498 19.485 19.498 cp_fm_upper_to_full 105 14.7 11.341 16.906 11.341 16.906 make_images_data 5014 15.6 0.063 0.068 14.038 16.610 sum_up_and_integrate 128 10.3 0.142 0.154 16.394 16.419 integrate_v_rspace 128 11.3 0.004 0.004 16.252 16.281 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.852 15.958 calculate_rho_elec 128 8.7 0.175 0.190 15.851 15.957 hybrid_alltoall_any 5200 16.5 0.526 2.199 12.101 14.807 dbcsr_complete_redistribute 395 12.7 1.526 1.645 10.540 14.410 copy_fm_to_dbcsr 209 11.7 0.003 0.011 8.566 12.420 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 12.096 12.406 multiply_cannon_sync_h2d 30084 15.6 11.606 12.266 11.606 12.266 dbcsr_mm_accdrv_process 62264 16.2 7.472 8.523 11.587 12.096 init_scf_run 11 5.9 0.002 0.006 11.945 11.947 scf_env_initial_rho_setup 11 6.9 0.003 0.006 11.944 11.947 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.923 10.700 pw_transfer 1547 11.6 0.086 0.100 10.345 10.429 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 10.120 10.207 mp_alltoall_i22 716 14.1 6.069 9.913 6.069 9.913 cp_fm_cholesky_decompose 22 10.9 9.057 9.211 9.057 9.211 density_rs2pw 128 9.7 0.006 0.006 8.414 8.680 wfi_extrapolate 11 7.9 0.001 0.001 8.666 8.666 fft_wrap_pw1pw2_140 523 13.2 0.476 0.485 8.519 8.610 fft3d_ps 1291 14.7 2.791 2.888 8.360 8.443 mp_alltoall_d11v 2415 14.1 7.220 7.851 7.220 7.851 grid_integrate_task_list 128 12.3 7.490 7.771 7.490 7.771 multiply_cannon_metrocomm4 25070 15.6 0.076 0.086 2.883 7.436 mp_irecv_dv 76098 16.2 2.744 7.180 2.744 7.180 calculate_dm_sparse 128 9.5 0.001 0.001 6.718 6.825 rs_pw_transfer 1046 11.9 0.013 0.014 6.515 6.801 mp_sum_l 7870 13.0 4.579 6.230 4.579 6.230 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.173 6.218 potential_pw2rs 128 12.3 0.023 0.023 6.043 6.077 grid_collocate_task_list 128 9.7 5.205 5.520 5.205 5.520 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.217 5.393 mp_alltoall_z22v 1291 16.7 4.987 5.066 4.987 5.066 dbcsr_dot_sd 1318 12.0 1.345 1.455 3.533 4.912 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.743 4.884 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.752 4.825 mp_allgather_i34 2507 14.6 2.678 4.801 2.678 4.801 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 4.593 4.594 mp_sum_d 4457 12.1 3.191 4.585 3.191 4.585 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=222.979000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1066.818182, yerr=17.235366 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.522393E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+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 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57658. MP_Allreduce 11099 1171. MP_Sync 88 MP_Alltoall 1724 18848021. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.078 191.421 191.421 qs_mol_dyn_low 1 2.0 0.003 0.003 190.886 190.898 qs_forces 11 3.9 0.003 0.003 190.780 190.784 qs_energies 11 4.9 0.011 0.023 183.159 183.168 scf_env_do_scf 11 5.9 0.002 0.017 164.930 164.940 velocity_verlet 10 3.0 0.011 0.012 126.767 126.772 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 124.445 124.446 qs_scf_new_mos 118 7.6 0.001 0.001 85.903 86.047 qs_scf_loop_do_ot 118 8.6 0.001 0.001 85.903 86.046 ot_scf_mini 118 9.6 0.004 0.004 81.236 81.297 dbcsr_multiply_generic 2535 12.6 0.184 0.188 78.660 79.270 multiply_cannon 2535 13.6 0.587 0.616 56.356 60.925 multiply_cannon_loop 2535 