=== This is the CP2K Performance-Test === Updating c7724263c..e0ef5d2e5 Fast-forward tools/docker/Dockerfile.test_cmake | 21 ++++++------------ tools/docker/generate_dockerfiles.py | 26 ++++++----------------- tools/docker/scripts/test_cmake.sh | 41 ++++++++++++++++++++++++++++++++++++ 3 files changed, 53 insertions(+), 35 deletions(-) create mode 100755 tools/docker/scripts/test_cmake.sh Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: e0ef5d2e5812615faf6ac12d806b53b060af4958 ################# 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/01 job id: 46412632 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/02 job id: 46412633 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/03 job id: 46412634 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/04 job id: 46412635 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/05 job id: 46412637 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/06 job id: 46412638 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/07 job id: 46412639 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/08 job id: 46412640 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/09 job id: 46412641 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/10 job id: 46412642 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/11 job id: 46412644 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/12 job id: 46412645 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/13 job id: 46412646 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/14 job id: 46412650 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/15 job id: 46412657 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/16 job id: 46412659 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/17 job id: 46412660 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/18 job id: 46412662 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/19 job id: 46412663 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/20 job id: 46412665 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/21 job id: 46412669 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/22 job id: 46412671 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/23 job id: 46412672 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/24 job id: 46412675 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/25 job id: 46412677 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/26 job id: 46412679 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/27 job id: 46412681 --- 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/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.064 0.065 137.662 137.663 farming_run 1 2.0 135.031 135.299 137.541 137.583 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452663E+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 491 2254389. 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 0.007 0.028 115.212 115.213 qs_energies 1 2.0 0.000 0.000 114.977 114.979 mp2_main 1 3.0 0.000 0.000 112.913 112.915 mp2_gpw_main 1 4.0 0.019 0.025 111.888 111.890 mp2_ri_gpw_compute_in 1 5.0 0.172 0.178 92.928 93.344 mp2_ri_gpw_compute_in_loop 1 6.0 0.005 0.005 55.500 55.919 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.827 47.204 get_2c_integrals 1 6.0 0.008 0.009 36.741 37.256 integrate_v_rspace 273 8.0 0.440 0.453 25.188 30.303 pw_transfer 6555 10.6 0.378 0.384 27.308 27.897 grid_integrate_task_list 273 9.0 20.995 26.606 20.995 26.606 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.978 26.576 fft_wrap_pw1pw2_100 2178 12.4 1.146 1.235 23.506 24.107 rpa_ri_compute_en 1 5.0 0.019 0.021 18.852 19.203 compute_2c_integrals 1 7.0 0.002 0.003 19.142 19.145 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.757 18.876 mp2_eri_2c_integrate_gpw 1 9.0 2.394 2.434 18.755 18.875 cp_fm_cholesky_decompose 12 8.2 17.524 18.030 17.524 18.030 cholesky_decomp 1 7.0 0.000 0.000 16.438 16.944 fft3d_s 5443 13.4 16.134 16.672 16.156 16.695 ao_to_mo_and_store_B_mult_1 272 7.0 10.850 15.574 10.850 15.574 calculate_wavefunction 272 8.0 5.432 5.589 12.556 13.187 rpa_num_int 1 6.0 0.000 0.000 10.669 10.669 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.626 10.639 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.589 contract_S_to_Q 8 9.0 0.000 0.000 8.914 9.006 parallel_gemm_fm 14 9.1 0.000 0.000 8.495 8.607 parallel_gemm_fm_cosma 14 10.1 8.495 8.607 8.495 8.607 calc_potential_gpw 544 9.5 0.005 0.005 8.221 8.576 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.143 8.381 potential_pw2rs 545 10.0 0.108 0.110 7.657 8.259 create_integ_mat 1 6.0 0.023 0.027 7.668 7.669 collocate_single_gaussian 272 10.0 0.039 0.042 7.403 7.605 array2fm 1 7.0 0.000 0.000 6.673 7.101 pw_scatter_s 2720 13.7 4.421 4.614 4.421 4.614 pw_gather_s 2722 13.2 3.875 4.213 3.875 4.213 array2fm_buffer_send 1 8.0 2.950 3.116 2.950 3.116 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.890458, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2790.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.052 403.182 403.184 farming_run 1 2.0 400.923 400.931 403.127 403.131 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223344E+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 1941 22272. MP_Sync 37 MP_Alltoall 77 23054071. 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 0.010 0.028 210.880 210.881 qs_energies 1 2.0 0.039 0.307 210.624 210.633 scf_env_do_scf 1 3.0 0.000 0.000 106.730 106.731 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.839 105.847 rebuild_ks_matrix 4 6.0 0.000 0.000 105.838 105.846 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.838 105.846 hfx_ks_matrix 4 8.0 0.001 0.001 105.456 105.462 integrate_four_center 4 9.0 0.143 0.460 105.456 105.461 mp2_main 1 3.0 0.010 0.080 103.564 103.612 mp2_gpw_main 1 4.0 0.044 0.169 102.415 102.432 integrate_four_center_main 4 10.0 0.141 0.472 96.581 99.868 integrate_four_center_bin 263 11.0 96.440 99.805 96.440 99.805 init_scf_loop 1 4.0 0.000 0.000 92.531 92.531 mp2_ri_gpw_compute_in 1 5.0 0.070 0.119 74.932 76.079 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.462 55.601 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.159 42.177 47.289 integrate_v_rspace 95 8.0 0.399 0.571 28.549 33.434 pw_transfer 2240 10.6 0.144 0.166 29.926 30.382 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.929 29.393 ao_to_mo_and_store_B_mult_1 91 7.0 10.592 29.204 10.592 29.204 mp2_ri_gpw_compute_en 1 5.0 0.074 0.173 27.234 28.985 grid_integrate_task_list 95 9.0 23.848 28.943 23.848 28.943 fft_wrap_pw1pw2_100 730 12.4 1.285 1.431 26.609 27.021 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.824 1.904 25.392 25.407 get_2c_integrals 1 6.0 0.008 0.069 20.336 20.370 compute_2c_integrals 1 7.0 0.004 0.018 19.310 19.327 compute_2c_integrals_loop_lm 1 8.0 0.002 0.011 18.910 19.198 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.933 18.908 19.196 fft3d_s 1823 13.4 18.417 18.752 18.430 18.766 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.198 14.198 calculate_wavefunction 91 8.0 2.014 2.046 9.740 9.998 mp2_ri_gpw_compute_en_expansio 172 7.0 0.559 0.583 8.915 9.770 local_gemm 172 8.0 8.355 9.199 8.355 9.199 potential_pw2rs 186 10.0 0.034 0.035 8.561 9.047 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.522 8.267 8.841 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.002 0.018 8.283 8.481 calc_potential_gpw 182 9.5 0.002 0.002 8.017 8.238 collocate_single_gaussian 91 10.0 0.019 0.076 7.886 8.046 mp_sendrecv_dm3 2068 8.0 6.313 6.882 6.313 6.882 mp_sync 37 10.5 3.582 6.801 3.582 6.801 mp2_ri_gpw_compute_en_ener 172 7.0 6.342 6.468 6.342 6.468 pw_gather_s 912 13.2 4.857 5.324 4.857 5.324 pw_scatter_s 910 13.7 3.975 4.227 3.975 4.227 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.402973, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 448.040960E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 90.013 90.014 qs_mol_dyn_low 1 2.0 0.003 0.003 85.382 89.678 velocity_verlet 10 3.0 0.006 0.037 56.626 56.925 qs_forces 11 3.9 0.002 0.003 56.579 56.580 qs_energies 11 4.9 0.001 0.002 55.129 55.142 scf_env_do_scf 11 5.9 0.000 0.001 47.739 47.739 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 45.699 45.700 dbcsr_multiply_generic 2286 12.5 0.092 0.096 34.458 34.887 qs_scf_new_mos 108 7.5 0.000 0.001 33.661 33.942 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.660 33.941 md_output 10 3.0 0.000 0.000 0.369 33.013 ot_scf_mini 108 9.5 0.002 0.002 31.975 32.177 md_write_output 11 3.9 0.203 19.462 0.328 31.362 multiply_cannon 2286 13.5 0.193 0.202 27.274 28.877 mp_sum_dm 438 4.9 28.465 28.799 28.465 28.799 update_particle_set 20 4.0 0.000 0.000 28.417 28.721 multiply_cannon_loop 2286 14.5 1.480 1.577 26.592 28.244 ot_mini 108 10.5 0.001 0.001 19.255 19.508 qs_ot_get_derivative 108 11.5 0.001 0.001 16.275 16.493 mp_waitall_1 245248 16.5 8.894 15.184 8.894 15.184 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 6.368 13.668 multiply_cannon_multrec 54864 15.5 4.247 6.579 8.428 12.108 write_trajectory 44 4.9 0.124 11.883 0.124 11.900 qs_ot_get_p 119 10.4 0.001 0.002 8.102 8.383 rebuild_ks_matrix 119 8.3 0.000 0.000 7.846 7.993 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.846 7.993 multiply_cannon_sync_h2d 54864 15.5 5.930 7.164 5.930 7.164 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.922 7.057 mp_sum_l 7207 12.9 5.271 6.853 5.271 6.853 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.741 6.197 init_scf_run 11 5.9 0.000 0.001 6.181 6.181 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.181 6.181 dbcsr_mm_accdrv_process 76910 16.1 1.158 1.847 4.101 5.759 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.305 5.340 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.056 5.153 sum_up_and_integrate 119 10.3 0.012 0.014 4.483 4.490 integrate_v_rspace 119 11.3 0.002 0.003 4.470 4.479 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.463 4.463 mp_max_l 33 2.8 4.297 4.365 4.297 4.365 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.198 4.330 calculate_rho_elec 119 8.7 0.012 0.017 4.198 4.329 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.264 4.265 cp_fm_redistribute_end 50 14.0 2.179 4.241 2.184 4.243 cp_fm_diag_elpa_base 50 14.0 2.054 4.127 2.057 4.136 calculate_first_density_matrix 1 7.0 0.000 0.001 3.921 3.932 jit_kernel_multiply 13 15.8 2.881 3.819 2.881 3.819 calculate_dm_sparse 119 9.5 0.000 0.001 3.304 3.452 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.098 3.100 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.883 3.089 apply_single 119 13.6 0.000 0.000 2.882 3.089 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.903 3.054 rs_pw_transfer 974 11.9 0.012 0.013 2.719 2.843 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.761 2.814 ot_diis_step 108 11.5 0.006 0.006 2.687 2.687 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.432 2.495 density_rs2pw 119 9.7 0.004 0.004 2.227 2.385 acc_transpose_blocks 54864 15.5 0.227 0.249 1.771 2.308 wfi_extrapolate 11 7.9 0.001 0.001 2.191 2.191 mp_bcast_b_src 168 9.3 2.031 2.153 2.031 2.153 external_control 118 7.2 0.021 1.961 2.051 2.148 grid_integrate_task_list 119 12.3 2.035 2.132 2.035 2.132 init_scf_loop 11 6.9 0.000 0.000 2.021 2.021 mp_sum_d 4135 12.0 1.307 1.907 1.307 1.907 potential_pw2rs 119 12.3 0.004 0.004 1.823 1.834 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.778 1.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=90.014000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.272727, yerr=1.052349 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 482.873344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 995295. 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 0.015 0.033 62.471 62.472 qs_mol_dyn_low 1 2.0 0.003 0.003 54.808 62.242 qs_forces 11 3.9 0.002 0.003 42.184 42.185 qs_energies 11 4.9 0.027 0.207 40.405 40.430 scf_env_do_scf 11 5.9 0.001 0.002 33.634 33.634 velocity_verlet 10 3.0 0.025 0.044 31.426 31.685 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.959 30.960 dbcsr_multiply_generic 2286 12.5 0.099 0.102 23.009 23.432 qs_scf_new_mos 108 7.5 0.001 0.001 21.057 21.298 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.056 21.298 ot_scf_mini 108 9.5 0.002 0.003 20.143 20.314 md_output 10 3.0 0.000 0.000 0.420 19.999 multiply_cannon 2286 13.5 0.210 0.218 17.617 19.225 multiply_cannon_loop 2286 14.5 0.902 0.974 16.354 18.172 md_write_output 11 3.9 0.206 9.882 0.355 16.958 ot_mini 108 10.5 0.001 0.001 12.333 12.565 mp_sum_dm 438 4.9 12.200 12.476 12.200 12.476 mp_waitall_1 200699 16.5 6.675 12.459 6.675 12.459 update_particle_set 20 4.0 0.000 0.000 12.167 12.449 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.685 10.452 qs_ot_get_derivative 108 11.5 0.001 0.001 9.832 10.005 multiply_cannon_multrec 27432 15.5 1.958 4.521 6.395 9.528 rebuild_ks_matrix 119 8.3 0.000 0.000 7.625 7.777 qs_ks_build_kohn_sham_matrix 119 9.3 0.017 0.031 7.625 7.777 mp_max_l 33 2.8 7.435 7.593 7.435 7.593 write_trajectory 44 4.9 0.148 7.069 0.148 7.076 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.719 6.859 dbcsr_mm_accdrv_process 47894 16.0 3.563 6.086 4.366 6.770 init_scf_run 11 5.9 0.000 0.001 5.165 5.166 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.165 5.165 qs_ot_get_p 119 10.4 0.084 0.135 4.841 5.073 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.789 4.663 sum_up_and_integrate 119 10.3 0.024 0.027 4.568 4.573 integrate_v_rspace 119 11.3 0.002 0.003 4.543 4.550 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.307 4.356 calculate_rho_elec 119 8.7 0.021 0.024 4.307 4.356 mp_sum_l 7207 12.9 2.227 4.192 2.227 4.192 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.109 4.178 apply_single 119 13.6 0.000 0.000 3.109 4.178 rs_pw_transfer 974 11.9 0.010 0.011 3.347 3.794 calculate_first_density_matrix 1 7.0 0.001 0.002 3.457 3.460 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.199 3.303 density_rs2pw 119 9.7 0.004 0.004 2.608 3.076 write_restart 10 4.0 0.064 3.049 0.065 3.050 make_m2s 4572 13.5 0.052 0.054 2.727 3.021 make_images 4572 14.5 0.201 0.236 2.639 2.930 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.748 2.832 multiply_cannon_sync_h2d 27432 15.5 2.162 2.764 2.162 2.764 init_scf_loop 11 6.9 0.001 0.005 2.651 2.651 jit_kernel_multiply 10 16.2 0.750 2.620 0.750 2.620 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.599 2.602 calculate_dm_sparse 119 9.5 0.000 0.001 2.458 2.528 ot_diis_step 108 11.5 0.011 0.011 2.440 2.441 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.304 2.