=== 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: 65f55731d7d9d8bf99167de5ccd8ae8600377d2e ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (19.04.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/01 job id: 46667722 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/02 job id: 46667723 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/03 job id: 46667724 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/04 job id: 46667725 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/05 job id: 46667727 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/06 job id: 46667728 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/07 job id: 46667729 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/08 job id: 46667731 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/09 job id: 46667732 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/10 job id: 46667740 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/11 job id: 46667741 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/12 job id: 46667742 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/13 job id: 46667743 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/14 job id: 46667744 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/15 job id: 46667745 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/16 job id: 46667746 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/17 job id: 46667747 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/18 job id: 46667748 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/19 job id: 46667749 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/20 job id: 46667750 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/21 job id: 46667751 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/22 job id: 46667752 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/23 job id: 46667753 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/24 job id: 46667754 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/25 job id: 46667755 --- 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/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/26 job id: 46667756 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/27 job id: 46667758 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.035 136.267 136.267 farming_run 1 2.0 135.688 135.690 136.226 136.229 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.459143E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.028 115.426 115.426 qs_energies 1 2.0 0.000 0.000 115.170 115.171 mp2_main 1 3.0 0.000 0.000 112.875 112.877 mp2_gpw_main 1 4.0 0.020 0.028 111.839 111.841 mp2_ri_gpw_compute_in 1 5.0 0.171 0.173 92.704 92.781 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.327 55.404 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.172 41.644 46.579 get_2c_integrals 1 6.0 0.008 0.009 36.575 37.206 integrate_v_rspace 273 8.0 0.436 0.451 25.083 29.733 pw_transfer 6555 10.6 0.377 0.387 27.255 27.535 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.926 26.143 grid_integrate_task_list 273 9.0 20.917 26.000 20.917 26.000 fft_wrap_pw1pw2_100 2178 12.4 1.168 1.274 23.450 23.706 compute_2c_integrals 1 7.0 0.002 0.002 19.250 19.251 rpa_ri_compute_en 1 5.0 0.019 0.021 19.028 19.122 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.779 18.945 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.428 18.777 18.944 cp_fm_cholesky_decompose 12 8.2 17.317 17.976 17.317 17.976 cholesky_decomp 1 7.0 0.000 0.000 16.167 16.824 fft3d_s 5443 13.4 16.070 16.233 16.093 16.253 ao_to_mo_and_store_B_mult_1 272 7.0 10.851 15.564 10.851 15.564 calculate_wavefunction 272 8.0 5.408 5.494 12.499 13.133 rpa_num_int 1 6.0 0.000 0.000 10.711 10.711 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.691 10.697 calc_mat_Q 8 8.0 0.000 0.000 9.497 9.581 contract_S_to_Q 8 9.0 0.000 0.000 8.921 9.005 parallel_gemm_fm 14 9.1 0.000 0.000 8.521 8.612 parallel_gemm_fm_cosma 14 10.1 8.521 8.612 8.521 8.612 calc_potential_gpw 544 9.5 0.005 0.005 8.158 8.543 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.152 8.406 potential_pw2rs 545 10.0 0.107 0.109 7.684 8.381 create_integ_mat 1 6.0 0.023 0.028 8.144 8.145 collocate_single_gaussian 272 10.0 0.039 0.042 7.423 7.694 array2fm 1 7.0 0.000 0.000 6.956 7.511 pw_scatter_s 2720 13.7 4.403 4.588 4.403 4.588 pw_gather_s 2722 13.2 3.876 4.241 3.876 4.241 array2fm_buffer_send 1 8.0 2.993 3.211 2.993 3.211 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.840948, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2803.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.042 401.093 401.097 farming_run 1 2.0 399.967 400.004 401.036 401.041 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.213620E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 24538762. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.032 210.364 210.366 qs_energies 1 2.0 0.000 0.001 210.102 210.112 scf_env_do_scf 1 3.0 0.000 0.000 106.717 106.717 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.708 105.716 rebuild_ks_matrix 4 6.0 0.000 0.000 105.707 105.715 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.066 105.707 105.715 hfx_ks_matrix 4 8.0 0.001 0.001 105.297 105.302 integrate_four_center 4 9.0 0.143 0.456 105.296 105.301 mp2_main 1 3.0 0.000 0.002 103.091 103.100 mp2_gpw_main 1 4.0 0.032 0.048 101.997 102.005 integrate_four_center_main 4 10.0 0.141 0.635 96.716 99.879 integrate_four_center_bin 263 11.0 96.575 99.867 96.575 99.867 init_scf_loop 1 4.0 0.000 0.000 92.472 92.472 mp2_ri_gpw_compute_in 1 5.0 0.064 0.083 74.832 75.969 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.485 55.621 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.161 42.209 47.222 integrate_v_rspace 95 8.0 0.398 0.568 28.613 33.491 pw_transfer 2240 10.6 0.147 0.167 29.909 30.345 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.935 29.415 ao_to_mo_and_store_B_mult_1 91 7.0 10.599 29.336 10.599 29.336 grid_integrate_task_list 95 9.0 23.876 28.954 23.876 28.954 mp2_ri_gpw_compute_en 1 5.0 0.055 0.063 27.012 28.715 fft_wrap_pw1pw2_100 730 12.4 1.244 1.402 26.637 27.147 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.839 1.891 25.234 25.245 get_2c_integrals 1 6.0 0.000 0.000 20.265 20.283 compute_2c_integrals 1 7.0 0.003 0.004 19.252 19.255 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.788 19.104 mp2_eri_2c_integrate_gpw 1 9.0 1.745 1.851 18.787 19.103 fft3d_s 1823 13.4 18.431 18.840 18.445 18.853 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.241 14.241 calculate_wavefunction 91 8.0 2.004 2.041 9.733 10.000 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.588 8.749 9.380 potential_pw2rs 186 10.0 0.033 0.034 8.626 9.291 local_gemm 172 8.0 8.190 8.793 8.190 8.793 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.522 8.242 8.627 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.194 8.571 calc_potential_gpw 182 9.5 0.002 0.002 7.901 8.117 collocate_single_gaussian 91 10.0 0.017 0.019 7.850 8.086 mp_sendrecv_dm3 2068 8.0 6.289 6.677 6.289 6.677 mp2_ri_gpw_compute_en_ener 172 7.0 6.359 6.413 6.359 6.413 pw_gather_s 912 13.2 4.929 5.355 4.929 5.355 mp_sync 37 10.5 3.196 5.207 3.196 5.207 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.992778, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 448.724992E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 2293187. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.037 55.205 55.206 qs_mol_dyn_low 1 2.0 0.003 0.003 54.962 54.969 qs_forces 11 3.9 0.002 0.003 54.895 54.896 qs_energies 11 4.9 0.002 0.008 53.387 53.409 scf_env_do_scf 11 5.9 0.000 0.001 46.386 46.387 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.297 44.297 dbcsr_multiply_generic 2286 12.5 0.096 0.099 34.388 34.835 qs_scf_new_mos 108 7.5 0.000 0.001 33.928 34.223 qs_scf_loop_do_ot 108 8.5 0.001 0.001 33.928 34.223 ot_scf_mini 108 9.5 0.002 0.002 32.217 32.408 multiply_cannon 2286 13.5 0.189 0.206 26.826 28.482 multiply_cannon_loop 2286 14.5 1.483 1.573 26.128 27.825 velocity_verlet 10 3.0 0.001 0.002 26.052 26.052 ot_mini 108 10.5 0.001 0.001 19.299 19.562 qs_ot_get_derivative 108 11.5 0.001 0.001 16.369 16.585 mp_waitall_1 245248 16.5 8.755 15.677 8.755 15.677 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.251 13.875 multiply_cannon_multrec 54864 15.5 4.266 6.599 8.176 11.927 qs_ot_get_p 119 10.4 0.001 0.001 8.221 8.533 rebuild_ks_matrix 119 8.3 0.000 0.000 7.993 8.115 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.993 8.114 mp_sum_l 7287 12.8 5.602 7.267 5.602 7.267 multiply_cannon_sync_h2d 54864 15.5 5.969 7.172 5.969 7.172 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.035 7.147 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.576 6.059 init_scf_run 11 5.9 0.000 0.001 5.734 5.734 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.734 5.734 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.562 5.616 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.508 5.615 dbcsr_mm_accdrv_process 76910 16.1 1.142 1.844 3.831 5.529 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.717 4.718 sum_up_and_integrate 119 10.3 0.012 0.014 4.623 4.630 integrate_v_rspace 119 11.3 0.002 0.003 4.611 4.619 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.362 4.511 calculate_rho_elec 119 8.7 0.012 0.017 4.362 4.510 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.495 4.496 cp_fm_redistribute_end 50 14.0 2.290 4.467 2.298 4.471 cp_fm_diag_elpa_base 50 14.0 2.167 4.355 2.171 4.365 jit_kernel_multiply 13 15.8 2.627 3.620 2.627 3.620 calculate_first_density_matrix 1 7.0 0.000 0.002 3.535 3.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.297 3.299 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.816 3.098 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.019 3.067 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.872 3.053 apply_single 119 13.6 0.000 0.000 2.872 3.053 calculate_dm_sparse 119 9.5 0.000 0.001 2.934 3.049 rs_pw_transfer 974 11.9 0.012 0.013 2.805 2.931 ot_diis_step 108 11.5 0.006 0.006 2.643 2.643 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.531 2.600 density_rs2pw 119 9.7 0.004 0.005 2.270 2.418 acc_transpose_blocks 54864 15.5 0.225 0.247 1.722 2.191 grid_integrate_task_list 119 12.3 2.049 2.167 2.049 2.167 wfi_extrapolate 11 7.9 0.001 0.001 2.134 2.134 init_scf_loop 11 6.9 0.001 0.007 2.067 2.067 potential_pw2rs 119 12.3 0.004 0.004 1.918 1.930 pw_transfer 1439 11.6 0.052 0.058 1.798 1.875 mp_sum_d 4135 12.0 1.303 1.871 1.303 1.871 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.722 1.802 make_m2s 4572 13.5 0.054 0.056 1.649 1.693 make_images 4572 14.5 0.133 0.138 1.567 1.610 mp_alltoall_d11v 2130 13.8 1.332 1.552 1.332 1.552 grid_collocate_task_list 119 9.7 1.391 1.475 1.391 1.475 fft3d_ps 1201 14.6 0.375 0.482 1.366 1.438 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.388 1.414 fft_wrap_pw1pw2_140 487 13.2 0.184 0.201 1.325 1.407 mp_waitany 12084 13.8 1.271 1.402 1.271 1.402 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.733 1.107 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.206000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 484.192256E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1404311. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 41.538 41.539 qs_mol_dyn_low 1 2.0 0.003 0.003 41.363 41.370 qs_forces 11 3.9 0.002 0.003 40.704 40.705 qs_energies 11 4.9 0.001 0.001 38.989 38.991 scf_env_do_scf 11 5.9 0.000 0.001 32.123 32.124 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.551 29.552 dbcsr_multiply_generic 2286 12.5 0.099 0.102 23.102 23.441 qs_scf_new_mos 108 7.5 0.001 0.001 20.180 20.424 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.179 20.424 multiply_cannon 2286 13.5 0.210 0.219 18.086 19.670 ot_scf_mini 108 9.5 0.002 0.003 19.289 19.465 velocity_verlet 10 3.0 0.001 0.002 19.085 19.099 multiply_cannon_loop 2286 14.5 0.905 0.985 16.971 18.470 mp_waitall_1 200699 16.5 6.501 12.631 6.501 12.631 ot_mini 108 10.5 0.001 0.001 11.821 12.056 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.885 11.359 multiply_cannon_multrec 27432 15.5 1.975 4.390 6.501 9.804 qs_ot_get_derivative 108 11.5 0.001 0.001 9.369 9.546 rebuild_ks_matrix 119 8.3 0.000 0.000 7.306 7.438 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.305 7.438 dbcsr_mm_accdrv_process 47894 16.0 3.383 5.900 4.455 6.794 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.438 6.558 init_scf_run 11 5.9 0.000 0.001 5.659 5.660 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.659 5.659 qs_ot_get_p 119 10.4 0.001 0.001 4.580 4.815 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.585 4.478 sum_up_and_integrate 119 10.3 0.024 0.027 4.313 4.318 integrate_v_rspace 119 11.3 0.002 0.003 4.288 4.296 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.084 4.161 apply_single 119 13.6 0.000 0.000 3.084 4.160 calculate_first_density_matrix 1 7.0 0.000 0.000 4.117 4.120 mp_sum_l 7287 12.8 2.088 4.019 2.088 4.019 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.920 3.949 calculate_rho_elec 119 8.7 0.021 0.024 3.920 3.948 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.703 3.705 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.486 3.522 rs_pw_transfer 974 11.9 0.010 0.011 2.695 3.168 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.124 3.144 multiply_cannon_sync_h2d 27432 15.5 2.181 2.828 2.181 2.828 make_m2s 4572 13.5 0.052 0.054 2.481 2.693 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.685 2.685 density_rs2pw 119 9.7 0.004 0.004 2.194 2.676 make_images 4572 14.5 0.200 0.235 2.392 2.601 jit_kernel_multiply 10 16.1 1.020 2.569 1.020 2.569 init_scf_loop 11 6.9 0.000 0.000 2.549 2.549 ot_diis_step 108 11.5 0.010 0.011 2.401 2.402 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.341 2.342 cp_fm_redistribute_end 50 14.0 1.185 2.311 1.190 2.314 cp_fm_diag_elpa_base 50 14.0 1.091 2.211 1.120 2.254 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.157 2.249 acc_transpose_blocks 27432 15.5 0.109 0.113 1.545 2.176 calculate_dm_sparse 119 9.5 0.000 0.001 2.031 2.105 grid_integrate_task_list 119 12.3 1.877 1.998 1.877 1.998 potential_pw2rs 119 12.3 0.006 0.006 1.974 1.987 pw_transfer 1439 11.6 0.065 0.070 1.951 1.985 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.860 1.895 prepare_preconditioner 11 7.9 0.000 0.000 1.625 1.652 make_preconditioner 11 8.9 0.000 0.000 1.625 1.652 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.521 1.578 make_images_data 4572 15.5 0.045 0.051 1.141 1.564 acc_transpose_blocks_kernels 27432 16.5 0.183 0.272 1.008 1.551 fft_wrap_pw1pw2_140 487 13.2 0.202 0.211 1.453 1.490 wfi_extrapolate 11 7.9 0.001 0.001 1.489 1.489 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.004 1.448 fft3d_ps 1201 14.6 0.521 0.577 1.410 1.441 grid_collocate_task_list 119 9.7 1.283 1.428 1.283 1.428 mp_alltoall_d11v 2130 13.8 1.231 1.414 1.231 1.414 mp_allgather_i34 2286 14.5 0.553 1.297 0.553 1.297 jit_kernel_transpose 5 15.5 0.826 1.284 0.826 1.284 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.251 1.258 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.175 1.223 mp_waitany 5720 13.7 0.520 1.025 0.520 1.025 mp_sum_d 4135 12.0 0.568 1.016 0.568 1.016 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.541 1.008 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.925 0.937 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.933 0.