14.6 0.454 0.468 51.302 52.653 ot_mini 118 10.6 0.001 0.001 42.655 42.726 init_scf_loop 11 6.9 0.000 0.000 40.305 40.307 prepare_preconditioner 11 7.9 0.000 0.000 36.100 36.133 make_preconditioner 11 8.9 0.000 0.000 36.100 36.133 mp_waitall_1 131034 16.6 28.972 35.941 28.972 35.941 make_full_inverse_cholesky 11 9.9 0.000 0.000 33.652 33.963 rebuild_ks_matrix 129 8.3 0.001 0.001 30.475 30.646 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.474 30.645 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.816 27.966 qs_ot_get_p 129 10.4 0.001 0.001 25.145 25.217 qs_ot_get_derivative 118 11.6 0.002 0.002 22.295 22.350 multiply_cannon_multrec 10140 15.6 10.510 14.535 18.175 21.288 cp_fm_cholesky_invert 11 10.9 20.993 21.002 20.993 21.002 qs_ot_p2m_diag 84 11.4 0.501 0.506 20.592 20.608 multiply_cannon_metrocomm3 10140 15.6 0.022 0.023 12.700 20.603 ot_diis_step 118 11.6 0.020 0.021 20.252 20.252 make_m2s 5070 13.6 0.066 0.070 18.003 20.180 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.713 20.086 apply_single 129 13.6 0.001 0.001 19.713 20.086 make_images 5070 14.6 2.334 2.903 17.697 19.877 cp_dbcsr_syevd 84 12.4 0.005 0.005 19.292 19.295 cp_fm_diag_elpa 84 13.4 0.000 0.000 15.809 15.810 cp_fm_diag_elpa_base 84 14.4 15.528 15.628 15.798 15.798 sum_up_and_integrate 129 10.3 0.182 0.192 15.331 15.380 integrate_v_rspace 129 11.3 0.004 0.004 15.148 15.207 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.667 14.719 calculate_rho_elec 129 8.7 0.260 0.271 14.667 14.719 make_images_data 5070 15.6 0.053 0.061 11.096 13.563 hybrid_alltoall_any 5257 16.5 0.836 3.774 10.664 13.285 multiply_cannon_sync_h2d 10140 15.6 11.629 12.119 11.629 12.119 init_scf_run 11 5.9 0.000 0.001 11.274 11.274 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.273 11.274 pw_transfer 1559 11.6 0.084 0.091 9.160 9.189 cp_fm_cholesky_decompose 22 10.9 9.072 9.182 9.072 9.182 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.067 9.116 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 8.937 8.962 multiply_cannon_metrocomm1 10140 15.6 0.029 0.030 5.087 8.305 grid_integrate_task_list 129 12.3 7.777 8.168 7.777 8.168 mp_allgather_i34 2535 14.6 3.246 8.084 3.246 8.084 wfi_extrapolate 11 7.9 0.001 0.001 8.059 8.059 dbcsr_mm_accdrv_process 20958 16.1 2.735 3.637 7.294 7.954 fft_wrap_pw1pw2_140 527 13.2 0.506 0.528 7.691 7.724 density_rs2pw 129 9.7 0.006 0.006 7.069 7.387 fft3d_ps 1301 14.7 2.751 2.819 7.083 7.118 mp_alltoall_d11v 2429 14.1 6.339 6.960 6.339 6.960 calculate_dm_sparse 129 9.5 0.001 0.001 6.464 6.564 dbcsr_complete_redistribute 397 12.7 2.114 2.264 6.050 6.465 grid_collocate_task_list 129 9.7 5.532 5.821 5.532 5.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.610 5.616 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.280 5.281 rs_pw_transfer 1054 12.0 0.012 0.013 4.927 5.254 potential_pw2rs 129 12.3 0.026 0.027 4.890 4.914 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.293 4.333 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.944 4.287 copy_dbcsr_to_fm 187 11.8 0.004 0.004 4.019 4.115 multiply_cannon_metrocomm4 7605 15.6 0.024 0.026 1.877 4.045 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.996 4.013 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.955 3.992 mp_irecv_dv 29150 15.9 1.841 3.977 1.841 3.977 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.570 3.892 mp_alltoall_z22v 1301 16.7 