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.329 2.370 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.315 2.316 potential_pw2rs 119 12.3 0.006 0.006 2.270 2.288 cp_fm_redistribute_end 50 14.0 1.172 2.280 1.176 2.284 cp_fm_diag_elpa_base 50 14.0 1.074 2.177 1.102 2.221 pw_transfer 1439 11.6 0.065 0.070 2.063 2.111 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.971 2.021 grid_integrate_task_list 119 12.3 1.838 1.955 1.838 1.955 prepare_preconditioner 11 7.9 0.000 0.000 1.705 1.731 make_images_data 4572 15.5 0.045 0.051 1.225 1.731 make_preconditioner 11 8.9 0.000 0.002 1.705 1.731 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.658 1.668 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.599 1.660 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.055 1.618 fft3d_ps 1201 14.6 0.524 0.582 1.523 1.568 wfi_extrapolate 11 7.9 0.001 0.001 1.554 1.554 fft_wrap_pw1pw2_140 487 13.2 0.201 0.210 1.501 1.546 acc_transpose_blocks 27432 15.5 0.110 0.116 1.212 1.497 mp_allgather_i34 2286 14.5 0.702 1.491 0.702 1.491 grid_collocate_task_list 119 9.7 1.243 1.449 1.243 1.449 mp_alltoall_d11v 2130 13.8 1.281 1.396 1.281 1.396 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.233 1.275 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=62.472000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.000000, yerr=1.858641 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 517.398528E+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 10075 307. MP_Sync 54 MP_Alltoall 1821 1607820. 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 0.074 0.121 40.568 40.570 qs_mol_dyn_low 1 2.0 0.003 0.003 35.468 39.885 qs_forces 11 3.9 0.015 0.021 35.009 35.030 qs_energies 11 4.9 0.020 0.152 33.299 33.360 scf_env_do_scf 11 5.9 0.001 0.001 26.533 26.533 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.508 23.510 dbcsr_multiply_generic 2286 12.5 0.117 0.161 18.417 18.506 multiply_cannon 2286 13.5 0.196 0.203 15.089 15.823 velocity_verlet 10 3.0 0.001 0.001 15.364 15.373 qs_scf_new_mos 108 7.5 0.001 0.001 15.288 15.309 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.287 15.308 multiply_cannon_loop 2286 14.5 0.633 0.669 14.309 15.071 ot_scf_mini 108 9.5 0.002 0.003 14.572 14.581 multiply_cannon_multrec 18288 15.5 2.003 3.116 8.743 9.053 ot_mini 108 10.5 0.001 0.001 8.956 8.972 dbcsr_mm_accdrv_process 38222 16.0 5.616 7.270 6.654 7.714 qs_ot_get_derivative 108 11.5 0.001 0.001 7.472 7.483 rebuild_ks_matrix 119 8.3 0.000 0.000 6.598 6.614 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.021 6.597 6.613 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.790 5.804 init_scf_run 11 5.9 0.000 0.001 5.481 5.481 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.481 5.481 md_output 10 3.0 0.000 0.000 0.154 4.837 mp_max_l 33 2.8 4.418 4.561 4.418 4.561 calculate_first_density_matrix 1 7.0 0.001 0.002 4.268 4.269 write_restart 10 4.0 0.133 4.260 0.134 4.261 sum_up_and_integrate 119 10.3 0.030 0.031 4.003 4.017 integrate_v_rspace 119 11.3 0.002 0.003 3.973 3.990 mp_waitall_1 158411 16.6 2.587 3.693 2.587 3.693 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.970 3.617 qs_ot_get_p 119 10.4 0.026 0.206 3.472 3.499 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.481 3.486 calculate_rho_elec 119 8.7 0.031 0.032 3.480 3.485 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.406 3.408 jit_kernel_multiply 8 16.1 0.986 3.276 0.986 3.276 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.240 3.245 init_scf_loop 11 6.9 0.036 0.284 3.008 3.009 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.627 2.683 rs_pw_transfer 974 11.9 0.009 0.010 2.305 2.563 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.051 2.377 apply_single 119 13.6 0.000 0.000 2.051 2.377 density_rs2pw 119 9.7 0.004 0.004 1.997 2.260 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.199 2.205 pw_transfer 1439 11.6 0.066 0.069 1.951 1.968 calculate_dm_sparse 119 9.5 0.000 0.000 1.939 1.956 make_m2s 4572 13.5 0.044 0.046 1.786 1.952 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.939 1.939 grid_integrate_task_list 119 12.3 1.810 1.892 1.810 1.892 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.858 1.875 make_images 4572 14.5 0.192 0.204 1.701 1.866 potential_pw2rs 119 12.3 0.007 0.008 1.760 1.776 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.697 1.708 cp_fm_diag_elpa_base 50 14.0 1.674 1.686 1.696 1.706 prepare_preconditioner 11 7.9 0.000 0.000 1.700 1.703 make_preconditioner 11 8.9 0.000 0.001 1.700 1.702 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.553 1.637 multiply_cannon_sync_h2d 18288 15.5 1.430 1.607 1.430 1.607 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.546 1.553 mp_sum_l 7207 12.9 1.162 1.531 1.162 1.531 fft_wrap_pw1pw2_140 487 13.2 0.252 0.259 1.508 1.527 ot_diis_step 108 11.5 0.011 0.011 1.464 1.464 grid_collocate_task_list 119 9.7 1.215 1.377 1.215 1.377 fft3d_ps 1201 14.6 0.529 0.545 1.346 1.362 acc_transpose_blocks 18288 15.5 0.076 0.079 1.273 1.359 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.315 1.320 wfi_extrapolate 11 7.9 0.001 0.001 1.164 1.164 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.963 0.997 make_images_data 4572 15.5 0.045 0.048 0.774 0.929 acc_transpose_blocks_kernels 18288 16.5 0.212 0.223 0.826 0.897 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.862 0.886 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.881 0.882 dbcsr_complete_redistribute 329 12.2 0.106 0.130 0.718 0.864 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.662 0.852 mp_alltoall_d11v 2130 13.8 0.645 0.829 0.645 0.829 mp_waitany 9880 13.7 0.533 0.816 0.533 0.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=40.570000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.454545, yerr=1.558766 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 552.759296E+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 10074 349. 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 0.011 0.027 43.933 43.933 qs_mol_dyn_low 1 2.0 0.003 0.003 39.881 43.772 qs_forces 11 3.9 0.002 0.003 38.572 38.573 qs_energies 11 4.9 0.001 0.001 36.802 36.806 scf_env_do_scf 11 5.9 0.000 0.001 29.007 29.008 scf_env_do_scf_inner_loop 108 6.5 0.006 0.012 25.351 25.351 dbcsr_multiply_generic 2286 12.5 0.096 0.099 21.653 21.796 multiply_cannon 2286 13.5 0.223 0.258 18.149 18.957 velocity_verlet 10 3.0 0.001 0.002 18.503 18.550 multiply_cannon_loop 2286 14.5 0.935 0.969 17.241 17.916 qs_scf_new_mos 108 7.5 0.001 0.001 17.016 17.076 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.016 17.075 ot_scf_mini 108 9.5 0.002 0.003 16.093 16.151 multiply_cannon_multrec 27432 15.5 2.914 3.784 11.962 12.412 dbcsr_mm_accdrv_process 47916 15.9 6.844 10.122 8.951 10.845 ot_mini 108 10.5 0.001 0.001 10.199 10.318 qs_ot_get_derivative 108 11.5 0.001 0.001 8.400 8.508 rebuild_ks_matrix 119 8.3 0.000 0.000 6.644 6.701 qs_ks_build_kohn_sham_matrix 119 9.3 0.034 0.075 6.644 6.701 init_scf_run 11 5.9 0.000 0.001 6.447 6.447 scf_env_initial_rho_setup 11 6.9 0.005 0.019 6.447 6.447 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.847 5.901 md_output 10 3.0 0.000 0.000 0.219 5.177 calculate_first_density_matrix 1 7.0 0.001 0.001 5.095 5.101 md_write_output 11 3.9 0.073 1.729 0.198 4.715 jit_kernel_multiply 10 15.8 2.046 4.527 2.046 4.527 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.178 4.179 mp_max_l 33 2.8 3.891 4.060 3.891 4.060 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.901 3.913 sum_up_and_integrate 119 10.3 0.035 0.038 3.755 3.764 integrate_v_rspace 119 11.3 0.003 0.003 3.720 3.729 init_scf_loop 11 6.9 0.000 0.000 3.638 3.638 qs_ot_get_p 119 10.4 0.001 0.002 3.543 3.619 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.082 3.503 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.467 3.498 calculate_rho_elec 119 8.7 0.040 0.046 3.466 3.497 write_trajectory 44 4.9 0.125 2.980 0.125 2.986 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.045 2.584 apply_single 119 13.6 0.000 0.000 2.045 2.584 calculate_dm_sparse 119 9.5 0.000 0.000 2.461 2.513 prepare_preconditioner 11 7.9 0.000 0.000 2.454 2.462 make_preconditioner 11 8.9 0.000 0.000 2.454 2.462 mp_waitall_1 137007 16.6 1.770 2.431 1.770 2.431 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.074 2.392 make_m2s 4572 13.5 0.054 0.056 2.060 2.184 rs_pw_transfer 974 11.9 0.009 0.009 1.948 2.164 make_images 4572 14.5 0.271 0.332 1.953 2.076 density_rs2pw 119 9.7 0.004 0.004 1.884 2.072 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.035 2.044 pw_transfer 1439 11.6 0.066 0.071 1.974 2.007 grid_integrate_task_list 119 12.3 1.825 1.929 1.825 1.929 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.881 1.919 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.887 1.915 acc_transpose_blocks 27432 15.5 0.111 0.115 1.796 1.861 ot_diis_step 108 11.5 0.012 0.012 1.750 1.751 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.727 1.728 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.911 1.688 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.655 1.662 fft_wrap_pw1pw2_140 487 13.2 0.288 0.301 1.561 1.602 mp_sum_l 7207 12.9 1.021 1.546 1.021 1.546 potential_pw2rs 119 12.3 0.009 0.010 1.509 1.512 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.499 1.509 cp_fm_diag_elpa_base 50 14.0 1.466 1.483 1.497 1.507 fft3d_ps 1201 14.6 0.556 0.606 1.318 1.343 grid_collocate_task_list 119 9.7 1.221 1.333 1.221 1.333 wfi_extrapolate 11 7.9 0.001 0.001 1.301 1.301 acc_transpose_blocks_kernels 27432 16.5 0.272 0.279 1.159 1.216 mp_sum_dm 438 4.9 1.092 1.146 1.092 1.146 cp_fm_upper_to_full 72 14.2 0.803 1.138 0.803 1.138 update_particle_set 20 4.0 0.000 0.000 1.076 1.123 dbcsr_complete_redistribute 329 12.2 0.166 0.261 0.839 1.102 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.080 1.095 multiply_cannon_sync_h2d 27432 15.5 0.992 1.070 0.992 1.070 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.069 1.069 jit_kernel_transpose 5 15.6 0.887 0.941 0.887 0.941 mp_sum_d 4133 12.0 0.683 0.921 0.683 0.921 make_images_data 4572 15.5 0.045 0.048 0.794 0.912 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.655 0.911 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.885 0.889 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=43.933000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=524.181818, yerr=3.737237 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 608.182272E+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 10074 348. 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 0.029 32.229 32.230 qs_mol_dyn_low 1 2.0 0.003 0.003 31.906 31.912 qs_forces 11 3.9 0.002 0.002 31.359 31.360 qs_energies 11 4.9 0.001 0.001 29.636 29.640 scf_env_do_scf 11 5.9 0.000 0.001 22.408 22.408 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.982 19.983 dbcsr_multiply_generic 2286 12.5 0.091 0.096 15.967 16.042 velocity_verlet 10 3.0 0.011 0.021 14.561 14.595 multiply_cannon 2286 13.5 0.228 0.237 13.582 13.989 multiply_cannon_loop 2286 14.5 0.330 0.341 12.718 12.906 qs_scf_new_mos 108 7.5 0.001 0.001 12.133 12.151 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.132 12.150 ot_scf_mini 108 9.5 0.002 0.002 11.464 11.490 multiply_cannon_multrec 9144 15.5 2.778 3.172 9.518 9.923 ot_mini 108 10.5 0.001 0.001 7.095 7.126 dbcsr_mm_accdrv_process 12550 15.8 4.301 6.395 6.637 6.764 rebuild_ks_matrix 119 8.3 0.000 0.000 5.918 5.936 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.032 5.918 5.935 qs_ot_get_derivative 108 11.5 0.001 0.001 5.821 5.847 init_scf_run 11 5.9 0.000 0.001 5.802 5.802 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.801 5.802 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.255 5.271 calculate_first_density_matrix 1 7.0 0.001 0.001 4.690 4.690 jit_kernel_multiply 8 15.8 2.297 3.803 2.297 3.