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.539000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.909091, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 518.332416E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1927562. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.029 34.863 34.863 qs_mol_dyn_low 1 2.0 0.003 0.003 34.532 34.663 qs_forces 11 3.9 0.002 0.003 34.463 34.464 qs_energies 11 4.9 0.001 0.001 32.860 32.863 scf_env_do_scf 11 5.9 0.000 0.001 26.115 26.115 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.504 23.504 dbcsr_multiply_generic 2286 12.5 0.093 0.096 18.427 18.534 multiply_cannon 2286 13.5 0.196 0.203 15.071 15.752 velocity_verlet 10 3.0 0.001 0.001 15.587 15.588 qs_scf_new_mos 108 7.5 0.001 0.001 15.087 15.108 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.086 15.107 multiply_cannon_loop 2286 14.5 0.634 0.667 14.263 14.987 ot_scf_mini 108 9.5 0.002 0.002 14.357 14.374 ot_mini 108 10.5 0.001 0.001 8.745 8.757 multiply_cannon_multrec 18288 15.5 1.940 2.969 7.462 8.164 qs_ot_get_derivative 108 11.5 0.001 0.001 7.238 7.253 dbcsr_mm_accdrv_process 38222 16.0 4.675 6.687 5.437 6.746 rebuild_ks_matrix 119 8.3 0.000 0.000 6.502 6.515 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.012 6.501 6.514 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.732 5.744 init_scf_run 11 5.9 0.000 0.001 5.587 5.587 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.587 5.587 mp_waitall_1 158411 16.6 3.627 5.227 3.627 5.227 calculate_first_density_matrix 1 7.0 0.000 0.000 4.339 4.340 sum_up_and_integrate 119 10.3 0.030 0.031 4.098 4.103 integrate_v_rspace 119 11.3 0.003 0.003 4.068 4.076 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.805 3.807 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 2.005 3.749 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.638 3.648 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.576 3.583 calculate_rho_elec 119 8.7 0.031 0.031 3.575 3.583 qs_ot_get_p 119 10.4 0.001 0.001 3.435 3.454 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.720 3.322 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.735 2.944 jit_kernel_multiply 8 15.9 0.711 2.689 0.711 2.689 rs_pw_transfer 974 11.9 0.009 0.010 2.358 2.634 init_scf_loop 11 6.9 0.000 0.000 2.592 2.594 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.023 2.349 apply_single 119 13.6 0.000 0.000 2.023 2.349 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.333 2.340 density_rs2pw 119 9.7 0.004 0.004 2.045 2.322 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.064 2.064 make_m2s 4572 13.5 0.045 0.045 1.878 2.037 acc_transpose_blocks 18288 15.5 0.075 0.077 1.662 1.964 grid_integrate_task_list 119 12.3 1.862 1.964 1.862 1.964 pw_transfer 1439 11.6 0.065 0.070 1.945 1.956 make_images 4572 14.5 0.190 0.202 1.791 1.950 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.852 1.866 calculate_dm_sparse 119 9.5 0.000 0.000 1.816 1.834 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.781 1.803 prepare_preconditioner 11 7.9 0.000 0.000 1.799 1.801 make_preconditioner 11 8.9 0.000 0.000 1.799 1.801 cp_fm_diag_elpa_base 50 14.0 1.757 1.778 1.780 1.801 potential_pw2rs 119 12.3 0.007 0.008 1.783 1.794 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.648 1.735 multiply_cannon_sync_h2d 18288 15.5 1.366 1.597 1.366 1.597 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.584 1.594 acc_transpose_blocks_kernels 18288 16.5 0.210 0.220 1.219 1.510 ot_diis_step 108 11.5 0.011 0.011 1.484 1.484 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.467 1.482 mp_sum_l 7287 12.8 1.093 1.418 1.093 1.418 grid_collocate_task_list 119 9.7 1.258 1.409 1.258 1.409 fft3d_ps 1201 14.6 0.530 0.547 1.339 1.351 jit_kernel_transpose 5 15.6 1.009 1.299 1.009 1.299 wfi_extrapolate 11 7.9 0.001 0.001 1.202 1.202 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.183 1.188 make_images_data 4572 15.5 0.044 0.048 0.850 1.024 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.931 0.933 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.896 0.916 hybrid_alltoall_any 4725 16.4 0.055 0.111 0.725 0.904 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.840 0.842 mp_alltoall_d11v 2130 13.8 0.721 0.830 0.721 0.830 mp_waitany 9880 13.7 0.508 0.807 0.508 0.807 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.504 0.787 cp_fm_cholesky_invert 11 10.9 0.747 0.751 0.747 0.751 mp_alltoall_z22v 1201 16.6 0.659 0.728 0.659 0.728 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.648 0.721 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.863000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.818182, yerr=1.849726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 550.633472E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.028 37.346 37.346 qs_mol_dyn_low 1 2.0 0.003 0.003 36.678 37.175 qs_forces 11 3.9 0.022 0.024 36.580 36.581 qs_energies 11 4.9 0.003 0.022 34.814 34.819 scf_env_do_scf 11 5.9 0.001 0.002 28.308 28.309 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.871 24.872 dbcsr_multiply_generic 2286 12.5 0.098 0.104 19.533 19.646 velocity_verlet 10 3.0 0.001 0.001 17.960 17.962 multiply_cannon 2286 13.5 0.227 0.264 15.854 16.346 qs_scf_new_mos 108 7.5 0.001 0.001 16.282 16.344 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.282 16.343 ot_scf_mini 108 9.5 0.002 0.003 15.345 15.400 multiply_cannon_loop 2286 14.5 0.939 0.979 14.903 15.351 ot_mini 108 10.5 0.001 0.001 9.288 9.353 multiply_cannon_multrec 27432 15.5 2.381 3.066 8.724 9.242 qs_ot_get_derivative 108 11.5 0.001 0.001 7.473 7.527 dbcsr_mm_accdrv_process 47916 15.9 5.478 7.171 6.251 7.386 rebuild_ks_matrix 119 8.3 0.000 0.000 6.682 6.741 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.682 6.740 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.920 5.967 init_scf_run 11 5.9 0.000 0.001 5.137 5.137 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.136 5.137 sum_up_and_integrate 119 10.3 0.035 0.038 3.920 3.929 integrate_v_rspace 119 11.3 0.003 0.003 3.885 3.894 calculate_first_density_matrix 1 7.0 0.000 0.002 3.733 3.736 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.609 3.637 calculate_rho_elec 119 8.7 0.040 0.046 3.608 3.637 qs_ot_get_p 119 10.4 0.001 0.001 3.448 3.516 init_scf_loop 11 6.9 0.001 0.009 3.417 3.419 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.671 3.161 calculate_dm_sparse 119 9.5 0.000 0.000 2.897 2.952 mp_waitall_1 137007 16.6 2.285 2.946 2.285 2.946 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.137 2.581 apply_single 119 13.6 0.000 0.000 2.137 2.581 prepare_preconditioner 11 7.9 0.000 0.000 2.555 2.564 make_preconditioner 11 8.9 0.000 0.001 2.555 2.564 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.157 2.489 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.411 2.412 acc_transpose_blocks 27432 15.5 0.109 0.111 2.302 2.319 make_m2s 4572 13.5 0.054 0.057 2.177 2.282 rs_pw_transfer 974 11.9 0.009 0.009 2.067 2.258 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.232 2.246 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.176 2.186 density_rs2pw 119 9.7 0.004 0.004 1.982 2.176 make_images 4572 14.5 0.270 0.331 2.069 2.171 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.079 2.091 pw_transfer 1439 11.6 0.065 0.070 2.032 2.069 grid_integrate_task_list 119 12.3 1.874 1.994 1.874 1.994 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.940 1.982 qs_ot_get_derivative_diag 49 12.0 0.001 0.003 1.922 1.950 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.860 1.860 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 1.084 1.821 ot_diis_step 108 11.5 0.012 0.012 1.776 1.776 acc_transpose_blocks_kernels 27432 16.5 0.269 0.278 1.675 1.693 fft_wrap_pw1pw2_140 487 13.2 0.288 0.304 1.587 1.632 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.604 1.611 cp_fm_diag_elpa_base 50 14.0 1.569 1.584 1.601 1.609 potential_pw2rs 119 12.3 0.008 0.009 1.599 1.603 mp_sum_l 7287 12.8 1.081 1.591 1.081 1.591 jit_kernel_multiply 8 16.1 0.713 1.445 0.713 1.445 jit_kernel_transpose 5 15.6 1.406 1.428 1.406 1.428 fft3d_ps 1201 14.6 0.557 0.607 1.377 1.406 grid_collocate_task_list 119 9.7 1.260 1.391 1.260 1.391 wfi_extrapolate 11 7.9 0.001 0.001 1.352 1.352 cp_fm_upper_to_full 72 14.2 0.820 1.170 0.820 1.170 multiply_cannon_metrocomm1 27432 15.5 0.034 0.035 0.344 1.164 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.107 1.127 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.095 1.095 multiply_cannon_sync_h2d 27432 15.5 0.996 1.091 0.996 1.091 dbcsr_complete_redistribute 329 12.2 0.117 0.143 0.777 1.055 make_images_data 4572 15.5 0.044 0.048 0.851 0.969 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.746 0.908 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.869 mp_alltoall_d11v 2130 13.8 0.760 0.865 0.760 0.865 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.857 0.862 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.570 0.841 cp_fm_cholesky_invert 11 10.9 0.750 0.753 0.750 0.753 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.346000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=523.090909, yerr=3.203820 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 598.347776E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 30.981 30.982 qs_mol_dyn_low 1 2.0 0.003 0.003 30.286 30.294 qs_forces 11 3.9 0.002 0.002 30.228 30.228 qs_energies 11 4.9 0.001 0.001 28.481 28.484 scf_env_do_scf 11 5.9 0.000 0.001 21.785 21.785 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.253 19.254 velocity_verlet 10 3.0 0.001 0.001 14.525 14.528 dbcsr_multiply_generic 2286 12.5 0.090 0.096 14.289 14.408 multiply_cannon 2286 13.5 0.231 0.241 11.777 12.263 qs_scf_new_mos 108 7.5 0.001 0.001 11.162 11.185 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.161 11.184 multiply_cannon_loop 2286 14.5 0.330 0.341 10.869 11.100 ot_scf_mini 108 9.5 0.002 0.002 10.489 10.515 multiply_cannon_multrec 9144 15.5 1.713 1.949 6.644 7.256 rebuild_ks_matrix 119 8.3 0.000 0.000 6.078 6.096 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.078 6.096 ot_mini 108 10.5 0.001 0.001 5.921 5.952 dbcsr_mm_accdrv_process 12550 15.8 3.637 5.029 4.829 5.411 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.394 5.410 init_scf_run 11 5.9 0.000 0.001 5.268 5.269 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.268 5.268 qs_ot_get_derivative 108 11.5 0.001 0.001 4.626 4.652 calculate_first_density_matrix 1 7.0 0.000 0.000 4.109 4.109 sum_up_and_integrate 119 10.3 0.037 0.042 3.764 3.770 integrate_v_rspace 119 11.3 0.003 0.003 3.726 3.733 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.576 3.582 calculate_rho_elec 119 8.7 0.060 0.061 3.575 3.582 mp_waitall_1 115863 16.7 2.369 3.443 2.369 3.443 calculate_dm_sparse 119 9.5 0.000 0.000 3.289 3.308 qs_ot_get_p 119 10.4 0.001 0.001 2.783 2.831 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.711 2.717 init_scf_loop 11 6.9 0.000 0.000 2.512 2.514 jit_kernel_multiply 8 15.6 1.154 2.173 1.154 2.173 pw_transfer 1439 11.6 0.066 0.068 2.058 2.068 density_rs2pw 119 9.7 0.004 0.004 1.858 2.000 grid_integrate_task_list 119 12.3 1.911 1.992 1.911 1.992 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.964 1.975 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.969 1.971 make_m2s 4572 13.5 0.035 0.036 1.753 1.898 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.978 1.871 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.850 1.852 acc_transpose_blocks 9144 15.5 0.038 0.038 1.370 1.840 rs_pw_transfer 974 11.9 0.008 0.008 1.706 1.