3.709 3.838 3.709 3.838 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=191.421000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1415.363636, yerr=54.722893 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1971389988864 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 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 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 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.241596E+12 0.0% 0.0% 100.0% flops max/rank 11.787141E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806515136 0.0% 0.0% 100.0% number of processed stacks 1981760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.6 marketing flops 145.659521E+12 ------------------------------------------------------------------------------- # multiplications 2533 max memory usage/rank 2.942763E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101320 MPI messages size (bytes): total size 1.145104E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.301851E+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 45808 35567697920 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4069 58204. MP_Allreduce 11170 1510. MP_Sync 86 MP_Alltoall 1724 36993574. MP_ISendRecv 3612 218624. MP_Wait 11682 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.065 0.093 292.432 292.433 qs_mol_dyn_low 1 2.0 0.003 0.003 291.750 291.762 qs_forces 11 3.9 0.003 0.003 291.663 291.667 qs_energies 11 4.9 0.001 0.002 283.125 283.136 scf_env_do_scf 11 5.9 0.007 0.054 261.174 261.184 velocity_verlet 10 3.0 0.009 0.010 210.831 210.838 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 135.284 135.286 init_scf_loop 11 6.9 0.000 0.000 125.574 125.578 prepare_preconditioner 11 7.9 0.000 0.000 120.908 120.927 make_preconditioner 11 8.9 0.000 0.000 120.908 120.927 make_full_inverse_cholesky 11 9.9 0.000 0.000 96.978 118.066 qs_scf_new_mos 118 7.6 0.001 0.001 93.159 93.345 qs_scf_loop_do_ot 118 8.6 0.001 0.001 93.158 93.344 ot_scf_mini 118 9.6 0.004 0.004 88.346 88.436 dbcsr_multiply_generic 2533 12.6 0.216 0.222 84.035 84.791 cp_fm_upper_to_full 104 14.8 52.931 75.799 52.931 75.799 multiply_cannon 2533 13.6 0.703 0.759 59.247 60.060 multiply_cannon_loop 2533 14.6 0.478 0.489 55.570 57.382 ot_mini 118 10.6 0.001 0.001 45.291 45.396 dbcsr_complete_redistribute 393 12.7 3.997 4.047 29.745 42.378 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.264 38.885 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.885 36.372 cp_fm_cholesky_invert 11 10.9 34.595 34.601 34.595 34.601 mp_alltoall_i22 716 14.1 21.757 34.549 21.757 34.549 rebuild_ks_matrix 129 8.3 0.001 0.001 33.506 33.607 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.505 33.607 mp_waitall_1 106546 16.7 28.636 32.308 28.636 32.308 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.326 31.423 qs_ot_get_p 129 10.4 0.001 0.001 27.373 27.472 qs_ot_get_derivative 118 11.6 0.002 0.002 24.707 24.807 qs_ot_p2m_diag 82 11.4 0.868 0.873 23.136 23.167 cp_dbcsr_syevd 82 12.4 0.005 0.006 21.381 21.383 make_m2s 5066 13.6 0.076 0.079 19.855 20.819 multiply_cannon_metrocomm3 10132 15.6 0.023 0.023 19.120 20.728 ot_diis_step 118 11.6 0.022 0.023 20.536 20.536 make_images 5066 14.6 3.782 3.989 19.373 20.339 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.604 19.949 apply_single 129 13.6 0.001 0.001 19.604 19.949 multiply_cannon_multrec 10132 15.6 10.596 12.388 18.114 