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.718 3.719 sum_up_and_integrate 119 10.3 0.037 0.041 3.614 3.618 integrate_v_rspace 119 11.3 0.003 0.003 3.576 3.581 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.569 3.576 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.449 3.455 calculate_rho_elec 119 8.7 0.060 0.061 3.448 3.455 qs_ot_get_p 119 10.4 0.001 0.002 2.633 2.680 init_scf_loop 11 6.9 0.000 0.000 2.406 2.408 mp_waitall_1 115863 16.7 1.460 2.265 1.460 2.265 calculate_dm_sparse 119 9.5 0.000 0.000 2.066 2.085 pw_transfer 1439 11.6 0.066 0.069 2.019 2.029 grid_integrate_task_list 119 12.3 1.848 1.947 1.848 1.947 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.926 1.937 density_rs2pw 119 9.7 0.004 0.004 1.789 1.927 make_m2s 4572 13.5 0.034 0.035 1.645 1.789 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.753 1.771 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.752 1.755 rs_pw_transfer 974 11.9 0.008 0.008 1.617 1.743 make_images 4572 14.5 0.269 0.306 1.555 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.675 1.679 make_preconditioner 11 8.9 0.000 0.000 1.675 1.679 fft_wrap_pw1pw2_140 487 13.2 0.364 0.373 1.595 1.607 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.567 1.592 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.559 1.560 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.488 1.493 grid_collocate_task_list 119 9.7 1.272 1.387 1.272 1.387 potential_pw2rs 119 12.3 0.010 0.011 1.351 1.353 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.288 1.296 cp_fm_diag_elpa_base 50 14.0 1.261 1.278 1.286 1.295 fft3d_ps 1201 14.6 0.560 0.571 1.259 1.267 ot_diis_step 108 11.5 0.013 0.013 1.263 1.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.222 1.223 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.156 1.215 apply_single 119 13.6 0.000 0.000 1.156 1.215 acc_transpose_blocks 9144 15.5 0.038 0.040 1.154 1.162 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.106 1.121 wfi_extrapolate 11 7.9 0.001 0.001 1.064 1.064 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.756 0.954 make_images_data 4572 15.5 0.039 0.043 0.765 0.933 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.920 acc_transpose_blocks_kernels 9144 16.5 0.119 0.122 0.907 0.910 mp_alltoall_d11v 2130 13.8 0.734 0.884 0.734 0.884 cp_fm_cholesky_invert 11 10.9 0.831 0.834 0.831 0.834 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.410 0.815 jit_kernel_transpose 5 15.6 0.788 0.791 0.788 0.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.787 multiply_cannon_sync_h2d 9144 15.5 0.706 0.787 0.706 0.787 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.674 0.725 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.703 0.712 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.325 0.650 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=32.230000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=578.272727, yerr=3.544289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 741.990400E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 62.691 62.691 qs_mol_dyn_low 1 2.0 0.003 0.003 56.344 62.339 qs_forces 11 3.9 0.002 0.003 44.522 44.522 qs_energies 11 4.9 0.001 0.001 42.558 42.561 scf_env_do_scf 11 5.9 0.001 0.001 34.212 34.212 velocity_verlet 10 3.0 0.001 0.001 32.496 33.865 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.507 26.507 dbcsr_multiply_generic 2286 12.5 0.101 0.102 20.817 20.927 multiply_cannon 2286 13.5 0.301 0.306 17.115 17.987 md_output 10 3.0 0.000 0.000 2.227 17.796 qs_scf_new_mos 108 7.5 0.001 0.001 16.853 16.951 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.852 16.950 multiply_cannon_loop 2286 14.5 0.348 0.377 15.904 16.815 md_write_output 11 3.9 1.090 8.715 2.039 16.300 ot_scf_mini 108 9.5 0.002 0.002 15.786 15.883 multiply_cannon_multrec 9144 15.5 4.315 6.047 11.834 12.170 mp_sum_dm 438 4.9 9.602 10.985 9.602 10.985 update_particle_set 20 4.0 0.000 0.000 9.582 10.951 ot_mini 108 10.5 0.001 0.001 9.784 9.897 dbcsr_mm_accdrv_process 12550 15.8 4.664 8.445 7.393 8.664 qs_ot_get_derivative 108 11.5 0.001 0.001 7.813 7.912 init_scf_loop 11 6.9 0.000 0.000 7.679 7.679 write_trajectory 44 4.9 0.948 7.579 0.949 7.585 rebuild_ks_matrix 119 8.3 0.000 0.000 7.229 7.368 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.228 7.367 mp_max_l 33 2.8 5.995 6.851 5.995 6.851 prepare_preconditioner 11 7.9 0.000 0.000 6.734 6.747 make_preconditioner 11 8.9 0.000 0.000 6.734 6.747 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.539 6.665 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.304 6.630 init_scf_run 11 5.9 0.000 0.001 6.321 6.321 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.320 6.320 calculate_first_density_matrix 1 7.0 0.000 0.001 4.754 4.891 cp_fm_upper_to_full 72 14.2 3.170 4.547 3.170 4.547 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.296 4.430 calculate_rho_elec 119 8.7 0.118 0.121 4.296 4.430 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 4.102 4.103 sum_up_and_integrate 119 10.3 0.065 0.066 3.943 3.982 integrate_v_rspace 119 11.3 0.003 0.003 3.878 3.916 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.827 3.880 mp_waitall_1 94719 16.7 2.374 3.596 2.374 3.596 jit_kernel_multiply 6 15.6 2.702 3.302 2.702 3.302 qs_ot_get_p 119 10.4 0.001 0.001 3.048 3.183 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.658 3.123 dbcsr_complete_redistribute 329 12.2 0.287 0.295 1.938 2.747 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.461 2.626 pw_transfer 1439 11.6 0.069 0.070 2.591 2.595 calculate_dm_sparse 119 9.5 0.000 0.000 2.382 2.516 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.492 2.496 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.647 2.457 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.186 2.453 apply_single 119 13.6 0.000 0.000 2.186 2.453 make_m2s 4572 13.5 0.038 0.038 2.119 2.264 mp_alltoall_i22 627 13.8 1.430 2.245 1.430 2.245 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.424 2.228 make_images 4572 14.5 0.356 0.399 1.999 2.143 density_rs2pw 119 9.7 0.004 0.004 2.105 2.120 fft_wrap_pw1pw2_140 487 13.2 0.617 0.621 2.110 2.118 grid_integrate_task_list 119 12.3 2.045 2.057 2.045 2.057 ot_diis_step 108 11.5 0.014 0.015 1.948 1.948 qs_ot_p2m_diag 50 11.0 0.044 0.044 1.851 1.853 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.757 1.757 mp_sum_l 7207 12.9 1.046 1.752 1.046 1.752 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.633 1.683 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.461 1.594 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.581 1.581 fft3d_ps 1201 14.6 0.596 0.605 1.509 1.512 grid_collocate_task_list 119 9.7 1.478 1.506 1.478 1.506 write_restart 10 4.0 0.188 1.502 0.189 1.503 rs_pw_transfer 974 11.9 0.009 0.009 1.407 1.430 cp_fm_cholesky_invert 11 10.9 1.385 1.388 1.385 1.388 wfi_extrapolate 11 7.9 0.001 0.001 1.365 1.365 potential_pw2rs 119 12.3 0.014 0.015 1.341 1.343 mp_alltoall_d11v 2130 13.8 1.176 1.316 1.176 1.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.307 1.307 cp_fm_diag_elpa_base 50 14.0 1.163 1.215 1.305 1.306 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=62.691000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=699.909091, yerr=9.365843 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 497.422336E+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 9696 492. MP_Sync 52 MP_Alltoall 1938 1523618. 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 0.017 0.039 120.557 120.564 qs_mol_dyn_low 1 2.0 0.003 0.004 115.794 120.225 qs_forces 11 3.9 0.003 0.003 86.453 86.454 qs_energies 11 4.9 0.002 0.008 83.552 83.566 velocity_verlet 10 3.0 0.001 0.002 75.635 75.941 scf_env_do_scf 11 5.9 0.001 0.002 73.349 73.351 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 67.732 67.733 dbcsr_multiply_generic 2055 12.4 0.106 0.112 50.957 51.168 qs_scf_new_mos 99 7.5 0.000 0.001 46.570 46.682 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.570 46.682 ot_scf_mini 99 9.5 0.002 0.003 44.163 44.298 multiply_cannon 2055 13.4 0.185 0.189 42.664 43.269 multiply_cannon_loop 2055 14.4 1.509 1.552 41.693 42.309 md_write_output 11 3.9 0.177 16.888 0.340 32.503 mp_sum_dm 438 4.9 29.038 29.367 29.038 29.367 update_particle_set 20 4.0 0.000 0.000 28.970 29.276 md_output 10 3.0 0.000 0.000 0.294 27.795 ot_mini 99 10.5 0.001 0.001 25.973 26.062 qs_ot_get_derivative 99 11.5 0.001 0.001 19.221 19.325 multiply_cannon_multrec 49320 15.4 12.473 13.115 18.099 18.686 write_trajectory 44 4.9 0.163 15.564 0.163 15.615 rebuild_ks_matrix 110 8.3 0.000 0.001 14.685 14.783 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.025 14.685 14.783 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.872 12.957 mp_waitall_1 220248 16.4 10.393 11.318 10.393 11.318 multiply_cannon_sync_h2d 49320 15.4 10.330 10.924 10.330 10.924 qs_ot_get_p 110 10.4 0.001 0.003 9.873 10.013 init_scf_run 11 5.9 0.000 0.001 8.035 8.035 scf_env_initial_rho_setup 11 6.9 0.001 0.001 8.034 8.035 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.262 7.795 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.180 7.687 apply_single 110 13.6 0.000 0.000 7.180 7.687 multiply_cannon_metrocomm3 49320 15.4 0.078 0.081 6.235 7.400 sum_up_and_integrate 110 10.3 0.036 0.043 7.180 7.204 integrate_v_rspace 110 11.3 0.003 0.004 7.144 7.176 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.603 6.728 calculate_rho_elec 110 8.6 0.021 0.026 6.603 6.727 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.642 6.658 ot_diis_step 99 11.5 0.006 0.006 6.547 6.547 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.821 5.821 dbcsr_mm_accdrv_process 87628 16.1 1.939 2.021 5.500 5.781 init_scf_loop 11 6.9 0.001 0.005 5.590 5.590 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.235 5.286 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.227 5.246 cp_fm_diag_elpa_base 48 14.0 5.212 5.232 5.225 5.244 mp_sum_l 6514 12.8 3.980 4.632 3.980 4.632 mp_max_l 33 2.8 4.433 4.480 4.433 4.480 rs_pw_transfer 902 11.9 0.012 0.014 3.817 4.403 calculate_dm_sparse 110 9.5 0.001 0.001 4.033 4.193 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.166 4.171 wfi_extrapolate 11 7.9 0.001 0.001 4.121 4.122 make_m2s 4110 13.4 0.061 0.066 3.865 3.969 mp_bcast_b_src 157 10.2 3.827 3.967 3.827 3.967 external_control 109 7.1 0.040 3.770 3.865 3.967 density_rs2pw 110 9.6 0.004 0.005 3.437 3.934 make_images 4110 14.4 0.178 0.192 3.769 3.875 calculate_first_density_matrix 1 7.0 0.001 0.003 3.808 3.811 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.762 3.810 grid_integrate_task_list 110 12.3 3.267 3.423 3.267 3.423 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.278 3.418 prepare_preconditioner 11 7.9 0.000 0.001 3.349 3.368 make_preconditioner 11 8.9 0.000 0.001 3.349 3.368 jit_kernel_multiply 13 15.9 3.282 3.366 3.282 3.366 pw_transfer 1331 11.6 0.055 0.069 3.193 3.295 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.104 3.202 make_full_inverse_cholesky 11 9.9 0.000 0.001 3.126 3.177 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.105 3.172 fft_wrap_pw1pw2_140 451 13.1 0.451 0.497 2.636 2.743 potential_pw2rs 110 12.3 0.006 0.007 2.697 2.720 mp_alltoall_d11v 2046 13.8 2.015 2.454 2.015 2.454 mp_waitany 14300 13.8 1.886 2.449 1.886 2.449 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=120.564000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=472.000000, yerr=2.696799 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 585.326592E+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 9695 570. MP_Sync 52 MP_Alltoall 1717 3492339. 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 0.023 0.082 106.846 106.848 qs_mol_dyn_low 1 2.0 0.003 0.004 101.832 106.524 qs_forces 11 3.9 0.003 0.003 71.485 71.485 velocity_verlet 10 3.0 0.001 0.001 67.850 68.480 qs_energies 11 4.9 0.002 0.007 68.158 68.162 scf_env_do_scf 11 5.9 0.001 0.002 58.953 58.957 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 51.332 51.333 dbcsr_multiply_generic 2055 12.4 0.113 0.117 38.056 38.221 qs_scf_new_mos 99 7.5 0.001 0.001 32.840 32.942 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.840 32.941 multiply_cannon 2055 13.4 0.223 0.243 31.525 32.617 md_output 10 3.0 0.000 0.000 0.677 32.310 md_write_output 11 3.9 0.417 19.977 0.672 32.180 ot_scf_mini 99 9.5 0.003 0.003 31.189 31.297 multiply_cannon_loop 2055 14.4 0.920 0.940 30.239 31.243 mp_sum_dm 438 4.9 29.642 30.293 29.642 30.293 update_particle_set 20 4.0 0.000 0.000 29.599 30.229 ot_mini 99 10.5 0.001 0.001 18.176 18.284 multiply_cannon_multrec 24660 15.4 7.624 9.539 14.402 16.078 rebuild_ks_matrix 110 8.3 0.000 0.000 13.789 13.897 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 13.788 13.897 qs_ot_get_derivative 99 11.5 0.001 0.001 12.360 12.