=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: e790daffb71be1cee1ff37897efe550b44207090 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/01 job id: 47161676 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/02 job id: 47161677 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/03 job id: 47161678 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/04 job id: 47161679 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/05 job id: 47161680 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/06 job id: 47161681 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/07 job id: 47161683 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/08 job id: 47161684 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/09 job id: 47161685 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/10 job id: 47161686 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/11 job id: 47161687 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/12 job id: 47161688 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/13 job id: 47161689 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/14 job id: 47161690 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/15 job id: 47161691 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/16 job id: 47161692 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/17 job id: 47161693 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/18 job id: 47161694 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/19 job id: 47161696 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/20 job id: 47161697 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/21 job id: 47161698 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/22 job id: 47161699 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/23 job id: 47161700 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/24 job id: 47161701 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/25 job id: 47161702 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/26 job id: 47161703 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/27 job id: 47161704 --- 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/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.035 133.287 133.288 farming_run 1 2.0 132.731 132.732 133.257 133.260 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.469891E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.009 0.081 114.661 114.713 qs_energies 1 2.0 0.000 0.000 114.467 114.468 mp2_main 1 3.0 0.000 0.000 112.429 112.431 mp2_gpw_main 1 4.0 0.019 0.026 111.435 111.437 mp2_ri_gpw_compute_in 1 5.0 0.171 0.172 92.290 92.878 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.815 55.404 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.167 41.254 46.640 get_2c_integrals 1 6.0 0.008 0.011 36.671 37.303 integrate_v_rspace 273 8.0 0.434 0.450 24.700 29.789 pw_transfer 6555 10.6 0.375 0.388 27.293 28.011 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 25.943 26.667 grid_integrate_task_list 273 9.0 20.535 26.099 20.535 26.099 fft_wrap_pw1pw2_100 2178 12.4 1.165 1.231 23.472 24.189 compute_2c_integrals 1 7.0 0.002 0.002 19.213 19.214 rpa_ri_compute_en 1 5.0 0.019 0.021 19.037 19.171 compute_2c_integrals_loop_lm 1 8.0 0.003 0.003 18.838 18.939 mp2_eri_2c_integrate_gpw 1 9.0 2.398 2.437 18.836 18.938 cp_fm_cholesky_decompose 12 8.2 17.397 18.034 17.397 18.034 cholesky_decomp 1 7.0 0.000 0.000 16.307 16.946 fft3d_s 5443 13.4 16.115 16.655 16.136 16.677 ao_to_mo_and_store_B_mult_1 272 7.0 10.729 15.314 10.729 15.314 calculate_wavefunction 272 8.0 5.396 5.554 12.469 13.106 rpa_num_int 1 6.0 0.000 0.000 10.606 10.615 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.591 10.593 calc_mat_Q 8 8.0 0.000 0.000 9.458 9.538 contract_S_to_Q 8 9.0 0.000 0.000 8.880 8.959 calc_potential_gpw 544 9.5 0.005 0.006 8.193 8.601 parallel_gemm_fm 14 9.1 0.000 0.000 8.475 8.566 parallel_gemm_fm_cosma 14 10.1 8.475 8.566 8.475 8.566 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.174 8.412 potential_pw2rs 545 10.0 0.107 0.109 7.695 8.327 create_integ_mat 1 6.0 0.014 0.027 7.747 7.757 collocate_single_gaussian 272 10.0 0.039 0.041 7.455 7.679 array2fm 1 7.0 0.000 0.000 6.699 7.157 pw_scatter_s 2720 13.7 4.395 4.545 4.395 4.545 pw_gather_s 2722 13.2 3.868 4.229 3.868 4.229 array2fm_buffer_send 1 8.0 3.025 3.157 3.025 3.157 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.434140, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2812.