18.183 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.139 18.140 cp_fm_diag_elpa_base 82 14.4 13.824 15.429 18.135 18.135 multiply_cannon_sync_h2d 10132 15.6 15.783 15.815 15.783 15.815 sum_up_and_integrate 129 10.3 0.322 0.325 15.720 15.808 integrate_v_rspace 129 11.3 0.004 0.004 15.397 15.486 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.969 14.988 calculate_rho_elec 129 8.7 0.487 0.488 14.969 14.988 hybrid_alltoall_any 5251 16.5 1.309 3.044 10.624 12.272 make_images_data 5066 15.6 0.060 0.065 10.550 12.201 init_scf_run 11 5.9 0.000 0.001 11.898 11.899 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.898 11.898 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.457 9.526 cp_fm_cholesky_decompose 22 10.9 9.179 9.244 9.179 9.244 dbcsr_mm_accdrv_process 20950 16.1 3.813 5.550 7.279 9.144 grid_integrate_task_list 129 12.3 8.604 8.786 8.604 8.786 wfi_extrapolate 11 7.9 0.001 0.001 8.757 8.757 pw_transfer 1559 11.6 0.090 0.091 8.358 8.371 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.124 8.135 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.003 8.005 fft_wrap_pw1pw2_140 527 13.2 0.545 0.549 7.165 7.186 mp_alltoall_d11v 2417 14.1 6.846 6.975 6.846 6.975 calculate_dm_sparse 129 9.5 0.001 0.001 6.671 6.788 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.365 6.437 grid_collocate_task_list 129 9.7 6.387 6.424 6.387 6.424 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.189 6.258 fft3d_ps 1301 14.7 2.777 2.797 6.152 6.158 density_rs2pw 129 9.7 0.005 0.005 6.022 6.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=292.433000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2596.272727, yerr=177.072194 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 1.261408E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.082 0.136 85.011 85.027 qs_energies 1 2.0 0.001 0.008 84.362 84.369 ls_scf 1 3.0 0.000 0.001 83.440 83.450 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.555 72.779 multiply_cannon 111 7.7 0.017 0.020 55.890 57.334 multiply_cannon_loop 111 8.7 0.209 0.223 52.452 54.061 ls_scf_main 1 4.0 0.000 0.003 52.207 52.208 density_matrix_trs4 2 5.0 0.003 0.029 46.805 46.879 ls_scf_init_scf 1 4.0 0.000 0.002 28.189 28.190 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.113 27.171 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.015 25.038 mp_waitall_1 11316 10.9 22.168 24.411 22.168 24.411 multiply_cannon_multrec 2664 9.7 8.217 8.961 15.526 17.317 multiply_cannon_sync_h2d 2664 9.7 13.837 15.829 13.837 15.829 make_m2s 222 7.7 0.008 0.011 13.046 13.614 make_images 222 8.7 0.100 0.109 13.024 13.594 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.528 12.341 make_images_data 222 9.7 0.004 0.005 7.606 8.156 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.634 6.928 7.925 hybrid_alltoall_any 227 10.6 0.214 1.837 6.512 7.788 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.356 7.464 dbcsr_mm_accdrv_process_sort 4760 11.4 6.221 7.148 6.221 7.148 calculate_norms 4752 9.8 5.545 6.239 5.545 6.239 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.977 5.173 mp_sum_l 807 5.4 3.105 4.807 3.105 4.807 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.291 3.518 make_images_sizes 222 9.7 0.000 0.000 0.692 3.469 mp_alltoall_i44 222 10.7 0.691 3.468 0.691 3.468 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.062 3.321 mp_irecv_dv 6231 10.9 2.044 3.301 2.044 