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.142 12.240 write_trajectory 44 4.9 0.250 11.994 0.255 12.202 mp_waitall_1 176588 16.5 7.566 10.079 7.566 10.079 multiply_cannon_sync_h2d 24660 15.4 6.969 7.896 6.969 7.896 multiply_cannon_metrocomm3 24660 15.4 0.068 0.071 5.167 7.731 init_scf_loop 11 6.9 0.001 0.004 7.586 7.587 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.572 7.261 apply_single 110 13.6 0.000 0.001 6.572 7.261 dbcsr_mm_accdrv_process 52282 16.1 5.083 6.057 6.617 6.929 init_scf_run 11 5.9 0.000 0.002 6.807 6.808 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.807 6.808 sum_up_and_integrate 110 10.3 0.052 0.058 6.632 6.641 integrate_v_rspace 110 11.3 0.003 0.003 6.579 6.592 qs_ot_get_p 110 10.4 0.001 0.002 6.427 6.554 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.216 6.228 calculate_rho_elec 110 8.6 0.040 0.048 6.216 6.228 ot_diis_step 99 11.5 0.010 0.010 5.767 5.768 prepare_preconditioner 11 7.9 0.000 0.001 5.532 5.549 make_preconditioner 11 8.9 0.000 0.001 5.531 5.549 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.657 5.391 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.100 5.256 mp_max_l 33 2.8 4.692 4.793 4.692 4.793 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.582 4.605 make_m2s 4110 13.4 0.057 0.059 4.118 4.581 make_images 4110 14.4 0.399 0.444 4.010 4.470 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.156 4.156 density_rs2pw 110 9.6 0.004 0.005 3.333 3.801 pw_transfer 1331 11.6 0.067 0.076 3.637 3.779 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.530 3.672 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.656 3.658 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.577 3.587 cp_fm_diag_elpa_base 48 14.0 3.530 3.546 3.574 3.583 rs_pw_transfer 902 11.9 0.012 0.013 3.023 3.524 wfi_extrapolate 11 7.9 0.001 0.001 3.519 3.519 grid_integrate_task_list 110 12.3 3.149 3.332 3.149 3.332 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.280 3.316 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.184 3.241 calculate_dm_sparse 110 9.5 0.001 0.001 3.143 3.176 fft_wrap_pw1pw2_140 451 13.1 0.520 0.535 3.017 3.160 calculate_first_density_matrix 1 7.0 0.008 0.021 3.135 3.136 write_restart 10 4.0 0.059 2.842 0.062 2.844 make_images_data 4110 15.4 0.047 0.050 2.313 2.788 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.046 2.756 fft3d_ps 1111 14.6 1.107 1.330 2.502 2.653 cp_fm_cholesky_invert 11 10.9 2.558 2.565 2.558 2.565 mp_sum_l 6514 12.8 1.844 2.555 1.844 2.555 grid_collocate_task_list 110 9.6 2.063 2.488 2.063 2.488 potential_pw2rs 110 12.3 0.008 0.009 2.415 2.430 mp_bcast_b_src 157 10.2 2.216 2.294 2.216 2.294 external_control 109 7.1 0.046 2.215 2.256 2.280 jit_kernel_multiply 10 16.3 1.181 2.148 1.181 2.148 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=106.848000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=551.818182, yerr=6.712515 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 658.685952E+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 9694 567. 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 1.950 2.078 182.859 182.872 qs_mol_dyn_low 1 2.0 0.003 0.005 173.440 173.464 qs_forces 11 3.9 0.007 0.040 170.535 170.537 qs_energies 11 4.9 0.018 0.072 167.174 167.187 dbcsr_multiply_generic 2055 12.4 0.107 0.121 136.768 137.170 multiply_cannon 2055 13.4 0.214 0.223 117.292 131.722 multiply_cannon_loop 2055 14.4 0.617 0.631 116.136 130.682 scf_env_do_scf 11 5.9 0.001 0.002 101.716 101.720 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 92.866 92.867 qs_scf_new_mos 99 7.5 0.001 0.001 76.484 76.578 qs_scf_loop_do_ot 99 8.5 0.001 0.001 76.484 76.578 ot_scf_mini 99 9.5 0.003 0.003 75.271 75.384 ot_mini 99 10.5 0.001 0.001 65.400 65.514 init_scf_run 11 5.9 0.000 0.001 62.880 62.880 scf_env_initial_rho_setup 11 6.9 0.012 0.027 62.879 62.880 mp_waitall_1 139946 16.5 44.877 62.086 44.877 62.086 qs_ot_get_derivative 99 11.5 0.001 0.001 61.013 61.131 multiply_cannon_multrec 16440 15.4 3.913 4.874 48.671 59.569 calculate_first_density_matrix 1 7.0 0.001 0.003 59.166 59.276 dbcsr_mm_accdrv_process 34862 16.1 36.121 54.446 44.610 54.816 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 36.415 53.108 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 26.551 40.546 velocity_verlet 10 3.0 0.018 0.025 35.079 35.169 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 34.992 34.993 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 34.609 34.693 acc_transpose_blocks 16440 15.4 0.074 0.075 19.431 28.570 acc_transpose_blocks_kernels 16440 16.4 0.212 0.216 18.997 28.137 jit_kernel_transpose 5 15.6 18.786 27.924 18.786 27.924 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 18.328 27.147 apply_single 110 13.6 0.000 0.000 18.328 27.147 calculate_dm_sparse 110 9.5 0.001 0.001 27.006 27.140 jit_kernel_multiply 10 16.2 8.096 26.883 8.096 26.883 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 25.973 26.085 mp_sum_l 6514 12.8 14.907 23.360 14.907 23.360 multiply_cannon_metrocomm1 16440 15.4 0.028 0.029 6.276 21.852 rebuild_ks_matrix 110 8.3 0.000 0.000 12.556 12.684 qs_ks_build_kohn_sham_matrix 110 9.3 0.047 0.122 12.555 12.684 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.023 11.132 qs_rho_update_rho_low 110 7.6 0.048 0.382 6.610 9.652 calculate_rho_elec 110 8.6 0.060 0.071 6.562 9.652 init_scf_loop 11 6.9 0.025 0.099 8.816 8.820 create_qs_kind_set 1 2.0 0.000 0.001 6.641 8.526 read_qs_kind 2 3.0 0.278 8.496 6.641 8.526 prepare_preconditioner 11 7.9 0.000 0.001 6.710 6.742 make_preconditioner 11 8.9 0.002 0.017 6.710 6.742 sum_up_and_integrate 110 10.3 0.061 0.063 6.580 6.613 parser_read_line 2807 4.0 0.001 0.001 6.363 6.579 parser_read_line_low 5 5.0 0.002 0.024 6.362 6.579 broadcast_input_information 5 6.0 0.002 0.002 6.361 6.578 mp_bcast_i_src 20 7.0 6.356 6.573 6.356 6.573 integrate_v_rspace 110 11.3 0.003 0.014 6.518 6.553 make_full_inverse_cholesky 11 9.9 0.001 0.008 6.040 6.440 mp_alltoall_d11v 2046 13.8 2.205 5.400 2.205 5.400 qs_ot_get_p 110 10.4 0.003 0.010 5.232 5.350 density_rs2pw 110 9.6 0.004 0.005 3.205 4.443 make_m2s 4110 13.4 0.049 0.051 4.006 4.374 ot_diis_step 99 11.5 0.011 0.011 4.359 4.359 rs_scatter_matrices 121 9.7 0.390 0.456 1.272 4.323 make_images 4110 14.4 0.394 0.513 3.892 4.258 multiply_cannon_sync_h2d 16440 15.4 3.670 4.200 3.670 4.200 rs_pw_transfer 902 11.9 0.011 0.011 2.671 3.901 pw_transfer 1331 11.6 0.066 0.075 3.802 3.841 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.695 3.736 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=182.872000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=622.727273, yerr=9.705583 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 721.924096E+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 9694 608. 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 0.089 0.133 157.868 157.884 qs_mol_dyn_low 1 2.0 0.003 0.003 156.755 157.383 qs_forces 11 3.9 0.003 0.003 138.884 138.885 qs_energies 11 4.9 0.001 0.001 135.470 135.473 dbcsr_multiply_generic 2055 12.4 0.111 0.117 102.201 102.431 multiply_cannon 2055 13.4 0.248 0.267 94.137 95.703 multiply_cannon_loop 2055 14.4 0.882 0.902 92.736 94.365 scf_env_do_scf 11 5.9 0.000 0.001 84.092 84.094 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 72.567 72.568 multiply_cannon_multrec 24660 15.4 6.173 15.457 48.579 56.652 qs_scf_new_mos 99 7.5 0.001 0.001 56.446 56.548 qs_scf_loop_do_ot 99 8.5 0.001 0.001 56.446 56.548 ot_scf_mini 99 9.5 0.003 0.003 54.830 54.941 dbcsr_mm_accdrv_process 52304 16.0 24.234 36.822 42.255 53.740 velocity_verlet 10 3.0 0.001 0.002 53.022 53.711 init_scf_run 11 5.9 0.000 0.001 48.649 48.650 scf_env_initial_rho_setup 11 6.9 0.001 0.001 48.649 48.650 calculate_first_density_matrix 1 7.0 0.001 0.001 45.585 45.591 ot_mini 99 10.5 0.001 0.001 44.274 44.401 qs_ot_get_derivative 99 11.5 0.001 0.001 40.146 40.258 jit_kernel_multiply 10 15.7 17.688 38.941 17.688 38.941 mp_waitall_1 121746 16.5 28.159 35.344 28.159 35.344 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 18.861 30.010 calculate_dm_sparse 110 9.5 0.001 0.001 26.766 26.795 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 25.387 25.401 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 21.154 22.351 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 22.060 22.061 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 21.144 21.243 multiply_cannon_metrocomm1 24660 15.4 0.034 0.035 6.699 19.176 acc_transpose_blocks 24660 15.4 0.105 0.107 13.463 17.455 md_write_output 11 3.9 0.284 6.800 0.710 16.995 acc_transpose_blocks_kernels 24660 16.4 0.308 0.314 12.809 16.808 mp_sum_dm 438 4.9 15.898 16.606 15.898 16.606 update_particle_set 20 4.0 0.000 0.000 15.865 16.555 jit_kernel_transpose 5 15.6 12.501 16.498 12.501 16.498 md_output 10 3.0 0.000 0.000 0.653 13.703 rebuild_ks_matrix 110 8.3 0.000 0.000 12.259 12.350 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.259 12.349 init_scf_loop 11 6.9 0.000 0.000 11.485 11.486 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.833 10.909 write_trajectory 44 4.9 0.423 10.151 0.425 10.194 prepare_preconditioner 11 7.9 0.000 0.000 9.723 9.738 make_preconditioner 11 8.9 0.000 0.000 9.723 9.738 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.947 9.411 sum_up_and_integrate 110 10.3 0.067 0.071 6.546 6.558 integrate_v_rspace 110 11.3 0.003 0.003 6.478 6.495 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.102 6.124 calculate_rho_elec 110 8.6 0.078 0.081 6.101 6.124 make_m2s 4110 13.4 0.059 0.062 5.365 5.730 qs_ot_get_p 110 10.4 0.001 0.002 5.562 5.725 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.155 5.647 apply_single 110 13.6 0.000 0.000 5.155 5.647 make_images 4110 14.4 0.576 0.698 5.225 5.586 cp_fm_upper_to_full 70 14.2 3.323 4.828 3.323 4.828 mp_alltoall_i22 605 13.7 1.707 4.331 1.707 4.331 dbcsr_complete_redistribute 325 12.2 0.727 0.961 3.111 4.241 ot_diis_step 99 11.5 0.011 0.012 4.092 4.092 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.823 3.837 pw_transfer 1331 11.6 0.066 0.076 3.757 3.790 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.612 3.731 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.650 3.688 density_rs2pw 110 9.6 0.004 0.004 3.026 3.538 grid_integrate_task_list 110 12.3 3.279 3.471 3.279 3.471 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.352 3.353 multiply_cannon_sync_h2d 24660 15.4 3.186 3.297 3.186 3.297 fft_wrap_pw1pw2_140 451 13.1 0.668 0.693 3.132 3.169 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=157.884000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.818182, yerr=8.715902 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 829.104128E+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 9694 649. 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 0.037 142.254 142.255 qs_mol_dyn_low 1 2.0 0.003 0.003 137.442 141.298 qs_forces 11 3.9 0.003 0.003 110.304 110.304 qs_energies 11 4.9 0.003 0.004 106.610 106.616 scf_env_do_scf 11 5.9 0.000 0.001 75.526 75.527 dbcsr_multiply_generic 2055 12.4 0.104 0.108 74.984 75.107 multiply_cannon 2055 13.4 0.246 0.253 67.570 69.660 multiply_cannon_loop 2055 14.4 0.320 0.333 66.259 68.458 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 67.820 67.820 velocity_verlet 10 3.0 0.001 0.002 56.783 58.445 qs_scf_new_mos 99 7.5 0.001 0.001 49.807 49.860 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.807 49.860 ot_scf_mini 99 9.5 0.002 0.002 48.574 48.622 multiply_cannon_multrec 8220 15.4 10.939 18.402 39.211 46.689 dbcsr_mm_accdrv_process 17442 15.9 23.024 34.473 28.141 41.403 ot_mini 99 10.5 0.001 0.001 40.066 40.112 qs_ot_get_derivative 99 11.5 0.001 0.001 36.255 36.304 md_write_output 11 3.9 0.742 11.868 1.798 28.751 init_scf_run 11 5.9 0.000 0.002 27.628 27.628 scf_env_initial_rho_setup 11 6.9 0.001 0.001 27.628 27.628 md_output 10 3.0 0.000 0.000 1.692 27.001 mp_sum_dm 438 4.9 24.952 26.621 24.952 26.621 update_particle_set 20 4.0 0.000 0.000 24.926 26.588 calculate_first_density_matrix 1 7.0 0.013 0.014 24.790 24.804 calculate_dm_sparse 110 9.5 0.001 0.001 21.632 21.689 mp_waitall_1 103326 16.6 14.949 21.469 14.949 21.469 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 20.600 20.625 write_trajectory 44 4.9 1.054 16.860 1.056 16.882 jit_kernel_multiply 7 15.4 4.805 16.047 4.805 16.047 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 13.999 15.016 acc_transpose_blocks 8220 15.4 0.036 0.037 10.470 13.410 acc_transpose_blocks_kernels 8220 16.4 0.111 0.115 10.233 13.168 jit_kernel_transpose 5 15.6 10.122 13.054 10.122 13.054 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 5.599 12.358 rebuild_ks_matrix 110 8.3 0.000 0.000 11.973 12.014 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 11.973 12.014 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 7.184 11.968 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.627 10.663 init_scf_loop 11 6.9 0.000 0.000 7.660 7.660 mp_sum_l 6514 12.8 2.602 6.753 2.602 6.753 sum_up_and_integrate 110 10.3 0.080 0.081 6.566 6.580 integrate_v_rspace 110 11.3 0.003 0.003 6.486 6.500 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.714 6.437 apply_single 110 13.6 0.000 0.000 4.713 6.437 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.288 6.392 calculate_rho_elec 110 8.6 0.114 0.115 6.288 6.392 prepare_preconditioner 11 7.9 0.000 0.000 5.995 6.001 make_preconditioner 11 8.9 0.000 0.000 5.995 6.001 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 5.857 5.