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.038 394.798 394.799 farming_run 1 2.0 394.010 394.013 394.763 394.766 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.228550E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 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 0.031 208.641 208.642 qs_energies 1 2.0 0.000 0.000 208.432 208.438 scf_env_do_scf 1 3.0 0.000 0.000 106.300 106.300 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.400 105.408 rebuild_ks_matrix 4 6.0 0.000 0.000 105.399 105.406 qs_ks_build_kohn_sham_matrix 4 7.0 0.056 0.064 105.399 105.406 hfx_ks_matrix 4 8.0 0.001 0.001 105.014 105.018 integrate_four_center 4 9.0 0.144 0.457 105.013 105.017 mp2_main 1 3.0 0.000 0.000 101.771 101.778 mp2_gpw_main 1 4.0 0.032 0.044 100.928 100.938 integrate_four_center_main 4 10.0 0.101 0.560 96.637 99.482 integrate_four_center_bin 262 11.0 96.536 99.476 96.536 99.476 init_scf_loop 1 4.0 0.000 0.000 92.355 92.355 mp2_ri_gpw_compute_in 1 5.0 0.063 0.064 74.251 75.312 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.937 54.997 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.165 41.667 46.605 integrate_v_rspace 95 8.0 0.397 0.565 28.059 32.803 pw_transfer 2240 10.6 0.143 0.175 29.919 30.321 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.934 29.348 ao_to_mo_and_store_B_mult_1 91 7.0 10.588 28.486 10.588 28.486 grid_integrate_task_list 95 9.0 23.349 28.316 23.349 28.316 mp2_ri_gpw_compute_en 1 5.0 0.054 0.062 26.529 28.169 fft_wrap_pw1pw2_100 730 12.4 1.272 1.463 26.638 27.073 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.831 1.893 24.822 24.831 get_2c_integrals 1 6.0 0.000 0.000 20.227 20.250 compute_2c_integrals 1 7.0 0.002 0.003 19.209 19.211 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.854 19.090 mp2_eri_2c_integrate_gpw 1 9.0 1.741 1.874 18.853 19.088 fft3d_s 1823 13.4 18.447 18.734 18.460 18.747 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 13.944 13.944 calculate_wavefunction 91 8.0 2.010 2.048 9.731 9.986 potential_pw2rs 186 10.0 0.033 0.035 8.632 9.307 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.579 8.772 9.298 local_gemm 172 8.0 8.214 8.727 8.214 8.727 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.000 0.001 8.229 8.558 mp2_ri_gpw_compute_en_comm 22 7.0 0.503 0.518 7.825 8.235 collocate_single_gaussian 91 10.0 0.017 0.023 7.890 8.154 calc_potential_gpw 182 9.5 0.002 0.002 7.899 8.121 mp2_ri_gpw_compute_en_ener 172 7.0 6.350 6.418 6.350 6.418 mp_sendrecv_dm3 2068 8.0 5.851 6.236 5.851 6.236 mp_sync 37 10.5 3.021 6.068 3.021 6.068 pw_gather_s 912 13.2 4.897 5.431 4.897 5.431 pw_scatter_s 910 13.7 3.924 4.273 3.924 4.273 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.923092, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.837952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 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 0.012 0.029 53.033 53.035 qs_mol_dyn_low 1 2.0 0.003 0.004 52.740 52.817 qs_forces 11 3.9 0.002 0.003 52.289 52.290 qs_energies 11 4.9 0.001 0.001 50.826 50.839 scf_env_do_scf 11 5.9 0.000 0.001 44.877 44.878 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.856 42.856 qs_scf_new_mos 108 7.5 0.000 0.001 32.671 32.960 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.671 32.959 dbcsr_multiply_generic 2286 12.5 0.095 0.098 32.390 32.859 ot_scf_mini 108 9.5 0.002 0.002 30.983 31.152 multiply_cannon 2286 13.5 0.178 0.187 25.298 26.827 multiply_cannon_loop 2286 14.5 1.465 1.553 24.619 26.200 velocity_verlet 10 3.0 0.001 0.001 26.000 26.004 ot_mini 108 10.5 0.001 0.001 18.570 18.812 qs_ot_get_derivative 108 11.5 0.001 0.001 15.657 15.830 mp_waitall_1 245248 16.5 7.887 