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.815 1.822 make_images 4572 14.5 0.267 0.303 1.663 1.807 prepare_preconditioner 11 7.9 0.000 0.000 1.767 1.771 make_preconditioner 11 8.9 0.000 0.000 1.767 1.771 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.655 1.680 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.653 1.653 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.559 1.639 fft_wrap_pw1pw2_140 487 13.2 0.365 0.376 1.606 1.618 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 1.128 1.596 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.568 1.587 jit_kernel_transpose 5 15.6 1.011 1.477 1.011 1.477 grid_collocate_task_list 119 9.7 1.309 1.430 1.309 1.430 potential_pw2rs 119 12.3 0.010 0.011 1.420 1.426 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.362 1.370 cp_fm_diag_elpa_base 50 14.0 1.335 1.351 1.360 1.369 fft3d_ps 1201 14.6 0.561 0.575 1.297 1.308 ot_diis_step 108 11.5 0.012 0.013 1.281 1.282 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.215 1.215 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.157 1.174 apply_single 119 13.6 0.000 0.000 1.156 1.174 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.138 1.152 wfi_extrapolate 11 7.9 0.001 0.001 1.110 1.110 hybrid_alltoall_any 4725 16.4 0.062 0.176 0.802 1.027 make_images_data 4572 15.5 0.038 0.042 0.823 0.991 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.867 0.919 mp_alltoall_d11v 2130 13.8 0.797 0.912 0.797 0.912 cp_fm_cholesky_invert 11 10.9 0.891 0.894 0.891 0.894 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.808 multiply_cannon_sync_h2d 9144 15.5 0.712 0.788 0.712 0.788 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.723 0.730 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.669 0.720 mp_allgather_i34 2286 14.5 0.239 0.665 0.239 0.665 mp_alltoall_z22v 1201 16.6 0.610 0.642 0.610 0.642 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.982000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=569.090909, yerr=4.187743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 763.691008E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 43.263 43.263 qs_mol_dyn_low 1 2.0 0.003 0.003 43.063 43.070 qs_forces 11 3.9 0.002 0.002 42.997 42.998 qs_energies 11 4.9 0.001 0.001 40.974 40.977 scf_env_do_scf 11 5.9 0.001 0.001 33.689 33.689 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.689 25.691 velocity_verlet 10 3.0 0.001 0.001 23.581 23.587 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.982 19.161 multiply_cannon 2286 13.5 0.303 0.313 15.153 16.044 qs_scf_new_mos 108 7.5 0.001 0.001 15.862 15.961 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.862 15.961 ot_scf_mini 108 9.5 0.002 0.002 14.776 14.879 multiply_cannon_loop 2286 14.5 0.344 0.350 13.877 14.797 multiply_cannon_multrec 9144 15.5 3.738 5.816 9.484 9.694 ot_mini 108 10.5 0.001 0.001 8.784 8.903 init_scf_loop 11 6.9 0.000 0.000 7.972 7.977 rebuild_ks_matrix 119 8.3 0.000 0.000 7.366 7.508 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.366 7.507 dbcsr_mm_accdrv_process 12550 15.8 4.097 5.585 5.623 7.425 prepare_preconditioner 11 7.9 0.000 0.000 6.997 7.010 make_preconditioner 11 8.9 0.000 0.000 6.997 7.010 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.538 6.888 qs_ot_get_derivative 108 11.5 0.001 0.001 6.739 6.843 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.630 6.758 init_scf_run 11 5.9 0.000 0.001 5.261 5.261 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.261 5.261 cp_fm_upper_to_full 72 14.2 3.231 4.636 3.231 4.636 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.337 4.399 calculate_rho_elec 119 8.7 0.117 0.120 4.336 4.399 sum_up_and_integrate 119 10.3 0.064 0.066 4.088 4.095 integrate_v_rspace 119 11.3 0.003 0.004 4.024 4.029 calculate_first_density_matrix 1 7.0 0.000 0.000 3.730 3.795 mp_waitall_1 94719 16.7 2.500 3.448 2.500 3.448 qs_ot_get_p 119 10.4 0.001 0.001 3.257 3.390 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.373 3.374 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.121 3.164 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.477 2.887 dbcsr_complete_redistribute 329 12.2 0.288 0.290 2.008 2.872 pw_transfer 1439 11.6 0.069 0.069 2.678 2.682 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.579 2.583 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.689 2.533 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.197 2.482 apply_single 119 13.6 0.000 0.000 2.197 2.482 make_m2s 4572 13.5 0.038 0.038 2.269 2.454 make_images 4572 14.5 0.353 0.381 2.148 2.332 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.454 2.292 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.364 2.289 calculate_dm_sparse 119 9.5 0.000 0.000 2.213 2.273 mp_alltoall_i22 627 13.8 1.432 2.258 1.432 2.258 density_rs2pw 119 9.7 0.004 0.004 2.180 2.195 fft_wrap_pw1pw2_140 487 13.2 0.616 0.620 2.162 2.169 grid_integrate_task_list 119 12.3 2.122 2.136 2.122 2.136 jit_kernel_multiply 6 15.7 1.499 2.128 1.499 2.128 ot_diis_step 108 11.5 0.014 0.014 2.017 2.018 qs_ot_p2m_diag 50 11.0 0.043 0.043 2.001 2.003 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.753 1.754 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.685 1.736 mp_sum_l 7287 12.8 1.016 1.733 1.016 1.733 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.721 1.722 fft3d_ps 1201 14.6 0.597 0.607 1.596 1.603 acc_transpose_blocks 9144 15.5 0.038 0.039 1.405 1.543 grid_collocate_task_list 119 9.7 1.516 1.531 1.516 1.531 cp_fm_cholesky_invert 11 10.9 1.510 1.514 1.510 1.514 rs_pw_transfer 974 11.9 0.009 0.009 1.481 1.504 potential_pw2rs 119 12.3 0.014 0.014 1.416 1.418 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.415 1.416 cp_fm_diag_elpa_base 50 14.0 1.267 1.323 1.413 1.413 wfi_extrapolate 11 7.9 0.001 0.001 1.398 1.398 hybrid_alltoall_any 4725 16.4 0.088 0.149 1.115 1.389 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.277 1.339 make_images_data 4572 15.5 0.042 0.045 1.068 1.307 acc_transpose_blocks_kernels 9144 16.5 0.118 0.121 1.152 1.288 mp_alltoall_d11v 2130 13.8 1.136 1.200 1.136 1.200 jit_kernel_transpose 5 15.6 1.034 1.169 1.034 1.169 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.135 1.157 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.094 1.111 multiply_cannon_sync_h2d 9144 15.5 1.039 1.044 1.039 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.986 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.948 0.961 qs_create_task_list 11 7.9 0.000 0.001 0.935 0.946 generate_qs_task_list 11 8.9 0.369 0.389 0.935 0.946 mp_alltoall_z22v 1201 16.6 0.864 0.880 0.864 0.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.263000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.636364, yerr=13.213442 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 497.930240E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1706627. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.030 94.157 94.157 qs_mol_dyn_low 1 2.0 0.003 0.003 93.859 93.869 qs_forces 11 3.9 0.003 0.003 93.285 93.287 qs_energies 11 4.9 0.001 0.001 90.179 90.196 scf_env_do_scf 11 5.9 0.000 0.001 79.283 79.285 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 71.783 71.783 dbcsr_multiply_generic 2055 12.4 0.113 0.118 54.172 54.493 qs_scf_new_mos 99 7.5 0.000 0.001 51.458 51.591 qs_scf_loop_do_ot 99 8.5 0.000 0.001 51.457 51.590 ot_scf_mini 99 9.5 0.002 0.002 48.966 49.078 velocity_verlet 10 3.0 0.001 0.002 48.499 48.505 multiply_cannon 2055 13.4 0.183 0.192 44.100 45.125 multiply_cannon_loop 2055 14.4 1.522 1.572 42.687 43.802 ot_mini 99 10.5 0.001 0.001 27.752 27.871 qs_ot_get_derivative 99 11.5 0.001 0.001 20.926 21.043 multiply_cannon_multrec 49320 15.4 12.332 13.114 18.202 19.049 rebuild_ks_matrix 110 8.3 0.000 0.000 16.341 16.547 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 16.341 16.547 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.335 14.525 mp_waitall_1 220248 16.4 12.611 14.047 12.611 14.047 qs_ot_get_p 110 10.4 0.001 0.001 12.367 12.512 multiply_cannon_sync_h2d 49320 15.4 10.118 10.798 10.118 10.798 init_scf_run 11 5.9 0.000 0.001 8.683 8.683 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.683 8.683 qs_ot_p2m_diag 48 11.0 0.012 0.019 8.604 8.658 sum_up_and_integrate 110 10.3 0.036 0.042 8.541 8.556 integrate_v_rspace 110 11.3 0.003 0.004 8.505 8.526 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.313 8.450 calculate_rho_elec 110 8.6 0.021 0.026 8.312 8.450 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.579 8.163 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.435 7.898 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.258 7.702 apply_single 110 13.6 0.000 0.000 7.257 7.702 cp_dbcsr_syevd 48 12.0 0.002 0.003 7.586 7.587 init_scf_loop 11 6.9 0.000 0.000 7.457 7.458 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.748 6.775 cp_fm_diag_elpa_base 48 14.0 6.713 6.747 6.746 6.773 ot_diis_step 99 11.5 0.005 0.006 6.624 6.625 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.265 6.316 rs_pw_transfer 902 11.9 0.012 0.014 5.510 6.155 dbcsr_mm_accdrv_process 87628 16.1 1.913 2.000 5.746 6.114 mp_sum_l 6594 12.7 4.458 5.545 4.458 5.545 density_rs2pw 110 9.6 0.004 0.005 4.957 5.515 make_m2s 4110 13.4 0.061 0.065 5.145 5.359 make_images 4110 14.4 0.179 0.197 5.048 5.264 prepare_preconditioner 11 7.9 0.000 0.000 5.100 5.121 make_preconditioner 11 8.9 0.000 0.000 5.100 5.121 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.052 5.062 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.855 4.933 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 3.182 4.711 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.619 4.697 wfi_extrapolate 11 7.9 0.001 0.001 4.578 4.578 pw_transfer 1331 11.6 0.056 0.065 4.364 4.557 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.275 4.471 calculate_first_density_matrix 1 7.0 0.000 0.000 3.998 4.001 potential_pw2rs 110 12.3 0.006 0.007 3.851 3.952 calculate_dm_sparse 110 9.5 0.001 0.001 3.867 3.946 jit_kernel_multiply 13 15.9 3.556 3.707 3.556 3.707 fft3d_ps 1111 14.6 0.795 0.906 3.476 3.651 fft_wrap_pw1pw2_140 451 13.1 0.446 0.494 3.426 3.627 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.489 3.580 grid_integrate_task_list 110 12.3 3.306 3.507 3.306 3.507 mp_alltoall_d11v 2046 13.8 2.512 3.086 2.512 3.086 cp_fm_cholesky_invert 11 10.9 2.985 2.990 2.985 2.990 make_images_data 4110 15.4 0.043 0.046 2.118 2.643 mp_waitany 14300 13.8 2.021 2.641 2.021 2.641 mp_alltoall_z22v 1111 16.6 2.254 2.533 2.254 2.533 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.782 2.510 grid_collocate_task_list 110 9.6 2.183 2.398 2.183 2.398 acc_transpose_blocks 49320 15.4 0.208 0.217 2.123 2.341 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.137 2.154 make_images_sizes 4110 15.4 0.004 0.004 1.469 1.915 mp_alltoall_i44 4110 16.4 1.465 1.911 1.465 1.911 mp_sum_d 3889 11.9 1.411 1.899 1.411 1.899 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=94.157000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=471.909091, yerr=2.391445 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 582.311936E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2291634. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.102 73.330 73.362 qs_mol_dyn_low 1 2.0 0.007 0.033 72.843 72.895 qs_forces 11 3.9 0.003 0.005 71.097 71.101 qs_energies 11 4.9 0.002 0.005 67.672 67.677 scf_env_do_scf 11 5.9 0.001 0.002 57.792 57.796 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 50.002 50.003 dbcsr_multiply_generic 2055 12.4 0.116 0.121 38.831 39.045 velocity_verlet 10 3.0 0.001 0.001 38.278 38.317 qs_scf_new_mos 99 7.5 0.001 0.001 33.326 33.472 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.326 33.471 multiply_cannon 2055 13.4 0.222 0.245 32.163 33.404 multiply_cannon_loop 2055 14.4 0.924 0.945 30.907 32.025 ot_scf_mini 99 9.5 0.003 0.004 31.661 31.823 ot_mini 99 10.5 0.001 0.001 18.274 18.440 multiply_cannon_multrec 24660 15.4 7.612 9.530 14.829 16.977 rebuild_ks_matrix 110 8.3 0.000 0.001 14.065 14.228 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.064 14.227 qs_ot_get_derivative 99 11.5 0.001 0.001 12.455 12.617 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.333 12.479 mp_waitall_1 176588 16.5 7.848 10.610 7.848 10.610 multiply_cannon_sync_h2d 24660 15.4 7.046 8.081 7.046 8.081 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.265 7.919 init_scf_loop 11 6.9 0.001 0.004 7.753 7.756 dbcsr_mm_accdrv_process 52282 16.1 4.756 6.375 7.056 7.439 init_scf_run 11 5.9 0.000 0.001 7.413 7.414 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.412 7.413 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.514 7.221 apply_single 110 13.6 0.000 0.001 6.514 7.221 qs_ot_get_p 110 10.4 0.001 0.001 6.761 6.947 sum_up_and_integrate 110 10.3 0.053 0.060 6.895 6.912 integrate_v_rspace 110 11.3 0.003 0.003 6.842 6.857 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.375 6.387 calculate_rho_elec 110 8.6 0.040 0.047 6.375 6.386 ot_diis_step 99 11.5 0.010 0.010 5.768 5.768 prepare_preconditioner 11 7.9 0.000 0.000 5.684 5.708 make_preconditioner 11 8.9 0.000 0.002 5.684 5.708 make_full_inverse_cholesky 11 9.9 0.000 0.002 5.258 5.423 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.671 5.383 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.843 4.863 make_m2s 4110 13.4 0.057 0.060 4.220 4.643 make_images 4110 14.4 0.398 0.442 4.110 4.529 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.466 4.468 cp_dbcsr_syevd 48 12.0 0.003 0.005 4.399 4.400 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.099 4.171 density_rs2pw 110 9.6 0.005 0.006 3.439 3.977 pw_transfer 1331 11.6 0.067 0.074 3.798 3.943 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.691 3.837 cp_fm_diag_elpa 48 13.0 0.000 0.001 3.804 3.818 cp_fm_diag_elpa_base 48 14.0 3.754 3.771 3.800 3.814 calculate_first_density_matrix 1 7.0 0.001 0.003 3.756 3.759 rs_pw_transfer 902 11.9 0.012 0.015 3.157 3.735 wfi_extrapolate 11 7.9 0.001 0.001 3.551 3.551 grid_integrate_task_list 110 12.3 3.251 3.449 3.251 3.449 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.224 3.312 fft_wrap_pw1pw2_140 451 13.1 0.521 0.536 3.102 3.251 jit_kernel_multiply 11 16.3 1.950 3.112 1.950 3.112 calculate_dm_sparse 110 9.5 0.001 0.001 2.967 2.996 fft3d_ps 1111 14.6 1.110 1.340 2.661 2.830 make_images_data 4110 15.4 0.046 0.051 2.354 2.768 cp_fm_cholesky_invert 11 10.9 2.693 2.701 2.693 2.701 hybrid_alltoall_any 4261 16.3 0.102 0.450 2.026 2.620 potential_pw2rs 110 12.3 0.008 0.008 2.551 2.569 grid_collocate_task_list 110 9.6 2.116 2.549 2.116 2.549 mp_sum_l 6594 12.7 1.839 2.539 1.839 2.539 mp_alltoall_d11v 2046 13.8 1.795 2.149 1.795 2.149 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.974 1.999 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.793 1.952 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.938 1.