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.574 5.643 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.512 5.524 qs_ot_get_p 110 10.4 0.001 0.002 4.595 4.639 make_m2s 4110 13.4 0.039 0.040 4.113 4.361 make_images 4110 14.4 0.638 0.696 3.983 4.234 mp_max_l 33 2.8 3.857 4.115 3.857 4.115 pw_transfer 1331 11.6 0.066 0.071 4.058 4.066 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.950 3.963 ot_diis_step 99 11.5 0.012 0.012 3.790 3.791 grid_integrate_task_list 110 12.3 3.358 3.543 3.358 3.543 fft_wrap_pw1pw2_140 451 13.1 0.834 0.846 3.420 3.443 density_rs2pw 110 9.6 0.004 0.004 3.025 3.397 mp_bcast_b_src 157 10.2 3.058 3.364 3.058 3.364 external_control 109 7.1 0.201 3.215 3.252 3.341 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.274 3.279 multiply_cannon_sync_h2d 8220 15.4 2.894 3.088 2.894 3.088 qs_energies_init_hamiltonians 11 5.9 0.048 0.052 2.999 3.000 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.977 2.978 cp_fm_cholesky_invert 11 10.9 2.885 2.888 2.885 2.888 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=142.255000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.454545, yerr=11.388700 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 1.369010E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.035 185.810 185.810 qs_mol_dyn_low 1 2.0 0.003 0.003 185.361 185.369 qs_forces 11 3.9 0.003 0.004 173.878 173.878 qs_energies 11 4.9 0.027 0.028 169.714 169.716 scf_env_do_scf 11 5.9 0.001 0.001 116.613 116.614 dbcsr_multiply_generic 2055 12.4 0.119 0.123 113.851 113.963 multiply_cannon 2055 13.4 0.346 0.370 95.085 106.905 multiply_cannon_loop 2055 14.4 0.341 0.345 93.199 105.111 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 87.970 87.971 qs_scf_new_mos 99 7.5 0.001 0.001 68.548 68.634 qs_scf_loop_do_ot 99 8.5 0.001 0.001 68.547 68.633 velocity_verlet 10 3.0 0.001 0.001 66.670 68.103 ot_scf_mini 99 9.5 0.002 0.002 66.795 66.859 multiply_cannon_multrec 8220 15.4 13.907 15.784 52.909 63.341 ot_mini 99 10.5 0.001 0.001 56.025 56.087 qs_ot_get_derivative 99 11.5 0.001 0.001 51.322 51.386 init_scf_run 11 5.9 0.001 0.003 48.772 48.773 scf_env_initial_rho_setup 11 6.9 0.014 0.016 48.772 48.772 dbcsr_mm_accdrv_process 11614 15.7 7.433 14.238 38.859 47.415 calculate_first_density_matrix 1 7.0 0.000 0.001 45.283 45.286 jit_kernel_multiply 10 15.3 31.225 44.238 31.225 44.238 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 22.263 34.347 mp_waitall_1 84994 16.7 26.881 33.612 26.881 33.612 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 21.704 30.527 calculate_dm_sparse 110 9.5 0.001 0.001 29.164 29.189 init_scf_loop 11 6.9 0.000 0.000 28.572 28.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 27.706 27.720 prepare_preconditioner 11 7.9 0.000 0.000 26.535 26.542 make_preconditioner 11 8.9 0.000 0.000 26.535 26.542 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.644 26.005 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 16.413 25.515 apply_single 110 13.6 0.000 0.000 16.413 25.514 mp_sum_l 6514 12.8 12.565 21.519 12.565 21.519 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 19.774 19.778 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 19.254 19.275 cp_fm_upper_to_full 70 14.2 12.853 18.458 12.853 18.458 acc_transpose_blocks 8220 15.4 0.035 0.035 11.739 15.710 acc_transpose_blocks_kernels 8220 16.4 0.111 0.112 11.497 15.466 jit_kernel_transpose 5 15.6 11.386 15.355 11.386 15.355 rebuild_ks_matrix 110 8.3 0.000 0.001 14.089 14.160 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.089 14.160 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.766 12.829 mp_sum_dm 438 4.9 10.056 11.498 10.056 11.498 md_write_output 11 3.9 0.743 5.939 1.435 11.470 update_particle_set 20 4.0 0.000 0.000 10.033 11.466 dbcsr_complete_redistribute 325 12.2 1.012 1.038 7.437 10.818 multiply_cannon_metrocomm1 8220 15.4 0.011 0.011 2.428 10.055 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.438 9.813 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.875 9.211 mp_alltoall_i22 605 13.7 5.512 8.901 5.512 8.901 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.700 7.737 calculate_rho_elec 110 8.6 0.225 0.226 7.700 7.736 sum_up_and_integrate 110 10.3 0.150 0.152 7.299 7.313 integrate_v_rspace 110 11.3 0.004 0.004 7.149 7.161 md_output 10 3.0 0.000 0.000 0.854 6.808 make_m2s 4110 13.4 0.043 0.043 5.308 5.815 qs_ot_get_p 110 10.4 0.001 0.001 5.603 5.657 make_images 4110 14.4 0.879 0.929 5.121 5.628 write_trajectory 44 4.9 0.677 5.412 0.692 5.530 cp_fm_cholesky_invert 11 10.9 5.304 5.309 5.304 5.309 pw_transfer 1331 11.6 0.076 0.076 5.039 5.051 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.921 4.933 ot_diis_step 99 11.5 0.015 0.016 4.684 4.684 fft_wrap_pw1pw2_140 451 13.1 1.339 1.344 4.297 4.310 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.044 4.051 multiply_cannon_sync_h2d 8220 15.4 3.943 3.950 3.943 3.950 density_rs2pw 110 9.6 0.004 0.004 3.724 3.756 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=185.810000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1223.545455, yerr=63.799920 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 627.793920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11004 802. MP_Sync 87 MP_Alltoall 2226 3534508. 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 0.022 0.050 348.315 348.319 qs_mol_dyn_low 1 2.0 0.003 0.003 347.708 347.718 qs_forces 11 3.9 0.005 0.005 344.616 344.618 qs_energies 11 4.9 0.002 0.010 339.027 339.038 dbcsr_multiply_generic 2507 12.6 0.179 0.182 261.494 262.289 scf_env_do_scf 11 5.9 0.001 0.001 242.368 242.372 multiply_cannon 2507 13.6 0.237 0.246 235.531 240.183 multiply_cannon_loop 2507 14.6 2.083 2.126 233.345 237.573 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 221.841 221.844 qs_scf_new_mos 117 7.6 0.001 0.001 181.494 181.844 qs_scf_loop_do_ot 117 8.6 0.001 0.001 181.494 181.843 ot_scf_mini 117 9.6 0.003 0.003 174.940 175.229 velocity_verlet 10 3.0 0.140 0.371 126.975 127.007 ot_mini 117 10.6 0.001 0.001 123.400 123.713 multiply_cannon_multrec 60168 15.6 33.549 36.674 98.167 122.170 mp_waitall_1 267128 16.5 82.654 107.905 82.654 107.905 qs_ot_get_derivative 117 11.6 0.001 0.001 98.431 98.738 init_scf_run 11 5.9 0.000 0.001 92.483 92.483 scf_env_initial_rho_setup 11 6.9 0.001 0.001 92.482 92.483 dbcsr_mm_accdrv_process 124484 16.2 3.172 3.282 64.178 87.312 calculate_first_density_matrix 1 7.0 0.000 0.001 83.122 83.152 jit_kernel_multiply 13 15.9 58.936 82.574 58.936 82.574 multiply_cannon_metrocomm3 60168 15.6 0.112 0.117 44.198 78.246 multiply_cannon_metrocomm1 60168 15.6 0.086 0.090 31.864 55.529 calculate_dm_sparse 128 9.5 0.001 0.001 53.509 53.656 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 47.661 47.705 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 41.182 41.193 acc_transpose_blocks 60168 15.6 0.294 0.307 26.414 41.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 40.142 40.259 acc_transpose_blocks_kernels 60168 16.6 0.700 0.722 24.839 39.592 jit_kernel_transpose 5 15.6 24.138 38.890 24.138 38.890 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 34.554 37.048 rebuild_ks_matrix 128 8.3 0.001 0.001 33.821 34.201 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.030 33.820 34.200 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 26.913 31.880 apply_single 128 13.6 0.001 0.001 26.913 31.880 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.370 30.732 qs_ot_get_p 128 10.4 0.001 0.002 29.528 29.867 multiply_cannon_sync_h2d 60168 15.6 27.686 29.747 27.686 29.747 ot_diis_step 117 11.6 0.008 0.008 24.738 24.739 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.987 23.058 cp_dbcsr_syevd 83 12.4 0.005 0.005 20.550 20.551 init_scf_loop 11 6.9 0.000 0.000 20.453 20.455 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.311 19.540 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.606 17.640 cp_fm_diag_elpa_base 83 14.4 17.538 17.573 17.602 17.637 prepare_preconditioner 11 7.9 0.000 0.000 15.835 15.897 make_preconditioner 11 8.9 0.000 0.000 15.835 15.897 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.795 15.569 calculate_rho_elec 128 8.7 0.046 0.065 13.794 15.568 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.052 15.278 make_m2s 5014 13.6 0.103 0.111 14.197 14.493 mp_sum_l 7870 13.0 10.532 14.391 10.532 14.391 make_images 5014 14.6 0.402 0.421 14.017 14.325 sum_up_and_integrate 128 10.3 0.089 0.107 14.256 14.272 integrate_v_rspace 128 11.3 0.004 0.005 14.167 14.186 density_rs2pw 128 9.7 0.006 0.007 7.128 10.312 wfi_extrapolate 11 7.9 0.001 0.001 9.171 9.171 cp_fm_cholesky_invert 11 10.9 8.990 8.998 8.990 8.998 rs_pw_transfer 1046 11.9 0.017 0.019 5.783 8.966 make_images_data 5014 15.6 0.066 0.071 6.843 8.206 pw_transfer 1547 11.6 0.075 0.100 7.841 8.016 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.844 7.943 fft_wrap_pw1pw2 1291 12.7 0.011 0.014 7.638 7.819 hybrid_alltoall_any 5200 16.5 0.290 2.269 5.979 7.692 grid_integrate_task_list 128 12.3 7.031 7.504 7.031 7.504 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=348.319000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.181818, yerr=6.336305 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/16/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 2.166472E+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 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 835.571712E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58319. MP_Allreduce 10986 968. MP_Sync 86 MP_Alltoall 1955 6465122. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.133 0.387 298.676 298.691 qs_mol_dyn_low 1 2.0 0.003 0.005 297.356 297.367 qs_forces 11 3.9 0.039 0.075 296.390 296.400 qs_energies 11 4.9 0.009 0.032 289.567 289.617 scf_env_do_scf 11 5.9 0.001 0.002 245.181 245.191 scf_env_do_scf_inner_loop 116 6.6 0.072 0.274 212.323 212.325 dbcsr_multiply_generic 2485 12.5 0.185 0.192 203.064 204.250 multiply_cannon 2485 13.5 0.472 0.527 167.645 184.603 multiply_cannon_loop 2485 14.5 1.233 1.274 164.782 181.634 qs_scf_new_mos 116 7.6 0.001 0.001 171.711 172.292 qs_scf_loop_do_ot 116 8.6 0.001 0.001 171.710 172.291 ot_scf_mini 116 9.6 0.004 0.005 166.969 167.590 ot_mini 116 10.6 0.003 0.017 128.690 129.224 velocity_verlet 10 3.0 0.005 0.032 118.951 118.970 qs_ot_get_derivative 116 11.6 0.002 0.020 106.952 107.564 mp_waitall_1 212858 16.6 74.933 99.405 74.933 99.405 multiply_cannon_metrocomm3 29820 15.5 0.090 0.096 64.181 89.471 multiply_cannon_multrec 29820 15.5 21.744 25.852 65.096 79.403 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 49.118 65.726 dbcsr_mm_accdrv_process 61726 16.2 23.065 45.352 42.809 59.285 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 37.701 49.362 apply_single 127 13.6 0.001 0.001 37.700 49.362 init_scf_run 11 5.9 0.000 0.001 39.319 39.321 scf_env_initial_rho_setup 11 6.9 0.001 0.002 39.319 39.321 jit_kernel_multiply 12 15.8 16.745 35.357 16.745 35.357 rebuild_ks_matrix 127 8.3 0.001 0.001 32.803 33.315 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.024 32.802 33.314 mp_sum_l 7804 13.0 19.819 32.916 19.819 32.916 init_scf_loop 11 6.9 0.001 0.004 32.771 32.773 calculate_dm_sparse 127 9.5 0.001 0.001 32.075 32.231 calculate_first_density_matrix 1 7.0 0.001 0.003 30.575 30.583 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.483 29.960 prepare_preconditioner 11 7.9 0.000 0.000 28.379 28.463 make_preconditioner 11 8.9 0.004 0.022 28.379 28.463 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 27.909 27.939 make_full_inverse_cholesky 11 9.9 0.002 0.009 27.051 27.626 ot_diis_step 116 11.6 0.020 0.062 21.557 21.583 qs_ot_get_p 127 10.4 0.001 0.002 20.388 21.045 multiply_cannon_sync_h2d 29820 15.5 18.947 21.017 18.947 21.017 cp_fm_cholesky_invert 11 10.9 16.544 16.556 16.544 16.556 qs_ot_p2m_diag 82 11.4 0.188 0.223 15.739 15.773 make_m2s 4970 13.5 0.088 0.094 13.954 15.132 make_images 4970 14.5 1.147 1.335 13.748 14.927 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.617 14.621 sum_up_and_integrate 127 10.3 0.115 0.133 14.494 14.525 integrate_v_rspace 127 11.3 0.003 0.004 14.378 14.416 qs_rho_update_rho_low 127 7.7 0.001 0.002 13.869 13.905 calculate_rho_elec 127 8.7 0.087 0.103 13.869 13.904 acc_transpose_blocks 29820 15.5 0.142 0.145 8.696 12.812 multiply_cannon_metrocomm1 29820 15.5 0.046 0.048 2.475 12.083 acc_transpose_blocks_kernels 29820 16.5 0.400 0.408 7.867 11.984 jit_kernel_transpose 5 15.6 7.467 11.585 7.467 11.585 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.528 11.553 cp_fm_diag_elpa_base 82 14.4 11.279 11.371 11.523 11.546 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.624 11.076 multiply_cannon_metrocomm4 27335 15.5 0.103 0.118 3.794 10.396 mp_irecv_dv 68888 16.3 3.593 10.011 3.593 10.011 make_images_data 4970 15.5 0.064 0.072 8.531 9.913 density_rs2pw 127 9.7 0.006 0.007 7.299 9.717 hybrid_alltoall_any 5155 16.4 0.339 1.500 7.274 9.288 pw_transfer 1535 11.6 0.085 0.095 8.646 8.692 wfi_extrapolate 11 7.9 0.028 0.219 8.563 8.565 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.423 8.473 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 8.213 8.251 rs_pw_transfer 1038 11.9 0.014 0.016 5.659 8.208 fft_wrap_pw1pw2_140 519 13.2 1.324 1.341 7.473 7.544 grid_integrate_task_list 127 12.3 7.112 7.506 7.112 7.506 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.178 7.344 cp_fm_cholesky_decompose 22 10.9 6.874 6.950 6.874 6.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=298.691000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.363636, yerr=3.497343 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 939.913216E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11002 1006. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.068 0.187 285.226 285.227 qs_mol_dyn_low 1 2.0 0.003 0.003 283.984 284.006 qs_forces 11 3.9 0.004 0.005 282.678 282.689 qs_energies 11 4.9 0.007 0.027 276.063 276.081 scf_env_do_scf 11 5.9 0.001 0.002 209.066 209.071 dbcsr_multiply_generic 2507 12.6 0.302 0.309 187.452 188.825 scf_env_do_scf_inner_loop 117 6.6 0.004 0.011 173.621 173.634 multiply_cannon 2507 13.6 0.504 0.525 155.893 170.641 multiply_cannon_loop 2507 14.6 0.857 0.886 152.735 167.656 qs_scf_new_mos 117 7.6 0.001 0.001 134.501 134.843 qs_scf_loop_do_ot 117 8.6 0.001 0.002 134.500 134.842 ot_scf_mini 117 9.6 0.004 0.005 130.380 130.778 velocity_verlet 10 3.0 0.001 0.002 113.954 113.993 ot_mini 117 10.6 0.001 0.001 97.319 97.721 mp_waitall_1 170520 16.6 63.776 88.844 63.776 88.844 qs_ot_get_derivative 117 11.6 0.002 0.002 77.671 78.063 multiply_cannon_metrocomm3 20056 15.6 0.058 0.062 49.642 74.078 multiply_cannon_multrec 20056 15.6 13.557 16.163 62.229 69.278 init_scf_run 11 5.9 0.000 0.001 61.972 61.973 scf_env_initial_rho_setup 11 6.9 0.001 0.001 61.972 61.973 dbcsr_mm_accdrv_process 41502 16.2 30.136 40.531 48.119 56.874 calculate_first_density_matrix 1 7.0 0.001 0.002 54.244 54.362 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 31.372 43.523 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 31.269 40.186 apply_single 128 13.6 0.001 0.001 31.268 40.186 calculate_dm_sparse 128 9.5 0.001 0.001 36.555 36.780 init_scf_loop 11 6.9 0.041 0.166 35.348 35.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 32.963 33.104 rebuild_ks_matrix 128 8.3 0.001 0.001 30.822 31.376 qs_ks_build_kohn_sham_matrix 128 9.3 0.036 0.100 30.821 31.375 prepare_preconditioner 11 7.9 0.000 0.001 30.898 30.956 make_preconditioner 11 8.9 0.017 0.092 30.898 30.956 make_full_inverse_cholesky 11 9.9 0.002 0.009 28.492 29.999 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.831 28.358 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 26.008 26.016 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 24.796 24.935 jit_kernel_multiply 6 15.5 15.053 23.939 15.053 23.939 mp_sum_l 7870 13.0 14.886 23.510 14.886 23.510 acc_transpose_blocks 20056 15.6 0.094 0.096 16.784 20.448 acc_transpose_blocks_kernels 20056 16.6 0.271 0.277 16.198 19.866 jit_kernel_transpose 5 15.6 15.927 19.596 15.927 19.596 ot_diis_step 117 11.6 0.018 0.022 19.550 19.550 qs_ot_get_p 128 10.4 0.015 0.111 18.589 19.061 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.331 17.861 calculate_rho_elec 128 8.7 0.130 0.145 14.330 17.860 multiply_cannon_sync_h2d 20056 15.6 14.296 15.922 14.296 15.922 make_m2s 5014 13.6 0.080 0.085 14.786 15.834 make_images 5014 14.6 1.167 1.271 14.555 15.596 sum_up_and_integrate 128 10.3 0.133 0.145 14.525 14.553 integrate_v_rspace 128 11.3 0.004 0.005 14.392 14.419 cp_fm_cholesky_invert 11 10.9 14.190 14.199 14.190 14.199 qs_ot_p2m_diag 83 11.4 0.266 0.273 13.734 13.740 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.740 12.742 multiply_cannon_metrocomm1 20056 15.6 0.037 0.039 5.286 11.058 make_images_data 5014 15.6 0.059 0.066 9.205 10.656 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.752 9.768 cp_fm_diag_elpa_base 83 14.4 9.343 9.505 9.749 9.765 hybrid_alltoall_any 5200 16.5 0.434 1.979 8.029 9.723 qs_ot_get_derivative_diag 77 12.4 0.004 0.016 8.908 9.190 density_rs2pw 128 9.7 0.006 0.006 7.083 9.104 multiply_cannon_metrocomm4 17549 15.6 0.061 0.071 3.387 9.070 pw_transfer 1547 11.6 0.086 0.106 8.813 8.915 mp_irecv_dv 50230 16.2 3.265 8.820 3.265 8.820 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.588 8.694 mp_alltoall_d11v 2415 14.1 4.251 7.971 4.251 7.971 fft_wrap_pw1pw2_140 523 13.2 1.415 1.445 7.641 7.759 grid_integrate_task_list 128 12.3 7.292 7.733 7.292 7.733 wfi_extrapolate 11 7.9 0.001 0.001 7.430 7.431 cp_fm_upper_to_full 105 14.8 5.878 7.392 5.878 7.392 cp_fm_cholesky_decompose 22 10.9 7.271 7.293 7.271 7.293 rs_pw_transfer 1046 11.9 0.014 0.014 5.044 7.102 dbcsr_complete_redistribute 395 12.7 1.266 1.609 4.784 6.595 rs_scatter_matrices 139 9.7 1.063 1.214 2.320 5.889 fft3d_ps 1291 14.7 2.738 2.964 5.706 5.777 grid_collocate_task_list 128 9.7 4.914 5.763 4.914 5.763 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=285.227000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=891.454545, yerr=9.178667 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/18/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 4.320339E+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 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.138123E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58138. MP_Allreduce 11015 1090. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.058 272.910 272.920 qs_mol_dyn_low 1 2.0 0.003 0.003 271.738 271.877 qs_forces 11 3.9 0.004 0.005 265.730 265.733 qs_energies 11 4.9 0.005 0.008 258.523 258.532 scf_env_do_scf 11 5.9 0.001 0.001 195.348 195.355 dbcsr_multiply_generic 2485 12.5 0.184 0.191 156.918 157.789 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 149.351 149.353 multiply_cannon 2485 13.5 0.554 0.586 130.253 134.407 velocity_verlet 10 3.0 0.001 0.002 129.827 130.066 multiply_cannon_loop 2485 14.5 1.169 1.197 126.429 129.316 qs_scf_new_mos 116 7.6 0.001 0.001 110.437 110.710 qs_scf_loop_do_ot 116 8.6 0.001 0.001 110.436 110.709 ot_scf_mini 116 9.6 0.003 0.003 105.869 106.167 multiply_cannon_multrec 29820 15.5 14.117 19.409 63.709 74.694 ot_mini 116 10.6 0.001 0.001 73.643 73.950 init_scf_run 11 5.9 0.000 0.002 57.264 57.265 scf_env_initial_rho_setup 11 6.9 0.001 0.003 57.263 57.265 mp_waitall_1 146592 16.7 41.637 57.149 41.637 57.149 dbcsr_mm_accdrv_process 61748 16.2 23.735 39.943 49.168 54.872 qs_ot_get_derivative 116 11.6 0.001 0.002 54.311 54.612 calculate_first_density_matrix 1 7.0 0.001 0.001 49.227 49.291 init_scf_loop 11 6.9 0.000 0.000 45.876 45.879 prepare_preconditioner 11 7.9 0.000 0.000 41.755 41.783 make_preconditioner 11 8.9 0.000 0.000 41.755 41.783 jit_kernel_multiply 13 15.8 22.707 41.355 22.707 41.355 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 23.955 40.630 make_full_inverse_cholesky 11 9.9 0.012 0.023 35.440 40.401 calculate_dm_sparse 127 9.5 0.001 0.001 30.651 30.794 rebuild_ks_matrix 127 8.3 0.001 0.001 29.861 30.149 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 29.861 30.148 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 27.430 27.475 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.869 27.127 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 26.739 26.813 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 26.198 26.378 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 22.265 24.369 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 20.790 22.861 apply_single 127 13.6 0.001 0.001 20.789 22.861 make_m2s 4970 13.5 0.094 0.098 20.260 21.180 acc_transpose_blocks 29820 15.5 0.135 0.138 15.511 21.017 make_images 4970 14.5 1.933 2.212 19.958 20.879 acc_transpose_blocks_kernels 29820 16.5 0.390 0.396 14.639 20.157 jit_kernel_transpose 5 15.6 14.248 19.773 14.248 19.773 ot_diis_step 116 11.6 0.017 0.018 19.204 19.205 qs_ot_get_p 127 10.4 0.001 0.002 18.574 18.897 multiply_cannon_metrocomm1 29820 15.5 0.045 0.046 7.267 18.112 cp_fm_upper_to_full 104 14.8 11.345 16.690 11.345 16.690 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.834 16.153 calculate_rho_elec 127 8.7 0.172 0.187 14.833 16.152 cp_fm_cholesky_invert 11 10.9 15.876 15.885 15.876 15.885 sum_up_and_integrate 127 10.3 0.139 0.149 14.654 14.680 integrate_v_rspace 127 11.3 0.004 0.004 14.515 14.544 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.409 14.460 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.155 13.157 dbcsr_complete_redistribute 393 12.7 1.509 1.627 8.904 12.704 multiply_cannon_sync_h2d 29820 15.5 11.679 12.644 11.679 12.644 make_images_data 4970 15.5 0.062 0.066 11.006 12.620 hybrid_alltoall_any 5155 16.4 0.521 2.186 9.630 11.655 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.553 11.306 mp_alltoall_i22 712 14.1 5.688 10.671 5.688 10.671 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.131 10.144 cp_fm_diag_elpa_base 82 14.4 9.189 9.498 10.125 10.137 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.292 9.969 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.357 9.571 pw_transfer 1535 11.6 0.086 0.101 9.094 9.169 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.870 8.951 density_rs2pw 127 9.7 0.006 0.006 6.945 8.424 fft_wrap_pw1pw2_140 519 13.2 1.548 1.578 7.895 7.993 grid_integrate_task_list 127 12.3 7.486 7.878 7.486 7.878 wfi_extrapolate 11 7.9 0.001 0.001 7.773 7.774 cp_fm_cholesky_decompose 22 10.9 7.468 7.581 7.468 7.581 mp_sum_l 7804 13.0 4.476 7.323 4.476 7.323 multiply_cannon_metrocomm4 24850 15.5 0.079 0.090 2.676 6.678 mp_alltoall_d11v 2401 14.1 5.224 6.600 5.224 6.600 mp_irecv_dv 75445 16.2 2.531 6.406 2.531 6.406 rs_pw_transfer 1038 11.9 0.013 0.014 4.630 6.183 fft3d_ps 1281 14.7 2.814 2.894 5.818 5.868 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.043 5.865 mp_sum_dm 438 4.9 5.588 5.859 5.588 5.859 grid_collocate_task_list 127 9.7 5.032 5.833 5.032 5.833 update_particle_set 20 4.0 0.000 0.000 5.520 5.760 md_write_output 11 3.9 0.075 1.795 0.241 5.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=272.920000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1067.818182, yerr=20.718100 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.530479E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+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 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57898. MP_Allreduce 11057 1172. MP_Sync 87 MP_Alltoall 1724 18848418. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 247.558 247.559 qs_mol_dyn_low 1 2.0 0.003 0.003 246.924 246.937 qs_forces 11 3.9 0.004 0.005 239.962 239.966 qs_energies 11 4.9 0.001 0.002 232.112 232.121 scf_env_do_scf 11 5.9 0.001 0.001 168.225 168.230 dbcsr_multiply_generic 2527 12.6 0.597 0.611 139.841 140.572 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.802 132.803 multiply_cannon 2527 13.6 0.588 0.613 118.943 123.655 velocity_verlet 10 3.0 0.002 0.002 120.858 121.289 multiply_cannon_loop 2527 14.6 0.447 0.462 114.230 115.547 qs_scf_new_mos 118 7.6 0.001 0.001 92.677 92.778 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.677 92.778 ot_scf_mini 118 9.6 0.003 0.004 88.213 88.331 multiply_cannon_multrec 10108 15.6 12.566 23.363 59.343 65.519 ot_mini 118 10.6 0.001 0.001 58.739 58.841 init_scf_run 11 5.9 0.000 0.002 57.058 57.058 scf_env_initial_rho_setup 11 6.9 0.001 0.001 57.057 57.058 dbcsr_mm_accdrv_process 20926 16.1 10.666 17.738 46.402 52.775 calculate_first_density_matrix 1 7.0 0.001 0.001 49.342 49.353 mp_waitall_1 126780 16.7 35.311 47.817 35.311 47.817 jit_kernel_multiply 12 15.4 33.135 43.700 33.135 43.700 qs_ot_get_derivative 118 11.6 0.002 0.002 38.452 38.562 init_scf_loop 11 6.9 0.000 0.000 35.279 35.281 calculate_dm_sparse 129 9.5 0.001 0.001 32.044 32.125 prepare_preconditioner 11 7.9 0.000 0.000 31.286 31.313 make_preconditioner 11 8.9 0.000 0.000 31.286 31.313 rebuild_ks_matrix 129 8.3 0.001 0.001 30.296 30.438 qs_ks_build_kohn_sham_matrix 129 9.3 0.155 0.294 30.295 30.437 make_full_inverse_cholesky 11 9.9 0.016 0.029 29.243 29.493 multiply_cannon_metrocomm3 10108 15.6 0.024 0.025 16.915 29.039 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 28.262 28.297 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.262 27.392 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 25.865 25.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 24.534 24.701 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.422 23.632 apply_single 129 13.6 0.001 0.001 20.422 23.632 ot_diis_step 118 11.6 0.020 0.021 20.219 20.219 acc_transpose_blocks 10108 15.6 0.046 0.048 14.301 18.695 make_m2s 5054 13.6 0.067 0.073 16.143 18.571 acc_transpose_blocks_kernels 10108 16.6 0.139 0.142 13.952 18.346 make_images 5054 14.6 2.309 2.804 15.835 18.252 jit_kernel_transpose 5 15.6 13.813 18.208 13.813 18.208 cp_fm_cholesky_invert 11 10.9 17.929 17.936 17.929 17.936 qs_ot_get_p 129 10.4 0.001 0.002 16.447 16.567 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.347 15.438 calculate_rho_elec 129 8.7 0.258 0.269 15.346 15.438 sum_up_and_integrate 129 10.3 0.183 0.193 15.044 15.090 multiply_cannon_metrocomm1 10108 15.6 0.029 0.029 9.412 14.971 integrate_v_rspace 129 11.3 0.004 0.004 14.861 14.915 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 12.484 13.217 qs_ot_p2m_diag 83 11.4 0.496 0.501 12.711 12.728 make_images_data 5054 15.6 0.052 0.060 9.795 12.615 multiply_cannon_sync_h2d 10108 15.6 11.617 12.443 11.617 12.443 hybrid_alltoall_any 5240 16.5 0.851 3.840 9.674 12.112 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.604 11.606 pw_transfer 1559 11.6 0.087 0.098 9.478 9.497 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.251 9.274 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.648 8.660 cp_fm_diag_elpa_base 83 14.4 8.416 8.486 8.645 8.657 fft_wrap_pw1pw2_140 527 13.2 1.933 1.974 8.213 8.239 grid_integrate_task_list 129 12.3 7.813 8.166 7.813 8.166 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.844 7.925 cp_fm_cholesky_decompose 22 10.9 7.779 7.884 7.779 7.884 density_rs2pw 129 9.7 0.006 0.006 6.837 7.741 wfi_extrapolate 11 7.9 0.001 0.001 7.490 7.490 mp_allgather_i34 2527 14.6 2.905 6.982 2.905 6.982 mp_sum_dm 438 4.9 6.524 6.970 6.524 6.970 md_write_output 11 3.9 0.269 4.303 0.434 6.918 update_particle_set 20 4.0 0.000 0.000 6.481 6.912 grid_collocate_task_list 129 9.7 5.442 6.106 5.442 6.106 fft3d_ps 1301 14.7 2.787 2.847 5.693 5.742 mp_alltoall_d11v 2423 14.1 4.968 5.715 4.968 5.715 dbcsr_complete_redistribute 395 12.7 2.189 2.277 5.152 5.544 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.614 5.428 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.902 5.330 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.269 5.270 rs_pw_transfer 1054 12.0 0.013 0.013 4.199 5.248 mp_irecv_dv 29086 15.9 1.865 5.243 1.865 5.