3.301 arnoldi_extremal 4 6.8 0.000 0.000 3.255 3.287 arnoldi_normal_ev 4 7.8 0.001 0.003 3.255 3.287 build_subspace 16 8.4 0.009 0.012 3.133 3.135 ls_scf_post 1 4.0 0.000 0.009 3.044 3.052 ls_scf_store_result 1 5.0 0.000 0.000 2.857 2.896 dbcsr_special_finalize 555 9.7 0.005 0.006 2.391 2.792 dbcsr_merge_single_wm 555 10.7 0.467 0.621 2.383 2.784 make_images_pack 222 9.7 2.210 2.633 2.212 2.634 dbcsr_matrix_vector_mult 304 9.0 0.003 0.010 2.321 2.567 dbcsr_sort_data 658 11.4 2.175 2.495 2.175 2.495 dbcsr_matrix_vector_mult_local 304 10.0 2.059 2.451 2.061 2.453 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.215 2.351 buffer_matrices_ensure_size 222 8.7 1.778 2.171 1.778 2.171 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.714 1.715 rebuild_ks_matrix 3 7.3 0.000 0.000 1.705 1.706 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.705 1.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.027000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 2.110763E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.055 0.083 91.476 91.478 qs_energies 1 2.0 0.000 0.000 90.545 90.558 ls_scf 1 3.0 0.004 0.017 89.236 89.249 dbcsr_multiply_generic 111 6.7 0.016 0.021 75.163 75.545 multiply_cannon 111 7.7 0.028 0.041 53.151 57.354 ls_scf_main 1 4.0 0.000 0.000 54.877 54.881 multiply_cannon_loop 111 8.7 0.116 0.123 49.888 53.551 density_matrix_trs4 2 5.0 0.002 0.003 49.239 49.445 ls_scf_init_scf 1 4.0 0.000 0.001 30.834 30.835 mp_waitall_1 9246 10.9 20.894 29.971 20.894 29.971 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.496 29.637 multiply_cannon_multrec 1332 9.7 13.116 17.059 22.316 27.346 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.023 27.090 27.104 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.733 22.073 make_m2s 222 7.7 0.007 0.009 15.151 15.873 make_images 222 8.7 1.572 1.928 15.121 15.845 dbcsr_mm_accdrv_process 4041 10.4 0.272 0.449 8.806 10.249 dbcsr_mm_accdrv_process_sort 4041 11.4 8.395 9.824 8.395 9.824 make_images_data 222 9.7 0.008 0.037 8.570 9.634 hybrid_alltoall_any 227 10.6 0.525 2.468 7.940 9.079 mp_sum_l 807 5.4 5.562 8.973 5.562 8.973 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.183 7.780 mp_irecv_dv 3311 11.0 3.163 7.730 3.163 7.730 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.346 7.179 calculate_norms 2376 9.8 5.991 6.679 5.991 6.679 multiply_cannon_sync_h2d 1332 9.7 4.866 6.322 4.866 6.322 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.209 arnoldi_extremal 4 6.8 0.000 0.000 4.684 4.713 arnoldi_normal_ev 4 7.8 0.025 0.035 4.684 4.713 build_subspace 16 8.4 0.014 0.021 4.394 4.397 ls_scf_post 1 4.0 0.001 0.005 3.522 3.534 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.135 3.370 ls_scf_store_result 1 5.0 0.000 0.000 3.211 3.340 dbcsr_matrix_vector_mult_local 304 10.0 2.743 3.236 2.745 3.238 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.224 2.702 ls_scf_dm_to_ks 2 5.0 0.000 0.011 2.522 2.607 make_images_pack 222 9.7 2.030 2.391 2.032 2.393 mp_allgather_i34 111 8.7 0.977 2.369 0.977 2.369 dbcsr_sort_data 436 11.2 1.833 2.075 1.833 2.075 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.948 1.951 rebuild_ks_matrix 3 7.3 0.000 0.000 1.936 1.938 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.011 1.936 1.938 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.478000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 