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.890 1.904 mp_waitany 10164 13.8 1.290 1.902 1.290 1.902 mp_sum_dm 438 4.9 1.712 1.775 1.712 1.775 mp_irecv_dv 57340 16.2 0.665 1.730 0.665 1.730 md_output 10 3.0 0.000 0.000 0.039 1.721 update_particle_set 20 4.0 0.000 0.000 1.654 1.693 md_write_output 11 3.9 0.038 1.657 0.040 1.678 acc_transpose_blocks 24660 15.4 0.110 0.113 1.556 1.634 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.224 1.026 1.592 mp_allgather_i34 2055 14.4 0.586 1.573 0.586 1.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.554 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.476 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.362000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=551.636364, yerr=6.772575 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 654.831616E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 63.029 63.030 qs_mol_dyn_low 1 2.0 0.003 0.003 62.422 62.432 qs_forces 11 3.9 0.003 0.003 62.357 62.357 qs_energies 11 4.9 0.001 0.001 59.073 59.078 scf_env_do_scf 11 5.9 0.000 0.001 49.340 49.341 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.634 40.635 velocity_verlet 10 3.0 0.001 0.001 33.150 33.151 dbcsr_multiply_generic 2055 12.4 0.107 0.112 30.298 30.633 qs_scf_new_mos 99 7.5 0.001 0.001 25.261 25.388 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.261 25.387 multiply_cannon 2055 13.4 0.213 0.225 23.922 24.974 ot_scf_mini 99 9.5 0.002 0.003 24.032 24.187 multiply_cannon_loop 2055 14.4 0.616 0.631 22.758 23.824 ot_mini 99 10.5 0.001 0.001 13.659 13.814 rebuild_ks_matrix 110 8.3 0.000 0.000 12.608 12.795 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.608 12.795 mp_waitall_1 139946 16.5 7.572 12.181 7.572 12.181 multiply_cannon_multrec 16440 15.4 3.750 4.657 10.474 11.436 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.067 11.239 qs_ot_get_derivative 99 11.5 0.001 0.001 9.125 9.278 init_scf_loop 11 6.9 0.000 0.000 8.668 8.669 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.710 7.821 init_scf_run 11 5.9 0.000 0.001 7.337 7.337 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.337 7.337 dbcsr_mm_accdrv_process 34862 16.1 5.392 6.392 6.580 7.151 prepare_preconditioner 11 7.9 0.000 0.000 6.889 6.907 make_preconditioner 11 8.9 0.000 0.000 6.889 6.907 sum_up_and_integrate 110 10.3 0.060 0.062 6.769 6.784 integrate_v_rspace 110 11.3 0.003 0.003 6.709 6.723 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.220 6.578 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.153 6.232 calculate_rho_elec 110 8.6 0.059 0.059 6.152 6.231 qs_ot_get_p 110 10.4 0.001 0.001 5.563 5.728 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.039 5.488 apply_single 110 13.6 0.000 0.000 5.039 5.488 make_m2s 4110 13.4 0.050 0.052 4.195 4.519 ot_diis_step 99 11.5 0.010 0.011 4.503 4.503 density_rs2pw 110 9.6 0.004 0.005 3.184 4.436 make_images 4110 14.4 0.390 0.512 4.078 4.402 calculate_first_density_matrix 1 7.0 0.000 0.000 4.225 4.229 calculate_dm_sparse 110 9.5 0.001 0.001 4.166 4.197 multiply_cannon_sync_h2d 16440 15.4 3.627 4.082 3.627 4.082 rs_pw_transfer 902 11.9 0.011 0.011 2.781 4.018 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.913 3.917 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.118 3.768 pw_transfer 1331 11.6 0.066 0.073 3.713 3.728 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.606 3.618 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.568 3.568 grid_integrate_task_list 110 12.3 3.279 3.487 3.279 3.487 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.123 3.137 fft_wrap_pw1pw2_140 451 13.1 0.636 0.643 3.092 3.104 wfi_extrapolate 11 7.9 0.001 0.001 3.023 3.023 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.012 3.023 cp_fm_diag_elpa_base 48 14.0 2.946 2.977 3.010 3.021 jit_kernel_multiply 10 16.4 0.801 2.987 0.801 2.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.963 2.964 make_images_data 4110 15.4 0.043 0.046 2.449 2.861 cp_fm_cholesky_invert 11 10.9 2.740 2.746 2.740 2.746 hybrid_alltoall_any 4261 16.3 0.105 0.379 2.138 2.720 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.630 2.691 grid_collocate_task_list 110 9.6 2.143 2.540 2.143 2.540 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.925 2.534 mp_waitany 17072 13.8 1.196 2.483 1.196 2.483 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.379 2.464 fft3d_ps 1111 14.6 1.092 1.103 2.432 2.451 mp_irecv_dv 48980 15.7 0.854 2.412 0.854 2.412 potential_pw2rs 110 12.3 0.011 0.011 2.367 2.379 mp_alltoall_d11v 2046 13.8 1.811 2.286 1.811 2.286 mp_sum_l 6594 12.7 1.566 2.204 1.566 2.204 multiply_cannon_metrocomm1 16440 15.4 0.028 0.029 0.457 2.182 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.179 0.903 2.139 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.956 1.957 dbcsr_complete_redistribute 325 12.2 0.320 0.340 1.482 1.940 acc_transpose_blocks 16440 15.4 0.073 0.075 1.771 1.926 cp_fm_upper_to_full 70 14.2 1.402 1.775 1.402 1.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.735 1.753 cp_fm_cholesky_decompose 22 10.9 1.612 1.630 1.612 1.630 mp_allgather_i34 2055 14.4 0.490 1.576 0.490 1.576 acc_transpose_blocks_kernels 16440 16.4 0.208 0.212 1.344 1.493 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.365 1.483 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.974 1.437 rs_gather_matrices 110 12.3 0.234 0.261 0.983 1.393 jit_kernel_transpose 5 15.6 1.136 1.285 1.136 1.285 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.266 1.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.030000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=619.272727, yerr=7.161652 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 733.069312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.041 80.072 80.073 qs_mol_dyn_low 1 2.0 0.003 0.004 79.561 79.572 qs_forces 11 3.9 0.003 0.003 79.058 79.060 qs_energies 11 4.9 0.003 0.009 75.400 75.406 scf_env_do_scf 11 5.9 0.001 0.002 64.693 64.695 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 49.265 49.266 velocity_verlet 10 3.0 0.001 0.001 44.177 44.197 dbcsr_multiply_generic 2055 12.4 0.116 0.120 33.444 33.829 qs_scf_new_mos 99 7.5 0.001 0.001 31.882 31.988 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.882 31.987 ot_scf_mini 99 9.5 0.003 0.003 30.132 30.235 multiply_cannon 2055 13.4 0.244 0.270 24.738 26.418 multiply_cannon_loop 2055 14.4 0.890 0.912 22.961 23.797 ot_mini 99 10.5 0.001 0.001 16.055 16.194 init_scf_loop 11 6.9 0.001 0.005 15.369 15.369 multiply_cannon_multrec 24660 15.4 4.165 6.513 13.094 14.855 rebuild_ks_matrix 110 8.3 0.000 0.000 13.649 13.765 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 13.648 13.765 prepare_preconditioner 11 7.9 0.000 0.000 13.407 13.423 make_preconditioner 11 8.9 0.000 0.001 13.406 13.423 make_full_inverse_cholesky 11 9.9 0.000 0.001 11.575 13.066 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.042 12.144 qs_ot_get_derivative 99 11.5 0.001 0.001 11.629 11.733 mp_waitall_1 121746 16.5 6.678 10.636 6.678 10.636 dbcsr_mm_accdrv_process 52304 16.0 7.300 9.598 8.784 10.298 qs_ot_get_p 110 10.4 0.001 0.001 8.718 8.878 sum_up_and_integrate 110 10.3 0.067 0.071 7.765 7.778 init_scf_run 11 5.9 0.000 0.001 7.772 7.773 scf_env_initial_rho_setup 11 6.9 0.001 0.002 7.772 7.773 integrate_v_rspace 110 11.3 0.003 0.003 7.697 7.716 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.446 7.520 calculate_rho_elec 110 8.6 0.078 0.082 7.446 7.519 make_m2s 4110 13.4 0.060 0.061 6.573 6.814 make_images 4110 14.4 0.572 0.693 6.430 6.669 qs_ot_p2m_diag 48 11.0 0.055 0.064 6.483 6.504 cp_dbcsr_syevd 48 12.0 0.003 0.005 5.939 5.940 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 2.349 5.854 cp_fm_cholesky_invert 11 10.9 5.140 5.153 5.140 5.153 cp_fm_diag_elpa 48 13.0 0.000 0.001 5.065 5.074 cp_fm_diag_elpa_base 48 14.0 4.889 4.947 5.060 5.071 cp_fm_upper_to_full 70 14.2 3.453 5.003 3.453 5.003 density_rs2pw 110 9.6 0.004 0.005 4.217 4.740 calculate_dm_sparse 110 9.5 0.001 0.001 4.708 4.735 pw_transfer 1331 11.6 0.066 0.076 4.476 4.564 rs_pw_transfer 902 11.9 0.010 0.011 3.974 4.500 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.369 4.461 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.121 4.353 apply_single 110 13.6 0.000 0.000 4.121 4.353 ot_diis_step 99 11.5 0.011 0.011 4.317 4.318 dbcsr_complete_redistribute 325 12.2 0.408 0.454 3.023 4.254 calculate_first_density_matrix 1 7.0 0.001 0.002 4.121 4.127 make_images_data 4110 15.4 0.045 0.049 3.227 3.897 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.737 3.785 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.682 3.743 fft_wrap_pw1pw2_140 451 13.1 0.667 0.691 3.546 3.636 grid_integrate_task_list 110 12.3 3.345 3.538 3.345 3.538 wfi_extrapolate 11 7.9 0.001 0.001 3.526 3.526 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.299 3.518 multiply_cannon_sync_h2d 24660 15.4 3.166 3.402 3.166 3.402 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.217 3.237 hybrid_alltoall_any 4261 16.3 0.120 0.457 2.630 3.221 potential_pw2rs 110 12.3 0.012 0.013 3.173 3.219 fft3d_ps 1111 14.6 1.092 1.128 3.131 3.198 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.822 3.021 mp_alltoall_i22 605 13.7 1.732 3.000 1.732 3.000 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.951 2.954 jit_kernel_multiply 10 15.8 1.154 2.824 1.154 2.824 mp_alltoall_d11v 2046 13.8 2.292 2.807 2.292 2.807 cp_fm_cholesky_decompose 22 10.9 2.569 2.620 2.569 2.620 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.501 2.549 grid_collocate_task_list 110 9.6 2.241 2.475 2.241 2.475 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.308 2.310 acc_transpose_blocks 24660 15.4 0.103 0.105 1.753 2.182 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.982 2.044 mp_sum_l 6594 12.7 1.457 1.936 1.457 1.936 mp_alltoall_z22v 1111 16.6 1.821 1.925 1.821 1.925 mp_waitany 13376 13.8 1.397 1.895 1.397 1.895 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.853 1.867 multiply_cannon_metrocomm1 24660 15.4 0.035 0.036 0.617 1.857 multiply_cannon_metrocomm4 20550 15.4 0.059 0.062 0.856 1.735 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.596 1.698 mp_allgather_i34 2055 14.4 0.878 1.698 0.878 1.698 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=80.073000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=693.727273, yerr=10.643743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 828.780544E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.036 55.563 55.564 qs_mol_dyn_low 1 2.0 0.003 0.003 55.299 55.310 qs_forces 11 3.9 0.003 0.003 55.194 55.197 qs_energies 11 4.9 0.001 0.001 51.464 51.469 scf_env_do_scf 11 5.9 0.000 0.001 43.228 43.228 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 35.520 35.521 velocity_verlet 10 3.0 0.001 0.001 31.325 31.330 dbcsr_multiply_generic 2055 12.4 0.104 0.110 22.814 22.983 qs_scf_new_mos 99 7.5 0.001 0.001 20.331 20.392 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.331 20.392 ot_scf_mini 99 9.5 0.002 0.002 19.099 19.145 multiply_cannon 2055 13.4 0.243 0.264 17.409 18.503 multiply_cannon_loop 2055 14.4 0.320 0.331 16.082 16.369 rebuild_ks_matrix 110 8.3 0.000 0.000 12.065 12.108 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.064 12.108 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.675 10.713 ot_mini 99 10.5 0.001 0.001 10.456 10.497 multiply_cannon_multrec 8220 15.4 3.204 4.849 7.421 8.635 mp_waitall_1 103326 16.6 6.126 7.774 6.126 7.774 init_scf_loop 11 6.9 0.000 0.000 7.661 7.662 sum_up_and_integrate 110 10.3 0.080 0.081 6.694 6.708 qs_ot_get_derivative 99 11.5 0.001 0.001 6.600 6.645 integrate_v_rspace 110 11.3 0.003 0.003 6.614 6.627 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.339 6.355 calculate_rho_elec 110 8.6 0.114 0.114 6.339 6.354 prepare_preconditioner 11 7.9 0.000 0.000 5.990 5.996 make_preconditioner 11 8.9 0.000 0.000 5.990 5.996 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.569 5.641 init_scf_run 11 5.9 0.000 0.001 5.107 5.107 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.106 5.106 dbcsr_mm_accdrv_process 17442 15.9 2.899 3.930 4.087 4.996 qs_ot_get_p 110 10.4 0.001 0.001 4.738 4.793 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.187 4.484 make_m2s 4110 13.4 0.039 0.040 4.119 4.365 make_images 4110 14.4 0.635 0.687 3.988 4.234 pw_transfer 1331 11.6 0.066 0.071 3.930 3.937 ot_diis_step 99 11.5 0.012 0.012 3.837 3.837 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.822 3.834 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.741 3.820 apply_single 110 13.6 0.000 0.000 3.741 3.819 grid_integrate_task_list 110 12.3 3.455 3.670 3.455 3.670 density_rs2pw 110 9.6 0.004 0.004 3.027 3.403 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.383 3.387 fft_wrap_pw1pw2_140 451 13.1 0.834 0.846 3.299 3.316 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.087 3.088 multiply_cannon_sync_h2d 8220 15.4 2.896 3.014 2.896 3.014 cp_fm_cholesky_invert 11 10.9 2.916 2.919 2.916 2.919 wfi_extrapolate 11 7.9 0.001 0.001 2.748 2.748 make_images_data 4110 15.4 0.038 0.043 2.336 2.