13.754 7.887 13.754 multiply_cannon_metrocomm3 54864 15.5 0.067 0.073 5.630 12.323 multiply_cannon_multrec 54864 15.5 4.283 6.686 7.519 10.971 qs_ot_get_p 119 10.4 0.001 0.001 7.805 8.086 rebuild_ks_matrix 119 8.3 0.000 0.000 7.900 8.022 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.899 8.022 multiply_cannon_sync_h2d 54864 15.5 5.957 7.181 5.957 7.181 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.961 7.074 mp_sum_l 7287 12.8 5.062 6.681 5.062 6.681 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.567 6.017 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.268 5.314 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.029 5.124 init_scf_run 11 5.9 0.000 0.001 4.654 4.655 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.654 4.654 sum_up_and_integrate 119 10.3 0.012 0.014 4.540 4.546 integrate_v_rspace 119 11.3 0.002 0.002 4.528 4.536 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.464 4.464 dbcsr_mm_accdrv_process 76910 16.1 1.137 1.817 3.157 4.434 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.260 4.359 calculate_rho_elec 119 8.7 0.012 0.017 4.260 4.359 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.248 4.248 cp_fm_redistribute_end 50 14.0 2.167 4.221 2.175 4.224 cp_fm_diag_elpa_base 50 14.0 2.043 4.111 2.048 4.119 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.902 3.093 apply_single 119 13.6 0.000 0.000 2.902 3.093 calculate_dm_sparse 119 9.5 0.000 0.001 2.901 3.052 rs_pw_transfer 974 11.9 0.011 0.012 2.762 2.872 ot_diis_step 108 11.5 0.006 0.006 2.664 2.665 jit_kernel_multiply 13 15.8 1.959 2.565 1.959 2.565 multiply_cannon_metrocomm1 54864 15.5 0.053 0.059 1.579 2.515 calculate_first_density_matrix 1 7.0 0.000 0.000 2.502 2.506 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.412 2.491 density_rs2pw 119 9.7 0.004 0.005 2.231 2.356 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.233 2.235 acc_transpose_blocks 54864 15.5 0.226 0.248 1.729 2.191 grid_integrate_task_list 119 12.3 2.024 2.133 2.024 2.133 wfi_extrapolate 11 7.9 0.001 0.001 2.081 2.081 init_scf_loop 11 6.9 0.000 0.000 2.005 2.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.926 1.969 potential_pw2rs 119 12.3 0.004 0.004 1.868 1.875 pw_transfer 1439 11.6 0.052 0.056 1.764 1.833 mp_sum_d 4135 12.0 1.208 1.812 1.208 1.812 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.689 1.757 make_m2s 4572 13.5 0.053 0.056 1.716 1.755 make_images 4572 14.5 0.133 0.139 1.633 1.673 mp_alltoall_d11v 2130 13.8 1.282 1.454 1.282 1.454 mp_waitany 12084 13.8 1.264 1.447 1.264 1.447 grid_collocate_task_list 119 9.7 1.359 1.437 1.359 1.437 fft3d_ps 1201 14.6 0.376 0.479 1.331 1.398 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.374 1.395 fft_wrap_pw1pw2_140 487 13.2 0.185 0.201 1.307 1.377 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.670 1.082 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.035000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.727273, yerr=0.445362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.915520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1861874. 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 0.018 0.051 40.325 40.326 qs_mol_dyn_low 1 2.0 0.003 0.009 40.070 40.078 qs_forces 11 3.9 0.003 0.007 38.760 38.761 qs_energies 11 4.9 0.002 0.008 37.060 37.063 scf_env_do_scf 11 5.9 0.001 0.003 31.906 31.907 scf_env_do_scf_inner_loop 108 6.5 0.003 0.011 29.417 29.418 dbcsr_multiply_generic 2286 12.5 0.102 0.105 21.311 21.672 qs_scf_new_mos 108 7.5 0.001 0.001 20.126 20.374 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.125 20.374 velocity_verlet 10 3.0 0.002 0.007 19.536 19.565 ot_scf_mini 108 9.5 0.003 0.004 19.230 19.402 multiply_cannon 2286 13.5 0.206 0.214 16.292 17.860 multiply_cannon_loop 2286 14.5 0.904 0.972 15.184 16.503 ot_mini 108 10.5 0.001 0.001 11.782 12.015 mp_waitall_1 200699 16.5 5.624 10.705 5.624 10.705 qs_ot_get_derivative 108 11.5 0.001 0.001 9.316 9.493 multiply_cannon_metrocomm3 27432 15.5 0.066 0.068 4.088 9.392 multiply_cannon_multrec 27432 15.5 2.005 4.339 5.896 8.707 