2.755547E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.211 0.343 94.111 94.131 qs_energies 1 2.0 0.000 0.000 93.022 93.079 ls_scf 1 3.0 0.000 0.001 91.515 91.572 dbcsr_multiply_generic 111 6.7 0.016 0.023 76.210 76.484 ls_scf_main 1 4.0 0.000 0.000 57.164 57.169 multiply_cannon 111 7.7 0.044 0.120 52.765 56.426 multiply_cannon_loop 111 8.7 0.100 0.105 49.154 53.309 density_matrix_trs4 2 5.0 0.002 0.003 51.196 51.427 mp_waitall_1 7374 11.0 23.657 34.141 23.657 34.141 ls_scf_init_scf 1 4.0 0.000 0.002 30.738 30.740 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.563 29.626 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.090 27.100 multiply_cannon_multrec 888 9.7 12.550 15.289 21.068 24.405 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.030 23.576 make_m2s 222 7.7 0.006 0.007 16.734 18.010 make_images 222 8.7 1.971 2.282 16.696 17.973 make_images_data 222 9.7 0.004 0.004 9.572 10.645 hybrid_alltoall_any 227 10.6 0.619 2.850 9.262 10.532 dbcsr_mm_accdrv_process 3754 10.4 0.245 0.413 8.036 9.198 dbcsr_mm_accdrv_process_sort 3754 11.4 7.667 8.785 7.667 8.785 mp_sum_l 807 5.4 5.162 8.454 5.162 8.454 multiply_cannon_sync_h2d 888 9.7 6.057 7.766 6.057 7.766 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.456 7.158 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.673 7.129 mp_irecv_dv 2335 11.1 2.441 7.109 2.441 7.109 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.895 6.470 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.988 5.203 arnoldi_extremal 4 6.8 0.000 0.000 5.085 5.101 arnoldi_normal_ev 4 7.8 0.001 0.005 5.085 5.101 build_subspace 16 8.4 0.014 0.020 4.783 4.791 calculate_norms 1584 9.8 4.249 4.548 4.249 4.548 mp_allgather_i34 111 8.7 1.452 3.829 1.452 3.829 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.463 3.777 ls_scf_post 1 4.0 0.005 0.043 3.613 3.671 dbcsr_matrix_vector_mult_local 304 10.0 3.024 3.602 3.026 3.604 ls_scf_store_result 1 5.0 0.000 0.000 3.318 3.415 ls_scf_dm_to_ks 2 5.0 0.000 0.009 2.776 2.898 dbcsr_sort_data 325 11.1 1.918 2.171 1.918 2.171 make_images_pack 222 9.7 1.818 2.135 1.820 2.137 dbcsr_data_release 9322 10.9 1.322 1.989 1.322 1.989 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.906 1.908 dbcsr_finalize 304 7.8 0.026 0.032 1.622 1.903 rebuild_ks_matrix 3 7.3 0.000 0.000 1.888 1.890 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.049 1.888 1.890 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.131000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2166.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 3.350209E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.095 0.329 98.145 98.150 qs_energies 1 2.0 0.007 0.056 96.933 96.984 ls_scf 1 3.0 0.026 0.043 95.288 95.298 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.619 78.869 ls_scf_main 1 4.0 0.000 0.002 59.205 59.232 multiply_cannon 111 7.7 0.061 0.128 51.674 55.821 density_matrix_trs4 2 5.0 0.003 0.023 53.100 53.210 multiply_cannon_loop 111 8.7 0.115 0.125 46.637 49.263 ls_scf_init_scf 1 4.0 0.000 0.001 32.767 32.769 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.320 31.392 mp_waitall_1 6438 11.0 23.161 29.974 23.161 29.974 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.841 28.860 multiply_cannon_multrec 1332 9.7 14.139 17.317 21.973 24.654 make_m2s 222 7.7 0.007 0.008 21.368 22.786 make_images 222 8.7 3.139 3.601 21.318 22.738 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.514 