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=247.559000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1421.272727, yerr=39.758569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456868864 0.0% 0.0% 100.0% flops 32 x 32 x 32 1971389988864 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992007581696 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753954734080 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613074706432 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176888320 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176888320 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911117062144 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.241596E+12 0.0% 0.0% 100.0% flops max/rank 11.787141E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806515136 0.0% 0.0% 100.0% number of processed stacks 1981760 0.0% 0.0% 100.0% average stack size 0.0 0.0 3434.6 marketing flops 145.659521E+12 ------------------------------------------------------------------------------- # multiplications 2533 max memory usage/rank 3.069432E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101320 MPI messages size (bytes): total size 1.145104E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.301851E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45808 35567697920 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592330608 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4065 58259. MP_Allreduce 11160 1510. MP_Sync 86 MP_Alltoall 1724 36993608. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.038 391.538 391.538 qs_mol_dyn_low 1 2.0 0.003 0.003 390.535 390.547 qs_forces 11 3.9 0.004 0.005 389.153 389.157 qs_energies 11 4.9 0.002 0.002 379.875 379.882 scf_env_do_scf 11 5.9 0.001 0.001 307.595 307.609 velocity_verlet 10 3.0 0.001 0.001 218.444 218.612 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 183.898 183.901 dbcsr_multiply_generic 2533 12.6 0.214 0.225 173.404 174.318 multiply_cannon 2533 13.6 0.729 0.795 140.844 146.845 multiply_cannon_loop 2533 14.6 0.475 0.480 137.081 143.258 qs_scf_new_mos 118 7.6 0.001 0.001 130.326 130.513 qs_scf_loop_do_ot 118 8.6 0.001 0.001 130.325 130.512 ot_scf_mini 118 9.6 0.004 0.004 125.438 125.551 init_scf_loop 11 6.9 0.000 0.000 123.443 123.448 prepare_preconditioner 11 7.9 0.000 0.000 118.330 118.363 make_preconditioner 11 8.9 0.000 0.000 118.330 118.363 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.768 115.419 ot_mini 118 10.6 0.001 0.001 84.713 84.857 cp_fm_upper_to_full 104 14.8 52.484 74.930 52.484 74.930 multiply_cannon_multrec 10132 15.6 15.045 21.960 62.107 65.207 qs_ot_get_derivative 118 11.6 0.002 0.002 63.709 63.819 init_scf_run 11 5.9 0.000 0.001 62.183 62.183 scf_env_initial_rho_setup 11 6.9 0.001 0.001 62.182 62.183 mp_waitall_1 104740 16.8 48.168 60.666 48.168 60.666 dbcsr_mm_accdrv_process 20950 16.1 10.913 23.201 46.821 55.626 calculate_first_density_matrix 1 7.0 0.001 0.001 52.741 52.755 multiply_cannon_metrocomm3 10132 15.6 0.023 0.025 38.016 50.188 jit_kernel_multiply 11 15.4 34.149 42.629 34.149 42.629 dbcsr_complete_redistribute 393 12.7 3.985 4.034 29.157 41.796 rebuild_ks_matrix 129 8.3 0.001 0.001 38.713 38.882 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 38.712 38.882 copy_fm_to_dbcsr 208 11.6 0.001 0.002 25.835 38.432 transfer_fm_to_dbcsr 11 9.9 0.029 0.032 23.516 36.000 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.754 35.918 qs_ot_get_derivative_taylor 42 13.0 0.001 0.001 26.871 34.713 mp_alltoall_i22 716 14.1 21.433 34.196 21.433 34.196 cp_fm_cholesky_invert 11 10.9 33.162 33.168 33.162 33.168 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 31.400 31.458 calculate_dm_sparse 129 9.5 0.001 0.001 31.250 31.342 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 27.449 29.735 apply_single 129 13.6 0.001 0.001 27.448 29.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 29.521 29.646 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 27.359 27.386 qs_ot_get_p 129 10.4 0.001 0.001 25.107 25.256 make_m2s 5066 13.6 0.077 0.080 20.312 21.291 qs_ot_p2m_diag 82 11.4 0.868 0.874 20.956 20.983 ot_diis_step 118 11.6 0.022 0.024 20.959 20.960 make_images 5066 14.6 3.802 3.988 19.828 20.810 acc_transpose_blocks 10132 15.6 0.046 0.047 19.248 20.729 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.579 20.626 calculate_rho_elec 129 8.7 0.485 0.486 20.578 20.625 acc_transpose_blocks_kernels 10132 16.6 0.139 0.141 18.883 20.362 jit_kernel_transpose 5 15.6 18.744 20.222 18.744 20.222 sum_up_and_integrate 129 10.3 0.324 0.325 19.734 19.821 integrate_v_rspace 129 11.3 0.004 0.004 19.409 19.495 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.261 19.262 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.143 16.143 cp_fm_diag_elpa_base 82 14.4 11.821 13.387 16.139 16.139 multiply_cannon_sync_h2d 10132 15.6 15.771 15.783 15.771 15.783 hybrid_alltoall_any 5251 16.5 1.310 3.057 11.198 13.079 make_images_data 5066 15.6 0.060 0.065 11.004 12.853 pw_transfer 1559 11.6 0.094 0.095 11.413 11.423 mp_sum_l 7948 13.1 9.523 11.243 9.523 11.243 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.175 11.186 mp_alltoall_d11v 2417 14.1 8.405 10.170 8.405 10.170 fft_wrap_pw1pw2_140 527 13.2 3.100 3.123 9.952 9.958 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.326 9.417 wfi_extrapolate 11 7.9 0.001 0.001 9.158 9.158 cp_fm_cholesky_decompose 22 10.9 8.833 8.862 8.833 8.862 grid_integrate_task_list 129 12.3 8.657 8.831 8.657 8.831 density_rs2pw 129 9.7 0.006 0.006 7.970 8.104 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.028 8.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=391.538000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2801.545455, yerr=150.417216 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 1.255981E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_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 2.527 2.567 130.458 130.459 qs_energies 1 2.0 0.000 0.000 120.693 120.703 ls_scf 1 3.0 0.000 0.000 119.723 119.734 dbcsr_multiply_generic 111 6.7 0.014 0.015 107.209 107.373 multiply_cannon 111 7.7 0.017 0.020 86.636 91.382 multiply_cannon_loop 111 8.7 0.210 0.227 83.214 87.883 ls_scf_init_scf 1 4.0 0.000 0.000 63.055 63.085 ls_scf_init_matrix_S 1 5.0 0.000 0.000 61.957 62.009 ls_scf_main 1 4.0 0.000 0.000 53.593 53.611 mp_waitall_1 11031 10.9 34.465 48.328 34.465 48.328 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 47.409 47.427 density_matrix_trs4 2 5.0 0.002 0.003 46.627 46.715 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 13.120 28.381 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 14.112 27.926 multiply_cannon_multrec 2664 9.7 8.150 8.893 20.345 25.065 compute_matrix_preconditioner 1 6.0 0.000 0.001 23.469 23.473 acc_transpose_blocks 2664 9.7 0.019 0.021 14.029 19.805 acc_transpose_blocks_kernels 2664 10.7 0.037 0.041 13.919 19.696 jit_kernel_transpose 1 13.0 13.882 19.659 13.882 19.659 apply_matrix_preconditioner 6 5.3 0.000 0.000 17.339 17.507 dbcsr_mm_accdrv_process 4760 10.4 0.515 0.640 11.816 16.361 mp_sum_l 807 5.4 7.022 15.692 7.022 15.692 multiply_cannon_sync_h2d 2664 9.7 13.725 15.037 13.725 15.037 make_m2s 222 7.7 0.008 0.010 13.014 13.510 make_images 222 8.7 0.099 0.111 12.992 13.490 jit_kernel_multiply 1 12.2 5.027 9.501 5.027 9.501 create_qs_kind_set 1 2.0 0.000 0.000 6.840 9.305 read_qs_kind 2 3.0 0.107 9.186 6.840 9.305 make_images_data 222 9.7 0.004 0.005 7.567 8.180 hybrid_alltoall_any 227 10.6 0.216 1.857 6.461 8.046 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.188 7.668 dbcsr_mm_accdrv_process_sort 4760 11.4 6.274 7.185 6.274 7.185 parser_read_line 3036 4.0 0.002 0.002 6.733 6.849 parser_read_line_low 5 5.0 0.003 0.099 6.731 6.848 broadcast_input_information 5 6.0 0.012 0.013 6.728 6.846 mp_bcast_i_src 20 7.0 6.709 6.827 6.709 6.827 calculate_norms 4752 9.8 5.532 6.161 5.532 6.161 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.056 3.651 mp_irecv_dv 6231 10.9 2.039 3.617 2.039 3.617 make_images_sizes 222 9.7 0.000 0.000 0.749 3.238 mp_alltoall_i44 222 10.7 0.749 3.238 0.749 3.238 arnoldi_extremal 4 6.8 0.000 0.000 3.202 3.223 arnoldi_normal_ev 4 7.8 0.001 0.003 3.202 3.223 build_subspace 16 8.4 0.009 0.012 3.102 3.104 ls_scf_post 1 4.0 0.000 0.000 3.074 3.085 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.894 dbcsr_special_finalize 555 9.7 0.005 0.006 2.344 2.765 dbcsr_merge_single_wm 555 10.7 0.458 0.593 2.336 2.757 make_images_pack 222 9.7 2.208 2.630 2.210 2.632 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=130.459000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1127.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 2.105487E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.054 93.485 93.486 qs_energies 1 2.0 0.000 0.000 92.337 92.341 ls_scf 1 3.0 0.000 0.000 90.729 90.732 dbcsr_multiply_generic 111 6.7 0.016 0.017 76.447 76.894 multiply_cannon 111 7.7 0.029 0.044 53.953 58.087 ls_scf_main 1 4.0 0.000 0.000 55.262 55.268 multiply_cannon_loop 111 8.7 0.116 0.123 50.704 54.742 density_matrix_trs4 2 5.0 0.002 0.003 49.281 49.465 ls_scf_init_scf 1 4.0 0.000 0.000 31.954 31.955 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.736 30.877 mp_waitall_1 9105 10.9 21.421 30.649 21.421 30.649 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.624 27.648 multiply_cannon_multrec 1332 9.7 13.242 16.183 22.706 25.869 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.843 21.437 make_m2s 222 7.7 0.006 0.007 15.453 16.138 make_images 222 8.7 1.577 1.961 15.423 16.109 dbcsr_mm_accdrv_process 4041 10.4 0.454 0.834 9.056 10.710 dbcsr_mm_accdrv_process_sort 4041 11.4 8.463 9.990 8.463 9.990 make_images_data 222 9.7 0.004 0.004 8.932 9.874 hybrid_alltoall_any 227 10.6 0.521 2.481 8.236 9.400 mp_sum_l 807 5.4 5.756 8.817 5.756 8.817 multiply_cannon_metrocomm4 1221 9.7 0.007 0.008 3.226 7.573 mp_irecv_dv 3311 11.0 3.205 7.515 3.205 7.515 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.286 6.853 calculate_norms 2376 9.8 6.005 6.686 6.005 6.686 multiply_cannon_sync_h2d 1332 9.7 4.832 5.960 4.832 5.960 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.727 5.907 arnoldi_extremal 4 6.8 0.000 0.000 4.676 4.708 arnoldi_normal_ev 4 7.8 0.001 0.005 4.676 4.708 build_subspace 16 8.4 0.014 0.021 4.413 4.416 ls_scf_post 1 4.0 0.000 0.000 3.512 3.517 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.122 3.348 ls_scf_store_result 1 5.0 0.000 0.000 3.237 3.340 dbcsr_matrix_vector_mult_local 304 10.0 2.732 3.214 2.734 3.216 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.577 2.666 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.280 2.623 mp_allgather_i34 111 8.7 0.955 2.450 0.955 2.450 make_images_pack 222 9.7 2.017 2.415 2.019 2.417 dbcsr_sort_data 436 11.2 1.799 2.055 1.799 2.055 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=93.486000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 2.786091E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.048 0.062 113.891 113.892 qs_energies 1 2.0 0.000 0.000 112.929 112.935 ls_scf 1 3.0 0.000 0.000 111.544 111.552 dbcsr_multiply_generic 111 6.7 0.016 0.020 94.725 95.033 multiply_cannon 111 7.7 0.052 0.143 67.328 72.697 multiply_cannon_loop 111 8.7 0.100 0.105 63.726 69.684 ls_scf_main 1 4.0 0.000 0.000 58.691 58.695 density_matrix_trs4 2 5.0 0.002 0.003 51.071 51.258 ls_scf_init_scf 1 4.0 0.000 0.000 49.300 49.303 ls_scf_init_matrix_S 1 5.0 0.000 0.000 48.125 48.199 mp_waitall_1 7281 11.0 33.230 44.049 33.230 44.049 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 35.220 35.230 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 18.640 33.519 multiply_cannon_multrec 888 9.7 12.662 15.464 24.345 30.221 make_m2s 222 7.7 0.006 0.007 16.756 18.062 make_images 222 8.7 1.972 2.260 16.718 18.022 mp_sum_l 807 5.4 9.082 15.779 9.082 15.779 apply_matrix_preconditioner 6 5.3 0.000 0.000 15.441 15.651 dbcsr_mm_accdrv_process 3754 10.4 1.686 7.225 11.205 15.547 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 5.662 13.948 make_images_data 222 9.7 0.003 0.004 9.546 10.601 hybrid_alltoall_any 227 10.6 0.618 2.849 9.218 10.554 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.285 10.451 dbcsr_mm_accdrv_process_sort 3754 11.4 7.747 8.939 7.747 8.939 compute_matrix_preconditioner 1 6.0 0.000 0.001 8.909 8.915 multiply_cannon_sync_h2d 888 9.7 6.046 7.139 6.046 7.139 jit_kernel_multiply 1 12.0 1.772 7.119 1.772 7.119 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.437 6.781 mp_irecv_dv 2335 11.1 2.421 6.736 2.421 6.736 acc_transpose_blocks 888 9.7 0.005 0.005 1.910 5.360 acc_transpose_blocks_kernels 888 10.7 0.012 0.013 1.870 5.316 jit_kernel_transpose 1 13.0 1.858 5.304 1.858 5.304 arnoldi_extremal 4 6.8 0.000 0.000 5.036 5.051 arnoldi_normal_ev 4 7.8 0.001 0.006 5.035 5.051 build_subspace 16 8.4 0.014 0.020 4.727 4.737 calculate_norms 1584 9.8 4.261 4.621 4.261 4.621 mp_allgather_i34 111 8.7 1.446 3.767 1.446 3.767 dbcsr_matrix_vector_mult 304 9.0 0.009 0.022 3.410 3.742 dbcsr_matrix_vector_mult_local 304 10.0 3.006 3.583 3.008 3.585 ls_scf_post 1 4.0 0.000 0.000 3.554 3.561 ls_scf_store_result 1 5.0 0.000 0.000 3.277 3.339 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.863 2.994 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=113.892000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2185.