rebuild_ks_matrix 119 8.3 0.000 0.000 7.250 7.389 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 7.250 7.389 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.403 6.528 dbcsr_mm_accdrv_process 47894 16.0 2.959 4.944 3.822 5.646 qs_ot_get_p 119 10.4 0.001 0.002 4.538 4.778 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.590 4.442 sum_up_and_integrate 119 10.3 0.025 0.028 4.237 4.243 integrate_v_rspace 119 11.3 0.002 0.003 4.212 4.219 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.049 4.122 apply_single 119 13.6 0.000 0.000 3.048 4.122 mp_sum_l 7287 12.8 2.053 4.006 2.053 4.006 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.877 3.908 calculate_rho_elec 119 8.7 0.021 0.025 3.877 3.907 init_scf_run 11 5.9 0.000 0.001 3.887 3.888 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.887 3.887 rs_pw_transfer 974 11.9 0.010 0.011 2.688 3.129 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.072 3.090 multiply_cannon_sync_h2d 27432 15.5 2.197 2.839 2.197 2.839 make_m2s 4572 13.5 0.052 0.054 2.523 2.738 make_images 4572 14.5 0.201 0.238 2.432 2.645 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.629 2.629 density_rs2pw 119 9.7 0.004 0.004 2.155 2.598 init_scf_loop 11 6.9 0.002 0.015 2.467 2.467 ot_diis_step 108 11.5 0.011 0.011 2.417 2.418 calculate_first_density_matrix 1 7.0 0.001 0.004 2.352 2.354 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.296 2.296 cp_fm_redistribute_end 50 14.0 1.164 2.270 1.168 2.273 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 2.169 2.259 cp_fm_diag_elpa_base 50 14.0 1.072 2.175 1.101 2.216 calculate_dm_sparse 119 9.5 0.000 0.001 2.033 2.109 pw_transfer 1439 11.6 0.066 0.070 1.937 1.970 potential_pw2rs 119 12.3 0.006 0.006 1.958 1.968 grid_integrate_task_list 119 12.3 1.835 1.925 1.835 1.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 1.906 1.908 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.845 1.880 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.683 1.721 jit_kernel_multiply 8 16.2 0.810 1.590 0.810 1.590 make_images_data 4572 15.5 0.044 0.050 1.156 1.581 prepare_preconditioner 11 7.9 0.000 0.000 1.541 1.570 make_preconditioner 11 8.9 0.000 0.001 1.541 1.570 acc_transpose_blocks 27432 15.5 0.111 0.116 1.209 1.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.439 1.501 fft_wrap_pw1pw2_140 487 13.2 0.201 0.213 1.445 1.481 wfi_extrapolate 11 7.9 0.001 0.001 1.473 1.473 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.004 1.472 fft3d_ps 1201 14.6 0.524 0.581 1.397 1.425 grid_collocate_task_list 119 9.7 1.269 1.423 1.269 1.423 mp_sum_dm 438 4.9 1.276 1.334 1.276 1.334 mp_alltoall_d11v 2130 13.8 1.186 1.324 1.186 1.324 mp_allgather_i34 2286 14.5 0.551 1.300 0.551 1.300 update_particle_set 20 4.0 0.000 0.000 1.248 1.278 md_write_output 11 3.9 0.030 1.262 0.031 1.277 md_output 10 3.0 0.000 0.000 0.029 1.256 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.243 1.251 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.172 1.221 mp_sum_d 4135 12.0 0.620 1.078 0.620 1.078 mp_waitany 5720 13.7 0.562 1.035 0.562 1.035 rs_pw_transfer_RS2PW_140 130 11.5 0.137 0.145 0.585 1.020 qs_energies_init_hamiltonians 11 5.9 0.000 0.003 0.971 0.972 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.909 0.923 acc_transpose_blocks_kernels 27432 16.5 0.182 0.274 0.658 0.868 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.608 0.774 0.816 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=40.326000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.636364, yerr=1.871933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 521.457664E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1434464. 