17.296 make_images_data 222 9.7 0.004 0.005 12.024 13.640 hybrid_alltoall_any 227 10.6 0.796 3.820 11.353 13.297 dbcsr_mm_accdrv_process 3641 10.4 0.194 0.402 7.466 9.021 dbcsr_mm_accdrv_process_sort 3641 11.4 7.095 8.596 7.095 8.596 mp_sum_l 807 5.4 4.036 7.221 4.036 7.221 multiply_cannon_sync_h2d 1332 9.7 5.537 6.344 5.537 6.344 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.057 6.071 mp_irecv_dv 3229 10.9 2.034 5.996 2.034 5.996 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.968 5.575 arnoldi_extremal 4 6.8 0.000 0.000 5.362 5.385 arnoldi_normal_ev 4 7.8 0.004 0.009 5.361 5.385 build_subspace 16 8.4 0.014 0.021 4.965 4.975 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.446 4.907 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.565 4.695 mp_allgather_i34 111 8.7 2.173 4.623 2.173 4.623 calculate_norms 2376 9.8 4.173 4.507 4.173 4.507 dbcsr_matrix_vector_mult 304 9.0 0.006 0.016 3.634 3.965 dbcsr_matrix_vector_mult_local 304 10.0 3.197 3.694 3.199 3.696 dbcsr_sort_data 658 11.4 3.094 3.509 3.094 3.509 ls_scf_post 1 4.0 0.003 0.015 3.291 3.301 dbcsr_special_finalize 555 9.7 0.006 0.007 2.838 3.225 dbcsr_merge_single_wm 555 10.7 0.538 0.666 2.829 3.217 ls_scf_store_result 1 5.0 0.000 0.000 3.025 3.079 ls_scf_dm_to_ks 2 5.0 0.001 0.009 2.956 3.007 dbcsr_data_release 10477 10.7 1.585 2.433 1.585 2.433 dbcsr_finalize 304 7.8 0.056 0.102 1.815 2.013 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.150000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 4.629008E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.081 99.048 99.050 qs_energies 1 2.0 0.000 0.000 98.254 98.260 ls_scf 1 3.0 0.000 0.000 96.341 96.346 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.933 78.161 ls_scf_main 1 4.0 0.000 0.000 62.144 62.145 multiply_cannon 111 7.7 0.091 0.172 55.399 60.398 density_matrix_trs4 2 5.0 0.002 0.003 55.051 55.141 multiply_cannon_loop 111 8.7 0.070 0.077 50.830 52.349 mp_waitall_1 5481 11.0 26.543 31.955 26.543 31.955 ls_scf_init_scf 1 4.0 0.000 0.000 30.549 30.555 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.378 29.424 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.195 27.207 multiply_cannon_multrec 444 9.7 14.071 16.389 21.090 23.152 make_m2s 222 7.7 0.004 0.005 17.768 20.307 make_images 222 8.7 3.721 4.413 17.706 20.248 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.116 16.413 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.273 15.655 make_images_data 222 9.7 0.003 0.004 9.952 12.523 hybrid_alltoall_any 227 10.6 0.789 3.775 9.761 12.297 dbcsr_mm_accdrv_process 3003 10.4 0.164 0.343 6.728 7.848 multiply_cannon_sync_h2d 444 9.7 6.521 7.796 6.521 7.796 dbcsr_mm_accdrv_process_sort 3003 11.4 6.414 7.505 6.414 7.505 mp_allgather_i34 111 8.7 2.799 6.990 2.799 6.990 arnoldi_extremal 4 6.8 0.000 0.000 5.890 5.904 arnoldi_normal_ev 4 7.8 0.002 0.005 5.890 5.904 build_subspace 16 8.4 0.015 0.020 5.496 5.506 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.638 4.749 mp_sum_l 807 5.4 2.815 4.705 2.815 4.705 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.221 4.393 dbcsr_matrix_vector_mult_local 304 10.0 3.712 4.179 3.714 4.182 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.645 3.783 mp_irecv_dv 1241 11.2 1.624 3.756 1.624 3.756 calculate_norms 792 9.8 3.547 3.698 3.547 3.698 ls_scf_post 1 4.0 0.000 0.000 3.647 3.653 