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.672 2.674 hybrid_alltoall_any 4261 16.3 0.200 0.864 2.262 2.621 rs_pw_transfer 902 11.9 0.010 0.010 2.209 2.616 grid_collocate_task_list 110 9.6 2.336 2.592 2.336 2.592 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.568 2.575 cp_fm_diag_elpa_base 48 14.0 2.514 2.531 2.566 2.574 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.520 2.522 calculate_dm_sparse 110 9.5 0.001 0.001 2.448 2.495 fft3d_ps 1111 14.6 1.133 1.157 2.366 2.378 calculate_first_density_matrix 1 7.0 0.000 0.000 2.263 2.264 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.177 2.191 potential_pw2rs 110 12.3 0.015 0.015 2.094 2.099 mp_alltoall_d11v 2046 13.8 1.668 1.995 1.668 1.995 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.978 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.912 1.937 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.787 1.812 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.719 1.725 cp_fm_cholesky_decompose 22 10.9 1.636 1.646 1.636 1.646 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.513 1.642 mp_allgather_i34 2055 14.4 0.514 1.567 0.514 1.567 dbcsr_complete_redistribute 325 12.2 0.559 0.592 1.434 1.526 mp_waitany 9240 13.8 1.084 1.513 1.084 1.513 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.432 qs_create_task_list 11 7.9 0.001 0.001 1.229 1.332 generate_qs_task_list 11 8.9 0.375 0.444 1.228 1.331 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.486 1.299 rs_gather_matrices 110 12.3 0.328 0.370 0.974 1.282 mp_irecv_dv 24056 15.7 0.463 1.258 0.463 1.258 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.805 1.212 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.119 1.154 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.564000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=778.272727, yerr=12.113786 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 1.391714E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.039 90.220 90.221 qs_mol_dyn_low 1 2.0 0.003 0.003 89.859 89.868 qs_forces 11 3.9 0.003 0.004 89.782 89.783 qs_energies 11 4.9 0.001 0.001 85.529 85.532 scf_env_do_scf 11 5.9 0.001 0.001 75.330 75.331 velocity_verlet 10 3.0 0.001 0.001 57.695 57.702 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 45.719 45.720 init_scf_loop 11 6.9 0.000 0.000 29.533 29.535 dbcsr_multiply_generic 2055 12.4 0.119 0.124 28.935 29.012 prepare_preconditioner 11 7.9 0.000 0.000 27.464 27.473 make_preconditioner 11 8.9 0.000 0.000 27.464 27.473 qs_scf_new_mos 99 7.5 0.001 0.001 27.022 27.067 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.022 27.066 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.395 26.943 ot_scf_mini 99 9.5 0.002 0.002 25.239 25.272 multiply_cannon 2055 13.4 0.344 0.373 21.746 22.620 multiply_cannon_loop 2055 14.4 0.340 0.343 19.914 20.206 cp_fm_upper_to_full 70 14.2 13.034 18.838 13.034 18.838 rebuild_ks_matrix 110 8.3 0.001 0.001 14.445 14.494 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.444 14.493 ot_mini 99 10.5 0.001 0.001 13.819 13.850 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.044 13.088 dbcsr_complete_redistribute 325 12.2 1.022 1.049 7.698 11.305 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.650 10.255 multiply_cannon_multrec 8220 15.4 4.370 4.548 9.547 9.631 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.054 9.619 mp_alltoall_i22 605 13.7 5.693 9.305 5.693 9.305 qs_ot_get_derivative 99 11.5 0.001 0.001 9.208 9.241 mp_waitall_1 84994 16.7 7.807 8.707 7.807 8.707 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.957 7.993 calculate_rho_elec 110 8.6 0.224 0.225 7.957 7.993 sum_up_and_integrate 110 10.3 0.150 0.151 7.588 7.603 integrate_v_rspace 110 11.3 0.004 0.004 7.438 7.453 qs_ot_get_p 110 10.4 0.001 0.001 6.129 6.158 init_scf_run 11 5.9 0.000 0.001 5.891 5.891 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.891 5.891 make_m2s 4110 13.4 0.043 0.044 5.363 5.877 cp_fm_cholesky_invert 11 10.9 5.697 5.702 5.697 5.702 make_images 4110 14.4 0.882 0.939 5.175 5.688 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.976 5.286 dbcsr_mm_accdrv_process 11614 15.7 3.264 3.617 5.034 5.285 pw_transfer 1331 11.6 0.075 0.075 5.230 5.238 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.705 5.180 apply_single 110 13.6 0.000 0.000 4.705 5.180 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.112 5.121 ot_diis_step 99 11.5 0.015 0.015 4.585 4.585 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.490 4.497 fft_wrap_pw1pw2_140 451 13.1 1.337 1.343 4.421 4.427 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.039 4.040 multiply_cannon_sync_h2d 8220 15.4 3.948 3.952 3.948 3.952 density_rs2pw 110 9.6 0.004 0.004 3.889 3.924 grid_integrate_task_list 110 12.3 3.812 3.874 3.812 3.874 qs_energies_init_hamiltonians 11 5.9 0.011 0.016 3.684 3.685 hybrid_alltoall_any 4261 16.3 0.256 0.551 2.929 3.618 make_images_data 4110 15.4 0.041 0.044 2.905 3.595 wfi_extrapolate 11 7.9 0.001 0.001 3.439 3.439 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.403 3.403 cp_fm_diag_elpa_base 48 14.0 2.849 3.059 3.400 3.400 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.937 3.382 calculate_dm_sparse 110 9.5 0.001 0.001 3.152 3.179 fft3d_ps 1111 14.6 1.308 1.331 3.042 3.054 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.928 2.931 grid_collocate_task_list 110 9.6 2.705 2.727 2.705 2.727 potential_pw2rs 110 12.3 0.021 0.021 2.520 2.522 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.464 2.480 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.411 2.421 rs_pw_transfer 902 11.9 0.010 0.011 2.377 2.401 calculate_first_density_matrix 1 7.0 0.000 0.000 2.328 2.328 qs_env_update_s_mstruct 11 6.9 0.002 0.009 2.189 2.246 cp_fm_cholesky_decompose 22 10.9 2.197 2.212 2.197 2.212 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.089 2.158 mp_alltoall_d11v 2046 13.8 2.028 2.157 2.028 2.157 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.006 2.012 qs_create_task_list 11 7.9 0.001 0.001 1.895 1.941 generate_qs_task_list 11 8.9 0.734 0.785 1.894 1.940 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.779 1.815 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.221000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1226.454545, yerr=67.398899 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.460992E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2790668. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.030 205.377 205.378 qs_mol_dyn_low 1 2.0 0.003 0.003 205.008 205.021 qs_forces 11 3.9 0.005 0.005 204.920 204.921 qs_energies 11 4.9 0.001 0.001 199.288 199.299 scf_env_do_scf 11 5.9 0.001 0.001 182.705 182.709 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 162.367 162.369 dbcsr_multiply_generic 2507 12.6 0.179 0.183 123.767 124.280 velocity_verlet 10 3.0 0.001 0.002 122.797 122.798 qs_scf_new_mos 117 7.6 0.001 0.001 122.282 122.507 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.281 122.506 ot_scf_mini 117 9.6 0.003 0.003 115.740 115.957 multiply_cannon 2507 13.6 0.239 0.248 100.834 102.766 multiply_cannon_loop 2507 14.6 2.063 2.126 98.750 100.792 ot_mini 117 10.6 0.001 0.001 65.199 65.463 multiply_cannon_multrec 60168 15.6 33.455 35.600 41.626 43.877 qs_ot_get_derivative 117 11.6 0.001 0.001 40.460 40.681 rebuild_ks_matrix 128 8.3 0.001 0.001 33.794 34.138 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.016 33.794 34.137 mp_waitall_1 267128 16.5 27.833 30.917 27.833 30.917 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.288 30.602 multiply_cannon_sync_h2d 60168 15.6 27.680 29.322 27.680 29.322 qs_ot_get_p 128 10.4 0.001 0.001 28.715 29.106 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.280 25.054 apply_single 128 13.6 0.001 0.001 24.280 25.054 ot_diis_step 117 11.6 0.007 0.008 24.477 24.478 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.179 22.267 init_scf_loop 11 6.9 0.000 0.001 20.266 20.267 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.683 19.684 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.252 19.403 multiply_cannon_metrocomm3 60168 15.6 0.113 0.119 15.526 17.426 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.802 16.833 cp_fm_diag_elpa_base 83 14.4 16.735 16.770 16.798 16.830 prepare_preconditioner 11 7.9 0.000 0.000 15.656 15.695 make_preconditioner 11 8.9 0.000 0.000 15.656 15.695 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.913 15.076 sum_up_and_integrate 128 10.3 0.089 0.106 14.347 14.363 integrate_v_rspace 128 11.3 0.004 0.005 14.258 14.277 make_m2s 5014 13.6 0.104 0.114 13.654 14.135 make_images 5014 14.6 0.400 0.418 13.472 13.960 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.756 13.869 calculate_rho_elec 128 8.7 0.046 0.065 13.756 13.869 init_scf_run 11 5.9 0.000 0.001 12.413 12.414 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.413 12.414 density_rs2pw 128 9.7 0.006 0.007 6.969 10.307 mp_sum_l 7950 12.9 8.010 9.262 8.010 9.262 wfi_extrapolate 11 7.9 0.001 0.001 9.058 9.058 cp_fm_cholesky_invert 11 10.9 8.939 8.947 8.939 8.947 rs_pw_transfer 1046 11.9 0.017 0.021 5.584 8.947 calculate_dm_sparse 128 9.5 0.001 0.001 8.345 8.435 dbcsr_mm_accdrv_process 124484 16.2 3.065 3.200 7.729 8.310 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.840 7.979 multiply_cannon_metrocomm1 60168 15.6 0.087 0.091 5.969 7.970 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.817 7.900 pw_transfer 1547 11.6 0.074 0.091 7.698 7.863 grid_integrate_task_list 128 12.3 7.271 7.783 7.271 7.783 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 7.495 7.659 make_images_data 5014 15.6 0.066 0.073 6.696 7.646 hybrid_alltoall_any 5200 16.5 0.289 2.255 5.896 7.360 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.770 6.782 fft_wrap_pw1pw2_140 523 13.2 1.277 1.327 6.605 6.776 grid_collocate_task_list 128 9.7 4.728 6.032 4.728 6.032 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.906 6.023 mp_waitany 16020 13.9 2.655 5.983 2.655 5.983 mp_alltoall_d11v 2415 14.1 4.214 5.808 4.214 5.808 fft3d_ps 1291 14.7 2.143 2.794 5.261 5.562 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.291 2.138 5.493 cp_fm_cholesky_decompose 22 10.9 4.575 4.588 4.575 4.588 potential_pw2rs 128 12.3 0.009 0.010 4.475 4.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.378000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.000000, yerr=7.058200 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 834.338816E+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 4002 58220. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6383282. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.061 200.852 200.855 qs_mol_dyn_low 1 2.0 0.004 0.009 200.325 200.341 qs_forces 11 3.9 0.007 0.023 200.220 200.222 qs_energies 11 4.9 0.002 0.008 193.168 193.187 scf_env_do_scf 11 5.9 0.001 0.001 175.789 175.799 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 140.222 140.226 velocity_verlet 10 3.0 0.001 0.002 125.974 125.977 dbcsr_multiply_generic 2507 12.6 0.191 0.209 100.063 101.581 qs_scf_new_mos 117 7.6 0.001 0.001 98.584 99.096 qs_scf_loop_do_ot 117 8.6 0.001 0.002 98.583 99.095 ot_scf_mini 117 9.6 0.004 0.005 93.636 94.180 multiply_cannon 2507 13.6 0.477 0.539 78.251 82.744 multiply_cannon_loop 2507 14.6 1.243 1.279 74.774 78.111 ot_mini 117 10.6 0.001 0.001 51.694 52.232 mp_waitall_1 214728 16.6 27.023 41.704 27.023 41.704 multiply_cannon_multrec 30084 15.6 21.271 25.953 31.053 35.937 init_scf_loop 11 6.9 0.001 0.006 35.469 35.477 rebuild_ks_matrix 128 8.3 0.001 0.001 34.409 35.006 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.035 34.409 35.005 multiply_cannon_metrocomm3 30084 15.6 0.091 0.098 17.321 32.016 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.879 31.429 prepare_preconditioner 11 7.9 0.000 0.001 30.910 30.992 make_preconditioner 11 8.9 0.000 0.002 30.910 30.992 make_full_inverse_cholesky 11 9.9 0.000 0.001 29.542 30.140 qs_ot_get_derivative 117 11.6 0.001 0.002 29.271 29.814 qs_ot_get_p 128 10.4 0.001 0.003 23.430 24.192 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.374 23.677 apply_single 128 13.6 0.001 0.001 22.374 23.677 ot_diis_step 117 11.6 0.014 0.015 22.228 22.229 multiply_cannon_sync_h2d 30084 15.6 18.523 20.572 18.523 20.572 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.265 18.316 cp_fm_cholesky_invert 11 10.9 18.277 18.290 18.277 18.290 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.024 17.025 make_m2s 5014 13.6 0.090 0.096 15.225 16.653 make_images 5014 14.6 1.148 1.333 15.013 16.434 sum_up_and_integrate 128 10.3 0.116 0.133 15.572 15.604 integrate_v_rspace 128 11.3 0.004 0.005 15.456 15.493 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.901 14.955 calculate_rho_elec 128 8.7 0.088 0.104 14.900 14.954 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.673 13.699 cp_fm_diag_elpa_base 83 14.4 13.386 13.490 13.664 13.692 init_scf_run 11 5.9 0.000 0.001 12.459 12.461 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.459 12.461 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.692 12.132 multiply_cannon_metrocomm4 27577 15.6 0.103 0.118 3.953 10.954 make_images_data 5014 15.6 0.064 0.072 9.296 10.929 mp_irecv_dv 69486 16.3 3.751 10.556 3.751 10.556 density_rs2pw 128 9.7 0.006 0.007 7.990 10.453 hybrid_alltoall_any 5200 16.5 0.341 1.514 7.941 10.266 dbcsr_mm_accdrv_process 62242 16.2 4.526 5.297 9.269 9.861 pw_transfer 1547 11.6 0.087 0.097 9.360 9.406 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.134 9.186 wfi_extrapolate 11 7.9 0.001 0.001 8.773 8.774 rs_pw_transfer 1046 11.9 0.014 0.016 6.299 8.757 fft_wrap_pw1pw2_140 523 13.2 1.335 1.362 7.955 8.024 grid_integrate_task_list 128 12.3 7.392 7.811 7.392 7.811 cp_fm_cholesky_decompose 22 10.9 7.684 7.762 7.684 7.762 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.420 7.206 mp_sum_l 7950 12.9 4.822 7.171 4.822 7.171 calculate_dm_sparse 128 9.5 0.001 0.001 6.754 6.