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 0.040 0.083 38.185 38.186 qs_mol_dyn_low 1 2.0 0.003 0.003 35.262 37.840 qs_forces 11 3.9 0.007 0.031 35.118 35.119 qs_energies 11 4.9 0.001 0.005 33.497 33.500 scf_env_do_scf 11 5.9 0.001 0.001 26.627 26.627 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.950 23.951 dbcsr_multiply_generic 2286 12.5 0.097 0.099 18.601 18.736 multiply_cannon 2286 13.5 0.196 0.205 15.109 15.874 velocity_verlet 10 3.0 0.001 0.001 15.765 15.767 qs_scf_new_mos 108 7.5 0.001 0.001 15.457 15.477 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.456 15.476 multiply_cannon_loop 2286 14.5 0.635 0.662 14.296 15.199 ot_scf_mini 108 9.5 0.002 0.003 14.707 14.724 ot_mini 108 10.5 0.001 0.001 8.957 8.975 multiply_cannon_multrec 18288 15.5 1.940 2.902 7.670 8.852 dbcsr_mm_accdrv_process 38222 16.0 5.124 7.293 5.645 7.550 qs_ot_get_derivative 108 11.5 0.001 0.001 7.421 7.439 rebuild_ks_matrix 119 8.3 0.000 0.000 6.530 6.557 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.530 6.556 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.767 5.791 init_scf_run 11 5.9 0.000 0.001 5.634 5.635 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.634 5.635 mp_waitall_1 158411 16.6 3.861 5.276 3.861 5.276 calculate_first_density_matrix 1 7.0 0.000 0.002 4.369 4.372 multiply_cannon_metrocomm3 18288 15.5 0.043 0.045 2.204 4.299 sum_up_and_integrate 119 10.3 0.030 0.031 4.094 4.100 integrate_v_rspace 119 11.3 0.003 0.003 4.064 4.073 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.668 3.669 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.647 3.653 calculate_rho_elec 119 8.7 0.031 0.031 3.647 3.652 qs_ot_get_p 119 10.4 0.001 0.001 3.516 3.545 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.492 3.503 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.749 3.376 rs_pw_transfer 974 11.9 0.009 0.010 2.457 2.710 md_output 10 3.0 0.000 0.000 0.087 2.699 mp_max_l 33 2.8 2.579 2.662 2.579 2.662 init_scf_loop 11 6.9 0.000 0.002 2.656 2.657 qs_ot_p2m_diag 50 11.0 0.012 0.012 2.396 2.402 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.065 2.396 apply_single 119 13.6 0.000 0.000 2.065 2.395 density_rs2pw 119 9.7 0.004 0.004 2.124 2.382 write_restart 10 4.0 0.072 2.301 0.073 2.302 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.743 2.138 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.116 2.117 make_m2s 4572 13.5 0.044 0.045 1.877 2.038 pw_transfer 1439 11.6 0.065 0.069 2.015 2.031 jit_kernel_multiply 6 16.5 0.469 2.025 0.469 2.025 make_images 4572 14.5 0.191 0.204 1.790 1.950 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.921 1.939 grid_integrate_task_list 119 12.3 1.803 1.884 1.803 1.884 calculate_dm_sparse 119 9.5 0.000 0.000 1.841 1.853 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.842 1.851 potential_pw2rs 119 12.3 0.007 0.008 1.839 1.849 cp_fm_diag_elpa_base 50 14.0 1.817 1.831 1.841 1.849 prepare_preconditioner 11 7.9 0.000 0.000 1.838 1.840 make_preconditioner 11 8.9 0.000 0.001 1.838 1.840 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.687 1.778 qs_ot_get_derivative_diag 49 12.0 0.001 0.002 1.649 1.661 mp_sum_l 7287 12.8 1.256 1.612 1.256 1.612 multiply_cannon_sync_h2d 18288 15.5 1.380 1.521 1.380 1.521 fft_wrap_pw1pw2_140 487 13.2 0.252 0.258 1.498 1.516 ot_diis_step 108 11.5 0.011 0.012 1.509 1.509 fft3d_ps 1201 14.6 0.532 0.549 1.410 1.427 grid_collocate_task_list 119 9.7 1.245 1.397 1.245 1.397 acc_transpose_blocks 18288 15.5 0.076 0.078 1.251 1.275 wfi_extrapolate 11 7.9 0.001 0.001 1.208 1.208 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.189 1.196 make_images_data 4572 15.5 0.044 0.048 0.849 1.020 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.974 0.975 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.931 0.944 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.730 0.916 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.836 0.838 mp_alltoall_d11v 2130 13.8 0.725 0.838 0.725 0.838 mp_waitany 9880 13.7 0.550 0.834 0.550 0.834 acc_transpose_blocks_kernels 18288 16.5 0.211 0.221 0.797 0.810 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.543 0.810 cp_fm_cholesky_invert 11 10.9 0.792 0.795 0.792 0.795 mp_alltoall_z22v 1201 16.6 0.721 0.790 0.721 0.