make_images_sizes 222 9.7 0.000 0.000 1.168 3.581 mp_alltoall_i44 222 10.7 1.168 3.580 1.168 3.580 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.469 3.560 ls_scf_store_result 1 5.0 0.000 0.000 3.432 3.467 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.904 3.447 dbcsr_finalize 304 7.8 0.062 0.078 2.205 2.267 dbcsr_merge_all 275 8.9 0.473 0.524 2.054 2.097 dbcsr_data_release 10123 10.8 1.329 2.011 1.329 2.011 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.050000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3608.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b0f1689a086be006f33214c134fa9a07bd867c23_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 8.750436E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.138 0.145 108.592 108.599 qs_energies 1 2.0 0.000 0.000 107.069 107.078 ls_scf 1 3.0 0.000 0.000 104.146 104.154 dbcsr_multiply_generic 111 6.7 0.024 0.028 78.133 78.239 ls_scf_main 1 4.0 0.000 0.000 66.300 66.301 density_matrix_trs4 2 5.0 0.002 0.003 57.383 57.434 multiply_cannon 111 7.7 0.148 0.216 50.313 51.965 multiply_cannon_loop 111 8.7 0.067 0.070 46.967 48.023 ls_scf_init_scf 1 4.0 0.000 0.000 34.119 34.120 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.775 32.789 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.950 29.962 mp_waitall_1 4569 11.1 23.135 26.442 23.135 26.442 make_m2s 222 7.7 0.005 0.005 24.225 25.082 make_images 222 8.7 4.586 4.967 24.116 24.974 multiply_cannon_multrec 444 9.7 17.809 18.520 22.407 23.118 hybrid_alltoall_any 227 10.6 1.658 3.617 13.196 16.113 make_images_data 222 9.7 0.003 0.003 13.427 15.749 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.079 11.634 multiply_cannon_sync_h2d 444 9.7 8.814 8.870 8.814 8.870 arnoldi_extremal 4 6.8 0.000 0.000 7.533 7.543 arnoldi_normal_ev 4 7.8 0.003 0.009 7.533 7.543 build_subspace 16 8.4 0.026 0.037 6.982 6.991 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.569 5.729 dbcsr_matrix_vector_mult_local 304 10.0 5.080 5.415 5.083 5.418 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.068 5.309 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.819 4.921 dbcsr_mm_accdrv_process 1814 10.4 0.229 0.315 4.420 4.538 dbcsr_mm_accdrv_process_sort 1814 11.4 4.123 4.248 4.123 4.248 ls_scf_post 1 4.0 0.000 0.000 3.727 3.735 make_images_sizes 222 9.7 0.000 0.000 1.468 3.580 mp_alltoall_i44 222 10.7 1.468 3.580 1.468 3.580 ls_scf_store_result 1 5.0 0.000 0.000 3.474 3.481 calculate_norms 792 9.8 3.276 3.321 3.276 3.321 dbcsr_finalize 304 7.8 0.082 0.089 3.081 3.138 mp_allgather_i34 111 8.7 0.997 2.952 0.997 2.952 dbcsr_merge_all 275 8.9 0.891 0.910 2.867 2.914 dbcsr_complete_redistribute 5 7.6 1.447 1.502 2.788 2.901 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.892 2.892 dbcsr_data_release 12724 10.6 2.333 2.877 2.333 2.877 matrix_ls_to_qs 2 6.0 0.000 0.000 2.446 2.571 dbcsr_sort_data 325 11.1 2.442 2.504 2.442 2.504 dbcsr_new_transposed 4 7.5 0.243 0.253 2.289 2.304 dbcsr_frobenius_norm 74 6.6 2.057 2.136 2.205 2.247 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.210 2.211 dbcsr_add_d 103 6.2 0.000 0.000 2.134 2.204 dbcsr_add_anytype 103 7.2 0.860 0.892 2.134 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.599000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6843.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b0f1689a086be006f33214c134fa9a07bd867c23 Summary: empty Status: OK