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.510 6.519 fft3d_ps 1291 14.7 2.816 2.980 6.444 6.488 grid_collocate_task_list 128 9.7 4.831 5.993 4.831 5.993 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.490 5.657 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.582 5.653 mp_allgather_i34 2507 14.6 2.055 5.534 2.055 5.534 potential_pw2rs 128 12.3 0.015 0.017 5.436 5.465 mp_alltoall_d11v 2415 14.1 4.693 5.448 4.693 5.448 mp_waitany 11748 13.9 2.682 5.214 2.682 5.214 rs_pw_transfer_RS2PW_140 139 11.5 0.349 0.372 2.161 4.594 mp_sum_d 4470 12.1 2.977 4.453 2.977 4.453 dbcsr_complete_redistribute 395 12.7 0.771 0.846 3.405 4.267 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=200.855000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=791.454545, yerr=6.415374 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 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 931.090432E+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 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58169. MP_Allreduce 11090 999. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 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 0.014 0.030 197.015 197.017 qs_mol_dyn_low 1 2.0 0.003 0.003 196.630 196.643 qs_forces 11 3.9 0.004 0.005 196.299 196.303 qs_energies 11 4.9 0.001 0.001 189.390 189.403 scf_env_do_scf 11 5.9 0.001 0.001 172.947 172.949 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 130.733 130.734 velocity_verlet 10 3.0 0.001 0.002 126.649 126.658 qs_scf_new_mos 116 7.6 0.001 0.001 89.838 90.240 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.837 90.239 dbcsr_multiply_generic 2485 12.5 0.187 0.193 84.825 86.913 ot_scf_mini 116 9.6 0.003 0.004 85.584 86.007 multiply_cannon 2485 13.5 0.495 0.520 62.662 70.689 multiply_cannon_loop 2485 14.5 0.851 0.876 59.405 64.286 ot_mini 116 10.6 0.001 0.001 45.432 45.876 init_scf_loop 11 6.9 0.000 0.000 42.101 42.102 prepare_preconditioner 11 7.9 0.000 0.000 37.849 37.911 make_preconditioner 11 8.9 0.000 0.000 37.849 37.911 mp_waitall_1 169034 16.6 28.034 37.714 28.034 37.714 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.389 36.877 rebuild_ks_matrix 127 8.3 0.001 0.001 32.381 33.013 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 32.381 33.013 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.093 29.684 multiply_cannon_metrocomm3 19880 15.5 0.058 0.063 16.664 28.663 qs_ot_get_derivative 116 11.6 0.001 0.002 25.270 25.693 qs_ot_get_p 127 10.4 0.001 0.001 25.165 25.669 multiply_cannon_multrec 19880 15.5 13.092 16.228 21.646 24.807 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.084 21.516 apply_single 127 13.6 0.001 0.001 20.084 21.515 qs_ot_p2m_diag 82 11.4 0.263 0.270 20.396 20.459 ot_diis_step 116 11.6 0.017 0.018 19.979 19.979 cp_fm_cholesky_invert 11 10.9 19.377 19.396 19.377 19.396 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.183 19.188 make_m2s 4970 13.5 0.080 0.086 15.760 16.306 sum_up_and_integrate 127 10.3 0.132 0.144 16.158 16.185 make_images 4970 14.5 1.154 1.242 15.529 16.071 integrate_v_rspace 127 11.3 0.004 0.004 16.026 16.049 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.753 15.782 calculate_rho_elec 127 8.7 0.129 0.144 15.752 15.781 multiply_cannon_sync_h2d 19880 15.5 13.854 15.552 13.854 15.552 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.357 15.373 cp_fm_diag_elpa_base 82 14.4 14.896 15.051 15.353 15.369 init_scf_run 11 5.9 0.000 0.001 11.328 11.329 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.327 11.328 make_images_data 4970 15.5 0.058 0.066 9.692 10.537 density_rs2pw 127 9.7 0.006 0.006 8.465 10.461 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.081 10.376 hybrid_alltoall_any 5155 16.4 0.431 1.971 8.371 10.040 pw_transfer 1535 11.6 0.086 0.104 9.496 9.667 multiply_cannon_metrocomm4 17395 15.5 0.061 0.071 3.528 9.590 fft_wrap_pw1pw2 1281 12.7 0.011 0.012 9.272 9.450 mp_irecv_dv 49801 16.2 3.408 9.342 3.408 9.342 cp_fm_cholesky_decompose 22 10.9 9.096 9.134 9.096 9.134 rs_pw_transfer 1038 11.9 0.013 0.014 7.070 9.016 dbcsr_mm_accdrv_process 41158 16.2 4.276 5.176 8.042 8.180 fft_wrap_pw1pw2_140 519 13.2 1.397 1.422 7.953 8.142 wfi_extrapolate 11 7.9 0.001 0.001 8.098 8.098 grid_integrate_task_list 127 12.3 7.431 8.059 7.431 8.059 cp_fm_upper_to_full 104 14.8 5.917 7.563 5.917 7.563 dbcsr_complete_redistribute 393 12.7 1.195 1.220 4.972 6.906 mp_sum_l 7884 12.9 4.599 6.563 4.599 6.563 fft3d_ps 1281 14.7 2.713 2.932 6.421 6.545 mp_alltoall_d11v 2401 14.1 4.886 6.128 4.886 6.128 potential_pw2rs 127 12.3 0.020 0.022 5.889 5.970 calculate_dm_sparse 127 9.5 0.001 0.001 5.875 5.967 grid_collocate_task_list 127 9.7 4.986 5.901 4.986 5.901 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.076 5.735 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.705 5.716 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.659 5.591 mp_allgather_i34 2485 14.5 1.793 4.964 1.793 4.964 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.650 4.874 mp_waitany 11660 13.9 2.594 4.708 2.594 4.708 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.400 4.435 mp_alltoall_i22 712 14.1 2.210 4.380 2.210 4.380 transfer_fm_to_dbcsr 11 9.9 0.020 0.028 2.441 4.357 rs_pw_transfer_RS2PW_140 138 11.5 0.328 0.350 1.986 3.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=197.017000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=885.000000, yerr=6.660603 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.134993E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57954. MP_Allreduce 11127 1081. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.032 190.252 190.253 qs_mol_dyn_low 1 2.0 0.003 0.003 189.878 189.891 qs_forces 11 3.9 0.004 0.005 189.761 189.768 qs_energies 11 4.9 0.002 0.013 182.466 182.475 scf_env_do_scf 11 5.9 0.001 0.003 165.736 165.745 velocity_verlet 10 3.0 0.001 0.002 125.996 125.999 scf_env_do_scf_inner_loop 117 6.6 0.003 0.013 119.024 119.025 qs_scf_new_mos 117 7.6 0.001 0.001 81.004 81.349 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.003 81.348 dbcsr_multiply_generic 2507 12.6 0.187 0.193 79.067 79.841 ot_scf_mini 117 9.6 0.003 0.004 76.522 76.812 multiply_cannon 2507 13.6 0.556 0.598 54.518 57.639 multiply_cannon_loop 2507 14.6 1.180 1.211 50.884 52.724 init_scf_loop 11 6.9 0.001 0.004 46.587 46.589 prepare_preconditioner 11 7.9 0.000 0.000 42.411 42.447 make_preconditioner 11 8.9 0.000 0.002 42.411 42.447 ot_mini 117 10.6 0.001 0.001 42.051 42.348 make_full_inverse_cholesky 11 9.9 0.012 0.021 36.048 41.021 multiply_cannon_multrec 30084 15.6 14.092 19.237 25.988 30.712 rebuild_ks_matrix 128 8.3 0.001 0.001 30.417 30.710 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 30.417 30.710 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.351 27.613 mp_waitall_1 147882 16.7 17.225 27.017 17.225 27.017 qs_ot_get_derivative 117 11.6 0.001 0.002 22.473 22.776 make_m2s 5014 13.6 0.095 0.099 20.196 21.173 qs_ot_get_p 128 10.4 0.001 0.001 20.659 20.969 make_images 5014 14.6 1.922 2.238 19.890 20.869 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.954 19.566 apply_single 128 13.6 0.001 0.001 18.953 19.565 ot_diis_step 117 11.6 0.017 0.019 19.453 19.455 cp_fm_upper_to_full 105 14.8 11.462 16.829 11.462 16.829 qs_ot_p2m_diag 83 11.4 0.343 0.390 16.592 16.644 cp_fm_cholesky_invert 11 10.9 16.293 16.302 16.293 16.302 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.308 15.310 sum_up_and_integrate 128 10.3 0.140 0.153 15.104 15.131 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.412 14.998 integrate_v_rspace 128 11.3 0.004 0.004 14.964 14.996 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.734 14.771 calculate_rho_elec 128 8.7 0.173 0.189 14.734 14.771 dbcsr_complete_redistribute 395 12.7 1.504 1.614 9.120 12.922 multiply_cannon_sync_h2d 30084 15.6 11.791 12.602 11.791 12.602 make_images_data 5014 15.6 0.061 0.066 10.898 12.445 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.011 12.027 cp_fm_diag_elpa_base 83 14.4 11.037 11.345 12.005 12.019 dbcsr_mm_accdrv_process 62264 16.2 7.351 8.109 11.466 11.954 hybrid_alltoall_any 5200 16.5 0.517 2.090 9.626 11.617 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.721 11.422 init_scf_run 11 5.9 0.000 0.001 10.838 10.839 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.837 10.839 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.340 9.969 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.303 9.524 mp_alltoall_i22 716 14.1 5.580 9.431 5.580 9.431 pw_transfer 1547 11.6 0.086 0.102 9.242 9.328 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.016 9.108 density_rs2pw 128 9.7 0.006 0.006 7.054 8.476 grid_integrate_task_list 128 12.3 7.768 8.158 7.768 8.158 fft_wrap_pw1pw2_140 523 13.2 1.557 1.586 7.993 8.096 wfi_extrapolate 11 7.9 0.001 0.001 7.698 7.699 cp_fm_cholesky_decompose 22 10.9 7.547 7.664 7.547 7.664 multiply_cannon_metrocomm4 25070 15.6 0.079 0.090 2.765 7.567 mp_irecv_dv 76098 16.2 2.617 7.289 2.617 7.289 rs_pw_transfer 1046 11.9 0.014 0.019 4.771 6.270 calculate_dm_sparse 128 9.5 0.001 0.001 6.157 6.228 mp_alltoall_d11v 2415 14.1 5.058 6.048 5.058 6.048 fft3d_ps 1291 14.7 2.848 2.940 5.944 6.010 grid_collocate_task_list 128 9.7 5.194 5.924 5.194 5.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.473 5.530 potential_pw2rs 128 12.3 0.024 0.024 4.626 4.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.411 4.483 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.451 4.451 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.333 4.418 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.196 4.251 mp_allgather_i34 2507 14.6 1.735 4.006 1.735 4.006 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.253000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1072.272727, yerr=19.207437 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.492705E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57953. MP_Allreduce 11127 1165. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.038 176.089 176.089 qs_mol_dyn_low 1 2.0 0.003 0.003 175.649 175.661 qs_forces 11 3.9 0.004 0.005 175.533 175.541 qs_energies 11 4.9 0.001 0.001 167.832 167.840 scf_env_do_scf 11 5.9 0.001 0.001 150.185 150.191 velocity_verlet 10 3.0 0.001 0.001 115.239 115.243 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 114.574 114.576 qs_scf_new_mos 117 7.6 0.001 0.001 75.665 75.731 qs_scf_loop_do_ot 117 8.6 0.001 0.001 75.665 75.730 dbcsr_multiply_generic 2507 12.6 0.184 0.197 74.752 75.232 ot_scf_mini 117 9.6 0.003 0.004 71.230 71.314 multiply_cannon 2507 13.6 0.584 0.612 54.827 59.868 multiply_cannon_loop 2507 14.6 0.444 0.457 50.129 51.066 ot_mini 117 10.6 0.001 0.001 39.991 40.060 init_scf_loop 11 6.9 0.000 0.000 35.462 35.464 mp_waitall_1 125778 16.7 26.351 33.788 26.351 33.788 prepare_preconditioner 11 7.9 0.000 0.000 31.485 31.507 make_preconditioner 11 8.9 0.000 0.000 31.485 31.507 rebuild_ks_matrix 128 8.3 0.001 0.001 30.209 30.315 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.209 30.314 make_full_inverse_cholesky 11 9.9 0.019 0.030 29.423 29.677 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.334 27.428 multiply_cannon_multrec 10028 15.6 10.428 15.478 17.697 22.055 ot_diis_step 117 11.6 0.020 0.020 20.160 20.160 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.732 20.061 apply_single 128 13.6 0.001 0.001 19.732 20.061 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.775 19.986 qs_ot_get_derivative 117 11.6 0.002 0.002 19.761 19.848 make_m2s 5014 13.6 0.068 0.073 16.240 18.639 qs_ot_get_p 128 10.4 0.001 0.001 18.241 18.353 make_images 5014 14.6 2.304 2.636 15.932 18.321 cp_fm_cholesky_invert 11 10.9 17.984 17.990 17.984 17.990 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.508 15.573 calculate_rho_elec 128 8.7 0.256 0.267 15.507 15.573 sum_up_and_integrate 128 10.3 0.181 0.191 15.294 15.344 integrate_v_rspace 128 11.3 0.004 0.004 15.113 15.170 qs_ot_p2m_diag 83 11.4 0.496 0.502 14.458 14.474 cp_dbcsr_syevd 83 12.4 0.005 0.005 13.321 13.322 make_images_data 5014 15.6 0.052 0.061 9.855 12.470 hybrid_alltoall_any 5200 16.5 0.789 3.617 9.700 12.235 multiply_cannon_sync_h2d 10028 15.6 11.546 12.164 11.546 12.164 init_scf_run 11 5.9 0.000 0.001 10.740 10.740 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.739 10.740 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.246 10.257 cp_fm_diag_elpa_base 83 14.4 10.006 10.083 10.239 10.250 pw_transfer 1547 11.6 0.086 0.096 9.558 9.580 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.334 9.361 grid_integrate_task_list 128 12.3 8.035 8.402 8.035 8.402 fft_wrap_pw1pw2_140 523 13.2 1.916 1.968 8.245 8.275 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.947 8.003 cp_fm_cholesky_decompose 22 10.9 7.808 7.938 7.808 7.938 mp_allgather_i34 2507 14.6 2.908 7.871 2.908 7.871 density_rs2pw 128 9.7 0.006 0.006 6.938 7.841 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.431 7.716 wfi_extrapolate 11 7.9 0.001 0.001 7.708 7.709 dbcsr_mm_accdrv_process 20762 16.1 2.589 3.308 6.900 7.535 grid_collocate_task_list 128 9.7 5.515 6.188 5.515 6.188 calculate_dm_sparse 128 9.5 0.001 0.001 6.107 6.184 mp_alltoall_d11v 2415 14.1 5.097 5.984 5.097 5.984 fft3d_ps 1291 14.7 2.760 2.828 5.810 5.847 dbcsr_complete_redistribute 395 12.7 2.130 2.192 5.212 5.657 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.334 5.378 multiply_cannon_metrocomm4 7521 15.6 0.023 0.027 1.885 5.353 qs_energies_init_hamiltonians 11 5.9 0.007 0.023 5.347 5.349 mp_irecv_dv 28860 15.9 1.848 5.267 1.848 5.267 rs_pw_transfer 1046 11.9 0.012 0.013 4.167 5.108 potential_pw2rs 128 12.3 0.026 0.026 4.431 4.445 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.203 4.242 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.858 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.756 3.798 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.777 3.788 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.377 3.703 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.611 3.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=176.089000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1401.636364, yerr=45.054054 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.090395E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108638608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 59213. MP_Allreduce 11070 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.071 300.118 300.118 qs_mol_dyn_low 1 2.0 0.003 0.003 299.537 299.550 qs_forces 11 3.9 0.022 0.023 299.438 299.441 qs_energies 11 4.9 0.001 0.001 290.008 290.016 scf_env_do_scf 11 5.9 0.001 0.001 267.643 267.662 velocity_verlet 10 3.0 0.001 0.001 215.104 215.112 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.644 141.646 init_scf_loop 11 6.9 0.000 0.000 125.748 125.751 prepare_preconditioner 11 7.9 0.000 0.000 120.642 120.677 make_preconditioner 11 8.9 0.000 0.000 120.642 120.677 make_full_inverse_cholesky 11 9.9 0.035 0.039 96.095 117.884 qs_scf_new_mos 116 7.6 0.001 0.001 91.137 91.300 qs_scf_loop_do_ot 116 8.6 0.001 0.001 91.136 91.299 ot_scf_mini 116 9.6 0.004 0.004 86.330 86.431 dbcsr_multiply_generic 2485 12.5 0.210 0.222 83.394 84.309 cp_fm_upper_to_full 104 14.8 53.398 77.030 53.398 77.030 multiply_cannon 2485 13.5 0.702 0.765 58.652 59.707 multiply_cannon_loop 2485 14.5 0.466 0.471 55.050 57.048 ot_mini 116 10.6 0.001 0.001 44.573 44.699 dbcsr_complete_redistribute 393 12.7 3.992 4.031 30.205 43.698 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.838 40.323 rebuild_ks_matrix 127 8.3 0.001 0.001 38.348 38.506 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.348 38.506 transfer_fm_to_dbcsr 11 9.9 0.030 0.034 24.502 37.853 mp_alltoall_i22 712 14.1 22.339 35.644 22.339 35.644 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.267 35.423 cp_fm_cholesky_invert 11 10.9 33.578 33.585 33.578 33.585 mp_waitall_1 102768 16.8 28.732 33.242 28.732 33.242 qs_ot_get_p 127 10.4 0.001 0.001 26.287 26.397 qs_ot_get_derivative 116 11.6 0.002 0.002 24.209 24.307 qs_ot_p2m_diag 82 11.4 0.867 0.873 22.114 22.146 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 19.193 21.041 make_m2s 4970 13.5 0.075 0.077 19.877 20.773 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.395 20.415 calculate_rho_elec 127 8.7 0.475 0.476 20.394 20.414 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.391 20.394 ot_diis_step 116 11.6 0.022 0.022 20.324 20.324 make_images 4970 14.5 3.717 3.952 19.402 20.302 sum_up_and_integrate 127 10.3 0.318 0.321 19.746 19.831 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.406 19.819 apply_single 127 13.6 0.001 0.001 19.406 19.819 integrate_v_rspace 127 11.3 0.004 0.004 19.427 19.512 multiply_cannon_multrec 9940 15.5 10.390 12.104 17.718 17.784 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.189 17.190 cp_fm_diag_elpa_base 82 14.4 12.807 14.351 17.185 17.185 multiply_cannon_sync_h2d 9940 15.5 15.538 15.548 15.538 15.548 hybrid_alltoall_any 5155 16.4 1.289 3.009 10.846 12.659 make_images_data 4970 15.5 0.058 0.062 10.707 12.491 init_scf_run 11 5.9 0.000 0.001 12.309 12.309 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.308 12.309 pw_transfer 1535 11.6 0.092 0.093 11.322 11.333 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.088 11.099 fft_wrap_pw1pw2_140 519 13.2 3.046 3.079 9.846 9.859 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.470 9.549 mp_alltoall_d11v 2401 14.1 8.322 9.200 8.322 9.200 wfi_extrapolate 11 7.9 0.001 0.001 9.174 9.174 grid_integrate_task_list 127 12.3 8.788 8.971 8.788 8.971 cp_fm_cholesky_decompose 22 10.9 8.887 8.964 8.887 8.964 dbcsr_mm_accdrv_process 20590 16.0 3.730 5.511 7.091 8.869 density_rs2pw 127 9.7 0.005 0.005 7.915 8.078 qs_energies_init_hamiltonians 11 5.9 0.009 0.011 7.977 7.977 calculate_dm_sparse 127 9.5 0.001 0.001 6.643 6.721 grid_collocate_task_list 127 9.7 6.434 6.507 6.434 6.507 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.346 6.453 rs_scatter_matrices 138 9.7 3.658 4.595 6.018 6.244 fft3d_ps 1281 14.7 2.824 2.834 6.206 6.240 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.019 6.135 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=300.118000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2784.090909, yerr=164.419892 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 1.256030E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.090 85.060 85.063 qs_energies 1 2.0 0.000 0.001 84.482 84.494 ls_scf 1 3.0 0.002 0.013 83.556 83.562 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.435 72.623 multiply_cannon 111 7.7 0.018 0.021 55.826 56.945 multiply_cannon_loop 111 8.7 0.209 0.224 52.480 53.772 ls_scf_main 1 4.0 0.000 0.003 52.254 52.255 density_matrix_trs4 2 5.0 0.002 0.003 46.678 46.775 ls_scf_init_scf 1 4.0 0.000 0.003 28.223 28.224 ls_scf_init_matrix_S 1 5.0 0.000 0.002 27.041 27.100 mp_waitall_1 11031 10.9 22.459 25.896 22.459 25.896 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 24.943 24.959 multiply_cannon_multrec 2664 9.7 8.160 8.992 15.454 17.288 multiply_cannon_sync_h2d 2664 9.7 13.590 15.367 13.590 15.367 make_m2s 222 7.7 0.009 0.011 13.036 13.480 make_images 222 8.7 0.099 0.110 13.014 13.459 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.709 12.988 make_images_data 222 9.7 0.004 0.005 7.639 8.174 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.447 8.158 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.615 6.911 7.889 hybrid_alltoall_any 227 10.6 0.216 1.830 6.623 7.762 dbcsr_mm_accdrv_process_sort 4760 11.4 6.207 7.106 6.207 7.106 calculate_norms 4752 9.8 5.515 6.232 5.515 6.232 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.020 5.219 mp_sum_l 887 5.1 3.077 4.550 3.077 4.550 make_images_sizes 222 9.7 0.000 0.000 0.658 3.770 mp_alltoall_i44 222 10.7 0.657 3.770 0.657 3.770 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.075 3.556 mp_irecv_dv 6231 10.9 2.059 3.523 2.059 3.523 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.260 3.279 arnoldi_extremal 4 6.8 0.000 0.000 3.249 3.276 arnoldi_normal_ev 4 7.8 0.001 0.003 3.249 3.276 build_subspace 16 8.4 0.009 0.012 3.154 3.157 ls_scf_post 1 4.0 0.000 0.001 3.078 3.085 ls_scf_store_result 1 5.0 0.000 0.000 2.896 2.935 dbcsr_special_finalize 555 9.7 0.005 0.006 2.386 2.758 dbcsr_merge_single_wm 555 10.7 0.455 0.594 2.378 2.750 make_images_pack 222 9.7 2.207 2.626 2.209 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.579 dbcsr_sort_data 658 11.4 2.181 2.487 2.181 2.487 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.473 2.066 2.475 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.348 2.441 buffer_matrices_ensure_size 222 8.7 1.741 2.090 1.741 2.090 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.846 1.851 rebuild_ks_matrix 3 7.3 0.000 0.000 1.837 1.842 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.033 1.837 1.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.063000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1133.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 2.117587E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.045 91.446 91.448 qs_energies 1 2.0 0.000 0.000 90.931 90.936 ls_scf 1 3.0 0.000 0.002 89.614 89.618 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.622 76.029 multiply_cannon 111 7.7 0.029 0.044 53.310 57.399 ls_scf_main 1 4.0 0.000 0.002 55.246 55.251 multiply_cannon_loop 111 8.7 0.116 0.123 50.025 53.351 density_matrix_trs4 2 5.0 0.002 0.003 49.477 49.664 ls_scf_init_scf 1 4.0 0.000 0.001 30.863 30.865 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.627 29.753 mp_waitall_1 9105 10.9 21.460 29.189 21.460 29.189 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.223 27.240 multiply_cannon_multrec 1332 9.7 13.094 16.437 22.288 26.709 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.844 20.900 make_m2s 222 7.7 0.006 0.008 15.554 16.225 make_images 222 8.7 1.573 1.950 15.524 16.196 dbcsr_mm_accdrv_process 4041 10.4 0.284 0.451 8.788 10.360 make_images_data 222 9.7 0.004 0.004 9.019 9.976 dbcsr_mm_accdrv_process_sort 4041 11.4 8.381 9.909 8.381 9.909 hybrid_alltoall_any 227 10.6 0.521 2.458 8.385 9.428 mp_sum_l 887 5.1 5.454 8.411 5.454 8.411 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.236 7.816 mp_irecv_dv 3311 11.0 3.217 7.758 3.217 7.758 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.217 6.770 calculate_norms 2376 9.8 6.000 6.626 6.000 6.626 multiply_cannon_sync_h2d 1332 9.7 4.776 5.939 4.776 5.939 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.080 5.299 arnoldi_extremal 4 6.8 0.000 0.000 4.629 4.651 arnoldi_normal_ev 4 7.8 0.001 0.005 4.629 4.651 build_subspace 16 8.4 0.014 0.021 4.369 4.372 ls_scf_post 1 4.0 0.000 0.001 3.505 3.510 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.131 3.348 ls_scf_store_result 1 5.0 0.000 0.000 3.237 3.331 dbcsr_matrix_vector_mult_local 304 10.0 2.727 3.206 2.729 3.207 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.226 2.770 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.637 2.727 mp_allgather_i34 111 8.7 0.992 2.505 0.992 2.505 make_images_pack 222 9.7 2.017 2.422 2.020 2.425 dbcsr_sort_data 436 11.2 1.822 2.034 1.822 2.034 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.883 1.887 rebuild_ks_matrix 3 7.3 0.000 0.000 1.870 1.874 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.009 1.870 1.874 dbcsr_data_new 4174 10.1 1.606 1.840 1.606 1.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.448000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 2.684002E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.078 93.142 93.150 qs_energies 1 2.0 0.000 0.000 92.573 92.577 ls_scf 1 3.0 0.000 0.000 91.180 91.184 dbcsr_multiply_generic 111 6.7 0.016 0.016 75.874 76.154 ls_scf_main 1 4.0 0.000 0.000 56.985 56.989 multiply_cannon 111 7.7 0.049 0.123 52.653 56.270 multiply_cannon_loop 111 8.7 0.100 0.106 49.055 52.795 density_matrix_trs4 2 5.0 0.002 0.003 50.924 51.127 mp_waitall_1 7281 11.0 23.459 33.409 23.459 33.409 ls_scf_init_scf 1 4.0 0.000 0.000 30.712 30.715 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.490 29.553 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.130 27.144 multiply_cannon_multrec 888 9.7 12.626 15.395 21.129 24.407 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.818 22.671 make_m2s 222 7.7 0.006 0.006 16.792 18.059 make_images 222 8.7 1.970 2.294 16.754 18.023 make_images_data 222 9.7 0.003 0.004 9.615 10.662 hybrid_alltoall_any 227 10.6 0.623 2.879 9.302 10.487 dbcsr_mm_accdrv_process 3754 10.4 0.247 0.445 8.035 9.283 dbcsr_mm_accdrv_process_sort 3754 11.4 7.658 8.838 7.658 8.838 mp_sum_l 887 5.1 4.868 8.086 4.868 8.086 multiply_cannon_sync_h2d 888 9.7 6.089 7.490 6.089 7.490 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.449 6.865 mp_irecv_dv 2335 11.1 2.434 6.802 2.434 6.802 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.652 6.739 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.717 6.389 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.914 5.114 arnoldi_extremal 4 6.8 0.000 0.000 5.061 5.079 arnoldi_normal_ev 4 7.8 0.001 0.005 5.061 5.079 build_subspace 16 8.4 0.014 0.020 4.756 4.766 calculate_norms 1584 9.8 4.256 4.566 4.256 4.566 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.409 3.739 mp_allgather_i34 111 8.7 1.439 3.717 1.439 3.717 dbcsr_matrix_vector_mult_local 304 10.0 2.999 3.577 3.001 3.579 ls_scf_post 1 4.0 0.000 0.000 3.482 3.485 ls_scf_store_result 1 5.0 0.000 0.000 3.241 3.308 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.910 3.030 make_images_pack 222 9.7 1.808 2.090 1.811 2.093 dbcsr_sort_data 325 11.1 1.921 2.092 1.921 2.092 dbcsr_data_release 9322 10.9 1.302 2.014 1.302 2.014 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.961 1.962 rebuild_ks_matrix 3 7.3 0.000 0.000 1.943 1.945 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.016 1.943 1.945 make_images_sizes 222 9.7 0.000 0.000 0.837 1.921 mp_alltoall_i44 222 10.7 0.837 1.921 0.837 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.150000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2189.