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=38.186000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.818182, yerr=2.289032 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 554.930176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.036 37.777 37.778 qs_mol_dyn_low 1 2.0 0.003 0.004 35.839 37.603 qs_forces 11 3.9 0.018 0.056 35.472 35.473 qs_energies 11 4.9 0.001 0.001 33.709 33.727 scf_env_do_scf 11 5.9 0.000 0.001 28.500 28.502 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.005 25.005 velocity_verlet 10 3.0 0.001 0.001 18.392 18.404 dbcsr_multiply_generic 2286 12.5 0.102 0.105 18.235 18.331 qs_scf_new_mos 108 7.5 0.001 0.001 16.399 16.454 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.398 16.454 ot_scf_mini 108 9.5 0.002 0.003 15.449 15.500 multiply_cannon 2286 13.5 0.221 0.234 14.604 14.930 multiply_cannon_loop 2286 14.5 0.948 0.979 13.651 14.032 ot_mini 108 10.5 0.001 0.001 9.290 9.361 multiply_cannon_multrec 27432 15.5 2.360 3.054 8.750 9.097 qs_ot_get_derivative 108 11.5 0.001 0.001 7.455 7.509 dbcsr_mm_accdrv_process 47916 15.9 5.370 7.103 6.301 7.497 rebuild_ks_matrix 119 8.3 0.000 0.000 6.687 6.737 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.686 6.736 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.922 5.966 sum_up_and_integrate 119 10.3 0.035 0.037 3.919 3.927 integrate_v_rspace 119 11.3 0.003 0.003 3.884 3.893 init_scf_run 11 5.9 0.000 0.001 3.831 3.832 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.831 3.831 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.619 3.655 calculate_rho_elec 119 8.7 0.040 0.046 3.618 3.655 qs_ot_get_p 119 10.4 0.001 0.001 3.523 3.600 init_scf_loop 11 6.9 0.000 0.000 3.476 3.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.809 3.204 prepare_preconditioner 11 7.9 0.000 0.000 2.621 2.631 make_preconditioner 11 8.9 0.000 0.000 2.621 2.631 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.231 2.561 mp_waitall_1 137007 16.6 1.866 2.525 1.866 2.525 calculate_first_density_matrix 1 7.0 0.000 0.000 2.421 2.423 make_m2s 4572 13.5 0.054 0.055 2.237 2.361 rs_pw_transfer 974 11.9 0.009 0.009 2.134 2.331 make_images 4572 14.5 0.271 0.332 2.130 2.252 qs_ot_p2m_diag 50 11.0 0.015 0.022 2.203 2.211 density_rs2pw 119 9.7 0.004 0.004 2.012 2.197 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.059 2.190 apply_single 119 13.6 0.000 0.000 2.058 2.189 md_output 10 3.0 0.000 0.000 0.090 2.108 pw_transfer 1439 11.6 0.065 0.070 2.068 2.102 calculate_dm_sparse 119 9.5 0.000 0.000 2.036 2.092 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.975 2.013 md_write_output 11 3.9 0.081 1.920 0.084 1.976 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.939 1.968 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.927 1.928 grid_integrate_task_list 119 12.3 1.829 1.916 1.829 1.916 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.886 1.886 mp_max_l 33 2.8 1.765 1.842 1.765 1.842 ot_diis_step 108 11.5 0.012 0.012 1.791 1.791 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.747 1.760 potential_pw2rs 119 12.3 0.009 0.009 1.655 1.658 fft_wrap_pw1pw2_140 487 13.2 0.287 0.302 1.609 1.650 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.628 1.637 cp_fm_diag_elpa_base 50 14.0 1.593 1.610 1.626 1.635 jit_kernel_multiply 8 15.7 0.869 1.623 0.869 1.623 acc_transpose_blocks 27432 15.5 0.112 0.115 1.477 1.493 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.759 1.446 fft3d_ps 1201 14.6 0.560 0.610 1.414 1.442 grid_collocate_task_list 119 9.7 1.248 1.368 1.248 1.368 wfi_extrapolate 11 7.9 0.001 0.001 1.353 1.353 mp_sum_l 7287 12.8 0.961 1.247 0.961 1.247 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.205 1.216 cp_fm_upper_to_full 72 14.2 0.811 1.158 0.811 1.158 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.127 1.144 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.118 1.135 multiply_cannon_sync_h2d 27432 15.5 0.986 1.061 0.986 1.061 dbcsr_complete_redistribute 329 12.2 0.119 0.145 0.774 1.044 make_images_data 4572 15.5 0.044 0.047 0.877 1.012 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.763 0.940 mp_alltoall_d11v 2130 13.8 0.752 0.904 0.752 0.904 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.855 0.860 acc_transpose_blocks_kernels 27432 16.5 0.272 0.280 0.827 0.837 