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 3.337347E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.167 122.843 122.845 qs_energies 1 2.0 0.000 0.000 121.403 121.415 ls_scf 1 3.0 0.000 0.002 119.776 119.782 dbcsr_multiply_generic 111 6.7 0.017 0.020 102.758 103.034 multiply_cannon 111 7.7 0.055 0.097 67.234 78.083 multiply_cannon_loop 111 8.7 0.115 0.127 62.195 71.709 ls_scf_main 1 4.0 0.019 0.099 59.143 59.144 ls_scf_init_scf 1 4.0 0.000 0.001 57.387 57.390 ls_scf_init_matrix_S 1 5.0 0.000 0.002 56.030 56.091 density_matrix_trs4 2 5.0 0.003 0.014 52.394 52.520 mp_waitall_1 6369 11.0 30.814 46.186 30.814 46.186 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 43.519 43.533 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 14.828 32.853 multiply_cannon_multrec 1332 9.7 14.181 17.556 25.621 28.802 make_m2s 222 7.7 0.007 0.008 21.035 22.475 make_images 222 8.7 3.138 3.599 20.985 22.426 mp_sum_l 807 5.4 12.938 20.055 12.938 20.055 compute_matrix_preconditioner 1 6.0 0.001 0.001 15.411 15.417 apply_matrix_preconditioner 6 5.3 0.000 0.000 14.595 14.762 dbcsr_mm_accdrv_process 3641 10.4 0.774 4.625 11.082 14.415 make_images_data 222 9.7 0.004 0.004 11.746 13.365 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 5.072 12.988 hybrid_alltoall_any 227 10.6 0.798 3.753 11.091 12.749 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 7.687 12.153 dbcsr_mm_accdrv_process_sort 3641 11.4 7.167 8.625 7.167 8.625 acc_transpose_blocks 1332 9.7 0.007 0.008 4.167 7.445 acc_transpose_blocks_kernels 1332 10.7 0.017 0.018 4.106 7.386 jit_kernel_transpose 1 13.0 4.089 7.369 4.089 7.369 jit_kernel_multiply 1 12.2 3.141 6.870 3.141 6.870 multiply_cannon_sync_h2d 1332 9.7 5.457 6.196 5.457 6.196 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 6.167 mp_irecv_dv 3229 10.9 2.060 6.087 2.060 6.087 arnoldi_extremal 4 6.8 0.000 0.000 5.191 5.207 arnoldi_normal_ev 4 7.8 0.012 0.035 5.191 5.207 build_subspace 16 8.4 0.014 0.021 4.829 4.837 calculate_norms 2376 9.8 4.230 4.635 4.230 4.635 mp_allgather_i34 111 8.7 2.161 4.440 2.161 4.440 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.548 3.841 dbcsr_matrix_vector_mult_local 304 10.0 3.171 3.660 3.173 3.662 dbcsr_sort_data 658 11.4 3.061 3.348 3.061 3.348 ls_scf_post 1 4.0 0.000 0.002 3.246 3.251 dbcsr_special_finalize 555 9.7 0.006 0.007 2.798 3.157 dbcsr_merge_single_wm 555 10.7 0.535 0.649 2.789 3.149 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.011 3.063 ls_scf_store_result 1 5.0 0.000 0.000 2.941 2.999 dbcsr_data_release 10477 10.7 1.594 2.464 1.594 2.464 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=122.845000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2717.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 4.649370E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.054 118.972 118.973 qs_energies 1 2.0 0.000 0.000 117.684 117.689 ls_scf 1 3.0 0.000 0.000 115.355 115.365 dbcsr_multiply_generic 111 6.7 0.018 0.019 95.134 95.362 multiply_cannon 111 7.7 0.089 0.185 69.157 75.661 multiply_cannon_loop 111 8.7 0.069 0.076 64.719 68.947 ls_scf_main 1 4.0 0.000 0.000 64.134 64.135 density_matrix_trs4 2 5.0 0.002 0.003 55.182 55.264 ls_scf_init_scf 1 4.0 0.000 0.000 47.583 47.587 ls_scf_init_matrix_S 1 5.0 0.000 0.000 46.337 46.370 mp_waitall_1 5436 11.0 30.328 38.945 30.328 38.945 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 36.178 36.187 multiply_cannon_multrec 444 9.7 14.205 16.329 26.720 29.837 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 13.446 22.685 make_m2s 222 7.7 0.004 0.005 17.816 20.240 make_images 222 8.7 3.724 4.434 17.754 20.181 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 7.514 20.011 dbcsr_mm_accdrv_process 3003 10.4 1.655 6.603 12.216 14.200 mp_sum_l 807 5.4 6.204 13.952 6.204 13.952 apply_matrix_preconditioner 6 5.3 0.000 0.000 12.595 12.727 make_images_data 222 9.7 0.003 0.004 10.172 12.415 hybrid_alltoall_any 227 10.6 0.792 3.793 9.923 12.257 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.602 9.869 compute_matrix_preconditioner 1 6.0 0.001 0.002 9.403 9.410 jit_kernel_multiply 1 12.1 4.083 7.956 4.083 7.956 multiply_cannon_sync_h2d 444 9.7 6.541 7.859 6.541 7.859 dbcsr_mm_accdrv_process_sort 3003 11.4 6.478 7.574 6.478 7.574 acc_transpose_blocks 444 9.7 0.002 0.003 4.876 7.346 acc_transpose_blocks_kernels 444 10.7 0.006 0.007 4.838 7.308 jit_kernel_transpose 1 13.0 4.832 7.302 4.832 7.302 mp_allgather_i34 111 8.7 2.669 6.993 2.669 6.993 arnoldi_extremal 4 6.8 0.000 0.000 5.800 5.811 arnoldi_normal_ev 4 7.8 0.002 0.005 5.800 5.811 build_subspace 16 8.4 0.015 0.020 5.412 5.420 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.143 4.349 dbcsr_matrix_vector_mult_local 304 10.0 3.669 4.149 3.671 4.151 calculate_norms 792 9.8 3.533 3.679 3.533 3.679 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.621 3.654 ls_scf_post 1 4.0 0.000 0.000 3.639 3.643 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.520 3.620 mp_irecv_dv 1241 11.2 1.604 3.613 1.604 3.613 make_images_sizes 222 9.7 0.000 0.000 1.002 3.514 mp_alltoall_i44 222 10.7 1.002 3.514 1.002 3.514 ls_scf_store_result 1 5.0 0.000 0.000 3.409 3.465 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=118.973000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3634.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_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 8.704700E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.082 0.092 130.832 130.833 qs_energies 1 2.0 0.000 0.000 128.581 128.600 ls_scf 1 3.0 0.000 0.000 125.653 125.672 dbcsr_multiply_generic 111 6.7 0.041 0.076 98.419 98.537 multiply_cannon 111 7.7 0.154 0.228 66.294 70.544 multiply_cannon_loop 111 8.7 0.067 0.069 62.949 68.309 ls_scf_main 1 4.0 0.000 0.000 66.123 66.125 density_matrix_trs4 2 5.0 0.002 0.003 56.351 56.416 ls_scf_init_scf 1 4.0 0.000 0.000 55.879 55.880 ls_scf_init_matrix_S 1 5.0 0.000 0.000 54.322 54.341 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 43.689 43.697 mp_waitall_1 4527 11.1 28.990 34.846 28.990 34.846 multiply_cannon_multrec 444 9.7 17.880 18.727 26.294 28.011 make_m2s 222 7.7 0.005 0.005 23.913 24.823 make_images 222 8.7 4.589 4.973 23.807 24.716 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 16.253 22.440 hybrid_alltoall_any 227 10.6 1.662 3.644 12.969 15.807 make_images_data 222 9.7 0.003 0.003 13.143 15.479 compute_matrix_preconditioner 1 6.0 0.002 0.002 14.601 14.611 apply_matrix_preconditioner 6 5.3 0.000 0.000 12.799 13.054 dbcsr_mm_accdrv_process 1814 10.4 3.092 6.018 8.248 10.255 multiply_cannon_sync_h2d 444 9.7 8.846 8.897 8.846 8.897 mp_sum_l 807 5.4 5.861 8.305 5.861 8.305 arnoldi_extremal 4 6.8 0.000 0.000 7.386 7.388 arnoldi_normal_ev 4 7.8 0.003 0.009 7.386 7.388 acc_transpose_blocks 444 9.7 0.002 0.002 6.114 7.255 acc_transpose_blocks_kernels 444 10.7 0.005 0.005 6.079 7.219 jit_kernel_transpose 1 13.0 6.074 7.214 6.074 7.214 build_subspace 16 8.4 0.026 0.036 6.834 6.844 multiply_cannon_metrocomm1 444 9.7 0.001 0.001 1.302 6.044 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.460 5.570 dbcsr_matrix_vector_mult_local 304 10.0 4.984 5.304 4.987 5.307 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.152 5.249 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.952 4.682 dbcsr_mm_accdrv_process_sort 1814 11.4 4.166 4.302 4.166 4.302 ls_scf_post 1 4.0 0.000 0.000 3.651 3.669 make_images_sizes 222 9.7 0.000 0.000 1.441 3.469 mp_alltoall_i44 222 10.7 1.441 3.469 1.441 3.469 ls_scf_store_result 1 5.0 0.000 0.000 3.401 3.412 calculate_norms 792 9.8 3.242 3.284 3.242 3.284 dbcsr_finalize 304 7.8 0.082 0.090 3.091 3.169 mp_allgather_i34 111 8.7 0.995 3.060 0.995 3.060 dbcsr_merge_all 275 8.9 0.892 0.914 2.875 2.948 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.898 2.898 dbcsr_complete_redistribute 5 7.6 1.439 1.467 2.766 2.881 dbcsr_data_release 12724 10.6 2.330 2.841 2.330 2.841 jit_kernel_multiply 1 12.0 0.990 2.762 0.990 2.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=130.833000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6832.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e0ef5d2e5812615faf6ac12d806b53b060af4958_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 586.452992E+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 13152 2094. MP_Sync 1064 MP_Alltoall 2588 640808658. 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 1.409 1.594 302.405 302.408 qs_mol_dyn_low 1 2.0 0.003 0.005 295.233 295.257 qs_forces 5 3.8 0.005 0.067 295.081 295.127 qs_energies 5 4.8 0.052 0.461 291.889 291.917 scf_env_do_scf 5 5.8 0.002 0.019 219.361 219.365 scf_env_do_scf_inner_loop 105 6.6 0.007 0.059 190.756 190.805 dbcsr_multiply_generic 1445 12.2 0.155 0.179 188.872 189.364 multiply_cannon 1445 13.2 0.274 0.297 169.347 171.703 multiply_cannon_loop 1445 14.2 2.394 2.517 167.454 168.830 qs_scf_new_mos 105 7.6 0.000 0.001 147.715 147.883 qs_scf_loop_do_ot 105 8.6 0.001 0.001 147.715 147.882 ot_scf_mini 105 9.6 0.003 0.004 137.961 138.105 velocity_verlet 4 3.0 0.007 0.051 107.788 107.788 mp_waitall_1 488190 16.1 52.019 78.467 52.019 78.467 multiply_cannon_multrec 69360 15.2 31.755 37.198 55.982 73.778 init_scf_run 5 5.8 0.001 0.020 69.280 69.285 scf_env_initial_rho_setup 5 6.8 0.003 0.022 69.279 69.285 calculate_first_density_matrix 1 7.0 0.006 0.058 64.075 64.116 multiply_cannon_metrocomm3 69360 15.2 0.195 0.209 36.624 61.166 ot_mini 105 10.6 0.001 0.003 59.312 59.435 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 52.212 52.235 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 51.369 51.448 qs_ot_get_p 112 10.4 0.001 0.002 48.804 49.084 acc_transpose_blocks 69360 15.2 0.361 0.382 28.894 42.688 acc_transpose_blocks_kernels 69360 16.2 0.869 0.921 26.993 40.806 dbcsr_mm_accdrv_process 154766 15.8 4.071 4.238 24.096 40.769 jit_kernel_transpose 5 15.0 26.124 39.958 26.124 39.958 qs_ot_get_derivative 55 11.6 0.001 0.002 37.735 37.872 qs_ot_p2m_diag 40 11.0 0.020 0.037 37.776 37.842 jit_kernel_multiply 13 15.2 19.150 35.179 19.150 35.179 multiply_cannon_sync_h2d 69360 15.2 30.622 34.747 30.622 34.747 cp_dbcsr_syevd 40 12.0 0.002 0.002 34.631 34.632 multiply_cannon_metrocomm1 69360 15.2 0.089 0.096 10.132 33.216 rebuild_ks_matrix 110 8.4 0.000 0.001 30.527 30.716 qs_ks_build_kohn_sham_matrix 110 9.4 0.061 0.402 30.526 30.716 cp_fm_syevd 40 13.0 0.000 0.001 29.356 29.499 init_scf_loop 7 6.6 0.023 0.211 28.569 28.576 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.153 28.350 cp_fm_redistribute_end 40 14.0 12.273 24.487 12.278 24.488 cp_fm_syevd_base 40 14.0 12.201 24.421 12.201 24.421 calculate_dm_sparse 110 9.5 0.000 0.001 23.272 23.417 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.935 23.229 apply_single 62 13.6 0.000 0.000 22.935 23.229 prepare_preconditioner 7 7.6 0.000 0.000 22.732 22.762 make_preconditioner 7 8.6 0.000 0.002 22.732 22.762 ot_new_cg_direction 55 11.6 0.001 0.004 20.930 20.930 qs_rho_update_rho_low 110 7.6 0.001 0.014 19.454 19.830 calculate_rho_elec 110 8.6 0.030 0.039 19.453 19.830 rs_pw_transfer 690 11.5 0.011 0.014 14.001 15.172 make_full_inverse_cholesky 7 9.6 0.000 0.001 15.056 15.111 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.732 14.868 density_rs2pw 110 9.6 0.006 0.008 13.402 14.636 cp_dbcsr_plus_fm_fm_t_native 10 8.8 0.000 0.000 13.981 14.035 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.569 13.662 mp_sum_l 4684 12.4 12.357 13.358 12.357 13.358 pw_transfer 1645 12.4 0.084 0.103 12.627 12.860 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 12.484 12.722 fft_wrap_pw1pw2_240 915 15.0 1.194 1.284 10.435 10.692 qs_vxc_create 110 10.4 0.008 0.079 10.563 10.613 cp_fm_cholesky_invert 7 10.6 10.550 10.560 10.550 10.560 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.199 10.265 check_diag 80 13.5 8.587 8.858 9.659 9.802 sum_up_and_integrate 60 10.3 0.028 0.031 9.572 9.589 integrate_v_rspace 60 11.3 0.002 0.002 9.544 9.562 fft3d_pb 915 16.0 2.401 2.701 8.471 8.744 xc_rho_set_and_dset_create 110 12.4 0.076 0.125 7.394 7.654 make_full_single_inverse 7 9.6 0.001 0.024 7.394 7.424 make_m2s 2890 13.2 0.076 0.084 6.701 7.275 make_images 2890 14.2 0.241 0.259 6.598 7.172 xc_vxc_pw_create 60 11.3 0.039 0.053 7.089 7.144 xc_pw_derive 510 13.4 0.005 0.006 6.283 6.365 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=302.408000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=556.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e0ef5d2e5812615faf6ac12d806b53b060af4958 Summary: empty Status: OK