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 3.343593E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.080 97.037 97.040 qs_energies 1 2.0 0.000 0.000 96.312 96.317 ls_scf 1 3.0 0.004 0.015 94.624 94.631 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.344 78.538 ls_scf_main 1 4.0 0.001 0.016 58.735 58.740 multiply_cannon 111 7.7 0.049 0.118 51.744 55.770 density_matrix_trs4 2 5.0 0.008 0.031 52.560 52.669 multiply_cannon_loop 111 8.7 0.114 0.126 46.801 49.719 ls_scf_init_scf 1 4.0 0.000 0.003 32.620 32.621 ls_scf_init_matrix_S 1 5.0 0.000 0.003 31.323 31.389 mp_waitall_1 6369 11.0 23.085 30.188 23.085 30.188 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.857 28.870 multiply_cannon_multrec 1332 9.7 14.111 16.951 21.955 24.776 make_m2s 222 7.7 0.007 0.008 21.121 22.478 make_images 222 8.7 3.121 3.604 21.070 22.430 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.633 16.707 make_images_data 222 9.7 0.004 0.004 11.813 13.313 hybrid_alltoall_any 227 10.6 0.798 3.773 11.217 13.034 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.404 7.467 8.939 dbcsr_mm_accdrv_process_sort 3641 11.4 7.094 8.523 7.094 8.523 mp_sum_l 887 5.1 3.934 6.811 3.934 6.811 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.260 mp_irecv_dv 3229 10.9 2.057 6.177 2.057 6.177 multiply_cannon_sync_h2d 1332 9.7 5.478 6.130 5.478 6.130 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.974 5.189 arnoldi_extremal 4 6.8 0.000 0.000 5.142 5.159 arnoldi_normal_ev 4 7.8 0.001 0.005 5.142 5.159 build_subspace 16 8.4 0.014 0.021 4.797 4.806 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.532 4.707 calculate_norms 2376 9.8 4.222 4.650 4.222 4.650 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.471 4.483 mp_allgather_i34 111 8.7 2.063 4.152 2.063 4.152 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.538 3.831 dbcsr_matrix_vector_mult_local 304 10.0 3.154 3.649 3.156 3.651 dbcsr_sort_data 658 11.4 3.071 3.469 3.071 3.469 ls_scf_post 1 4.0 0.000 0.001 3.265 3.270 dbcsr_special_finalize 555 9.7 0.006 0.007 2.821 3.181 dbcsr_merge_single_wm 555 10.7 0.536 0.661 2.812 3.173 ls_scf_dm_to_ks 2 5.0 0.008 0.049 3.075 3.149 ls_scf_store_result 1 5.0 0.000 0.000 3.002 3.058 dbcsr_data_release 10477 10.7 1.604 2.389 1.604 2.389 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.034 2.035 rebuild_ks_matrix 3 7.3 0.000 0.000 2.011 2.012 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.011 2.011 2.012 dbcsr_finalize 304 7.8 0.049 0.062 1.794 1.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.040000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2701.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 4.671021E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.049 0.061 99.279 99.280 qs_energies 1 2.0 0.000 0.000 98.452 98.456 ls_scf 1 3.0 0.000 0.000 96.515 96.523 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.027 78.231 ls_scf_main 1 4.0 0.000 0.000 62.132 62.133 multiply_cannon 111 7.7 0.082 0.158 55.343 60.917 density_matrix_trs4 2 5.0 0.002 0.003 54.967 55.044 multiply_cannon_loop 111 8.7 0.069 0.075 50.944 52.535 mp_waitall_1 5436 11.0 26.795 32.256 26.795 32.256 ls_scf_init_scf 1 4.0 0.000 0.000 30.760 30.765 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.473 29.507 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.305 27.310 multiply_cannon_multrec 444 9.7 13.907 16.745 20.920 22.623 make_m2s 222 7.7 0.004 0.005 17.861 20.234 make_images 222 8.7 3.711 4.411 17.800 20.173 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.192 16.716 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.358 13.694 make_images_data 222 9.7 0.003 0.004 10.048 12.403 hybrid_alltoall_any 227 10.6 0.789 3.775 9.900 12.349 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.342 6.716 7.861 multiply_cannon_sync_h2d 444 9.7 6.570 7.632 6.570 7.632 dbcsr_mm_accdrv_process_sort 3003 11.4 6.401 7.508 6.401 7.508 mp_allgather_i34 111 8.7 2.633 7.015 2.633 7.015 arnoldi_extremal 4 6.8 0.000 0.000 5.737 5.744 arnoldi_normal_ev 4 7.8 0.002 0.005 5.737 5.744 build_subspace 16 8.4 0.015 0.020 5.353 5.359 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.546 4.697 mp_sum_l 887 5.1 2.867 4.588 2.867 4.588 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.691 4.500 mp_irecv_dv 1241 11.2 1.673 4.469 1.673 4.469 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.116 4.315 dbcsr_matrix_vector_mult_local 304 10.0 3.638 4.117 3.640 4.119 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.612 3.688 calculate_norms 792 9.8 3.542 3.644 3.542 3.644 ls_scf_post 1 4.0 0.000 0.000 3.622 3.625 make_images_sizes 222 9.7 0.000 0.000 1.174 3.537 mp_alltoall_i44 222 10.7 1.174 3.536 1.174 3.536 ls_scf_store_result 1 5.0 0.000 0.000 3.403 3.445 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.984 3.372 dbcsr_finalize 304 7.8 0.062 0.078 2.202 2.293 dbcsr_merge_all 275 8.9 0.473 0.526 2.051 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.103 2.104 rebuild_ks_matrix 3 7.3 0.000 0.000 2.070 2.072 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.070 2.072 dbcsr_data_release 10123 10.8 1.343 2.023 1.343 2.023 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.280000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3681.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_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 8.813666E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.099 108.409 108.410 qs_energies 1 2.0 0.000 0.000 107.007 107.023 ls_scf 1 3.0 0.000 0.000 104.098 104.113 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.831 77.952 ls_scf_main 1 4.0 0.000 0.000 65.896 65.897 density_matrix_trs4 2 5.0 0.002 0.003 56.664 56.715 multiply_cannon 111 7.7 0.150 0.221 50.092 51.915 multiply_cannon_loop 111 8.7 0.067 0.069 46.630 47.746 ls_scf_init_scf 1 4.0 0.000 0.000 34.544 34.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.969 32.987 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.792 29.801 mp_waitall_1 4527 11.1 22.276 26.153 22.276 26.153 make_m2s 222 7.7 0.005 0.005 23.837 24.845 make_images 222 8.7 4.569 4.992 23.731 24.736 multiply_cannon_multrec 444 9.7 17.926 18.576 22.562 23.057 make_images_data 222 9.7 0.003 0.003 13.113 15.566 hybrid_alltoall_any 227 10.6 1.660 3.635 12.918 15.550 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.573 10.956 multiply_cannon_sync_h2d 444 9.7 8.851 8.901 8.851 8.901 arnoldi_extremal 4 6.8 0.000 0.000 7.210 7.221 arnoldi_normal_ev 4 7.8 0.005 0.013 7.210 7.221 build_subspace 16 8.4 0.026 0.036 6.641 6.649 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.330 5.600 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.304 5.463 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.221 5.312 dbcsr_matrix_vector_mult_local 304 10.0 4.905 5.194 4.907 5.196 dbcsr_mm_accdrv_process 1814 10.4 0.206 0.325 4.461 4.691 dbcsr_mm_accdrv_process_sort 1814 11.4 4.119 4.248 4.119 4.248 ls_scf_post 1 4.0 0.000 0.000 3.658 3.673 mp_allgather_i34 111 8.7 1.120 3.568 1.120 3.568 make_images_sizes 222 9.7 0.000 0.000 1.423 3.485 mp_alltoall_i44 222 10.7 1.423 3.484 1.423 3.484 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.419 calculate_norms 792 9.8 3.237 3.275 3.237 3.275 dbcsr_finalize 304 7.8 0.082 0.090 3.083 3.144 dbcsr_merge_all 275 8.9 0.893 0.921 2.868 2.923 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.879 2.879 dbcsr_data_release 12724 10.6 2.335 2.870 2.335 2.870 dbcsr_complete_redistribute 5 7.6 1.425 1.463 2.753 2.847 matrix_ls_to_qs 2 6.0 0.000 0.000 2.420 2.541 dbcsr_sort_data 325 11.1 2.441 2.514 2.441 2.514 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.482 2.483 rebuild_ks_matrix 3 7.3 0.000 0.000 2.416 2.418 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.416 2.418 mp_sum_l 887 5.1 1.570 2.347 1.570 2.347 dbcsr_new_transposed 4 7.5 0.244 0.253 2.298 2.330 dbcsr_frobenius_norm 74 6.6 2.055 2.132 2.190 2.230 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.205 dbcsr_add_anytype 103 7.2 0.859 0.891 2.132 2.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.410000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6799.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/65f55731d7d9d8bf99167de5ccd8ae8600377d2e_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 582.144000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 408834079. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.123 290.666 290.668 qs_mol_dyn_low 1 2.0 0.004 0.022 289.618 289.640 qs_forces 5 3.8 0.006 0.096 289.031 289.033 qs_energies 5 4.8 0.002 0.028 285.402 285.430 scf_env_do_scf 5 5.8 0.000 0.006 267.262 267.264 scf_env_do_scf_inner_loop 105 6.6 0.002 0.008 233.193 233.196 qs_scf_new_mos 105 7.6 0.000 0.001 180.943 181.452 qs_scf_loop_do_ot 105 8.6 0.001 0.001 180.943 181.451 ot_scf_mini 105 9.6 0.003 0.006 170.902 171.294 dbcsr_multiply_generic 1445 12.2 0.126 0.134 138.586 140.571 multiply_cannon 1445 13.2 0.280 0.296 117.037 123.339 velocity_verlet 4 3.0 0.001 0.009 121.348 121.349 multiply_cannon_loop 1445 14.2 2.387 2.530 114.425 118.755 qs_ot_get_p 112 10.4 0.001 0.001 78.353 78.752 qs_ot_p2m_diag 40 11.0 0.020 0.031 66.066 66.176 cp_dbcsr_syevd 40 12.0 0.002 0.004 62.577 62.580 ot_mini 105 10.6 0.001 0.002 62.151 62.300 cp_fm_syevd 40 13.0 0.000 0.003 56.338 56.488 cp_fm_redistribute_end 40 14.0 25.405 50.663 25.416 50.670 cp_fm_syevd_base 40 14.0 25.231 50.509 25.231 50.509 mp_waitall_1 488190 16.1 37.924 49.583 37.924 49.583 multiply_cannon_multrec 69360 15.2 31.651 37.411 39.469 45.640 multiply_cannon_metrocomm3 69360 15.2 0.193 0.203 25.361 41.196 qs_ot_get_derivative 55 11.6 0.001 0.002 40.400 40.676 multiply_cannon_sync_h2d 69360 15.2 30.502 34.962 30.502 34.962 init_scf_loop 7 6.6 0.000 0.006 34.018 34.021 rebuild_ks_matrix 110 8.4 0.000 0.000 33.496 33.710 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.023 33.496 33.710 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.645 30.837 prepare_preconditioner 7 7.6 0.000 0.000 28.408 28.441 make_preconditioner 7 8.6 0.000 0.002 28.408 28.441 qs_rho_update_rho_low 110 7.6 0.001 0.001 25.531 25.638 calculate_rho_elec 110 8.6 0.030 0.033 25.531 25.637 rs_pw_transfer 690 11.5 0.011 0.020 21.967 23.478 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.043 23.298 apply_single 62 13.6 0.000 0.000 23.042 23.298 ot_new_cg_direction 55 11.6 0.001 0.003 21.000 21.002 density_rs2pw 110 9.6 0.006 0.008 19.265 20.665 make_full_inverse_cholesky 7 9.6 0.000 0.002 20.443 20.510 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.091 15.541 init_scf_run 5 5.8 0.000 0.001 15.240 15.241 scf_env_initial_rho_setup 5 6.8 0.002 0.003 15.240 15.241 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.443 14.551 cp_fm_cholesky_invert 7 10.6 14.212 14.246 14.212 14.246 mp_sum_l 4764 12.2 12.592 13.794 12.592 13.794 pw_transfer 1645 12.4 0.085 0.113 13.398 13.667 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.254 13.527 multiply_cannon_metrocomm1 69360 15.2 0.090 0.098 4.611 12.690 sum_up_and_integrate 60 10.3 0.028 0.033 12.519 12.540 integrate_v_rspace 60 11.3 0.002 0.009 12.491 12.513 calculate_dm_sparse 110 9.5 0.000 0.001 11.707 11.885 check_diag 80 13.5 8.587 8.859 11.125 11.276 fft_wrap_pw1pw2_240 915 15.0 1.206 1.298 10.951 11.214 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.899 10.977 qs_vxc_create 110 10.4 0.002 0.005 10.798 10.847 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 10.212 10.228 calculate_first_density_matrix 1 7.0 0.000 0.003 9.754 9.774 acc_transpose_blocks 69360 15.2 0.356 0.375 9.284 9.766 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.635 9.696 fft3d_pb 915 16.0 2.379 2.663 9.000 9.344 rs_pw_transfer_RS2PW_30 110 11.6 1.571 1.666 8.514 9.186 make_m2s 2890 13.2 0.077 0.083 8.538 9.125 make_images 2890 14.2 0.245 0.265 8.434 9.022 mp_sendrecv_dv 168740 12.6 8.641 8.872 8.641 8.872 dbcsr_mm_accdrv_process 154766 15.8 4.009 4.183 7.689 8.518 potential_pw2rs 60 12.3 0.003 0.003 8.402 8.476 xc_rho_set_and_dset_create 110 12.4 0.076 0.100 7.543 7.791 acc_transpose_blocks_kernels 69360 16.2 0.850 0.895 7.391 7.762 make_full_single_inverse 7 9.6 0.001 0.004 7.580 7.614 xc_vxc_pw_create 60 11.3 0.039 0.052 7.245 7.295 jit_kernel_transpose 5 15.0 6.542 6.885 6.542 6.885 mp_alltoall_z22v 2340 17.7 6.115 6.649 6.115 6.649 xc_pw_derive 510 13.4 0.006 0.007 6.522 6.587 mp_waitany 7680 13.5 4.655 6.046 4.655 6.046 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=290.668000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=555.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 65f55731d7d9d8bf99167de5ccd8ae8600377d2e Summary: empty Status: OK