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.566 0.835 cp_fm_cholesky_invert 11 10.9 0.805 0.809 0.805 0.809 mp_alltoall_z22v 1201 16.6 0.738 0.761 0.738 0.761 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.778000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.181818, yerr=2.208174 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 603.324416E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.032 33.898 33.899 qs_mol_dyn_low 1 2.0 0.003 0.003 32.190 33.710 qs_forces 11 3.9 0.002 0.002 31.666 31.667 qs_energies 11 4.9 0.001 0.001 29.861 29.865 scf_env_do_scf 11 5.9 0.000 0.001 24.804 24.804 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 21.822 21.823 velocity_verlet 10 3.0 0.001 0.001 16.863 16.888 dbcsr_multiply_generic 2286 12.5 0.096 0.101 13.449 13.619 qs_scf_new_mos 108 7.5 0.001 0.001 12.637 12.670 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.637 12.670 ot_scf_mini 108 9.5 0.002 0.002 11.879 11.911 multiply_cannon 2286 13.5 0.233 0.241 10.116 10.608 multiply_cannon_loop 2286 14.5 0.333 0.343 9.119 9.407 rebuild_ks_matrix 119 8.3 0.000 0.000 6.721 6.758 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.721 6.758 ot_mini 108 10.5 0.001 0.001 6.623 6.662 multiply_cannon_multrec 9144 15.5 1.603 1.966 5.766 6.153 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.993 6.027 qs_ot_get_derivative 108 11.5 0.001 0.001 5.246 5.276 sum_up_and_integrate 119 10.3 0.037 0.041 4.264 4.269 integrate_v_rspace 119 11.3 0.003 0.003 4.226 4.232 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.177 4.194 calculate_rho_elec 119 8.7 0.060 0.061 4.177 4.193 dbcsr_mm_accdrv_process 12550 15.8 3.313 4.110 4.064 4.162 init_scf_run 11 5.9 0.000 0.001 3.527 3.527 scf_env_initial_rho_setup 11 6.9 0.009 0.009 3.527 3.527 qs_ot_get_p 119 10.4 0.001 0.001 3.283 3.335 mp_waitall_1 115863 16.7 2.602 3.123 2.602 3.123 init_scf_loop 11 6.9 0.000 0.000 2.953 2.954 rs_pw_transfer 974 11.9 0.008 0.008 2.494 2.633 density_rs2pw 119 9.7 0.004 0.004 2.423 2.575 make_m2s 4572 13.5 0.034 0.035 2.345 2.537 make_images 4572 14.5 0.268 0.299 2.254 2.445 pw_transfer 1439 11.6 0.066 0.069 2.407 2.418 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.313 2.325 prepare_preconditioner 11 7.9 0.000 0.000 2.136 2.142 make_preconditioner 11 8.9 0.000 0.000 2.136 2.142 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.123 2.126 calculate_first_density_matrix 1 7.0 0.000 0.000 2.121 2.123 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.996 2.042 md_output 10 3.0 0.000 0.000 0.129 2.020 grid_integrate_task_list 119 12.3 1.863 1.974 1.863 1.974 potential_pw2rs 119 12.3 0.010 0.011 1.918 1.921 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.893 1.895 calculate_dm_sparse 119 9.5 0.000 0.000 1.803 1.830 fft_wrap_pw1pw2_140 487 13.2 0.365 0.373 1.803 1.816 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.760 1.782 fft3d_ps 1201 14.6 0.566 0.575 1.646 1.656 mp_max_l 33 2.8 1.521 1.622 1.521 1.622 write_restart 10 4.0 0.101 1.618 0.102 1.619 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.608 1.610 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.560 1.568 cp_fm_diag_elpa_base 50 14.0 1.528 1.550 1.559 1.566 make_images_data 4572 15.5 0.038 0.041 1.122 1.447 grid_collocate_task_list 119 9.7 1.306 1.433 1.306 1.433 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.415 1.430 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.403 1.414 hybrid_alltoall_any 4725 16.4 0.063 0.176 1.034 1.406 ot_diis_step 108 11.5 0.013 0.013 1.347 1.347 wfi_extrapolate 11 7.9 0.001 0.001 1.325 1.325 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.217 1.255 apply_single 119 13.6 0.000 0.000 1.216 1.255 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.240 1.241 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.152 1.160 cp_fm_cholesky_invert 11 10.9 1.100 1.104 1.100 1.104 mp_alltoall_d11v 2130 13.8 0.965 1.058 0.965 1.058 jit_kernel_multiply 6 15.5 0.713 1.045 0.713 1.045 mp_alltoall_z22v 1201 16.6 0.953 0.980 0.953 0.980 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.721 0.979 acc_transpose_blocks 9144 15.5 0.039 0.040 0.912 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.923 make_images_sizes 4572 15.5 0.005 0.005 0.559 0.877 mp_alltoall_i44 4572 16.5 0.554 0.873 0.554 0.873 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.853 0.857 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.832 0.841 mp_allgather_i34 2286 14.5 0.324 0.832 0.324 0.832 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.392 0.780 multiply_cannon_sync_h2d 9144 15.5 0.695 0.778 0.695 0.778 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.695 0.749 dbcsr_complete_redistribute 329 12.2 0.162 0.175 0.662 0.704 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.662 0.689 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=33.899000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=572.636364, yerr=4.416348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 764.407808E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.033 44.361 44.361 qs_mol_dyn_low 1 2.0 0.003 0.004 44.157 44.164 qs_forces 11 3.9 0.002 0.002 41.584 41.585 qs_energies 11 4.9 0.001 0.001 39.572 39.575 scf_env_do_scf 11 5.9 0.001 0.001 33.719 33.719 velocity_verlet 10 3.0 0.001 0.001 25.917 26.237 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.704 25.705 dbcsr_multiply_generic 2286 12.5 0.101 0.104 17.471 17.617 qs_scf_new_mos 108 7.5 0.001 0.001 15.873 15.971 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.872 15.970 ot_scf_mini 108 9.5 0.002 0.002 14.779 14.878 multiply_cannon 2286 13.5 0.303 0.306 13.553 14.471 multiply_cannon_loop 2286 14.5 0.345 0.350 12.310 13.227 ot_mini 108 10.5 0.001 0.001 8.781 8.900 multiply_cannon_multrec 9144 15.5 3.425 4.841 8.555 8.634 init_scf_loop 11 6.9 0.000 0.000 7.987 7.988 rebuild_ks_matrix 119 8.3 0.000 0.000 7.323 7.469 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.323 7.469 prepare_preconditioner 11 7.9 0.000 0.000 7.023 7.036 make_preconditioner 11 8.9 0.000 0.000 7.023 7.036 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.529 6.913 qs_ot_get_derivative 108 11.5 0.001 0.001 6.746 6.847 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.599 6.732 dbcsr_mm_accdrv_process 12550 15.8 4.244 5.623 5.010 6.358 cp_fm_upper_to_full 72 14.2 3.246 4.684 3.246 4.684 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.337 4.359 calculate_rho_elec 119 8.7 0.118 0.121 4.336 4.358 sum_up_and_integrate 119 10.3 0.064 0.066 4.041 4.048 integrate_v_rspace 119 11.3 0.003 0.003 3.976 3.982 init_scf_run 11 5.9 0.000 0.001 3.810 3.810 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.810 3.810 qs_ot_get_p 119 10.4 0.001 0.001 3.252 3.393 mp_waitall_1 94719 16.7 2.370 3.301 2.370 3.301 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.443 2.890 dbcsr_complete_redistribute 329 12.2 0.286 0.289 2.024 2.870 pw_transfer 1439 11.6 0.069 0.069 2.706 2.711 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.607 2.612 mp_sum_dm 438 4.9 2.260 2.593 2.260 2.593 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.717 2.573 md_write_output 11 3.9 0.033 0.258 0.322 2.561 md_output 10 3.0 0.000 0.000 0.322 2.559 update_particle_set 20 4.0 0.000 0.000 2.238 2.558 make_m2s 4572 13.5 0.037 0.038 2.300 2.458 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.221 2.455 apply_single 119 13.6 0.000 0.000 2.221 2.455 mp_alltoall_i22 627 13.8 1.530 2.361 1.530 2.361 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.489 2.338 make_images 4572 14.5 0.351 0.380 2.178 2.336 calculate_first_density_matrix 1 7.0 0.000 0.000 2.304 2.329 write_trajectory 44 4.9 0.288 2.297 0.288 2.303 calculate_dm_sparse 119 9.5 0.000 0.000 2.233 2.252 density_rs2pw 119 9.7 0.004 0.004 2.210 2.234 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.291 2.192 fft_wrap_pw1pw2_140 487 13.2 0.615 0.617 2.181 2.187 grid_integrate_task_list 119 12.3 2.068 2.091 2.068 2.091 ot_diis_step 108 11.5 0.014 0.015 2.010 2.010 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.956 1.958 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.917 1.918 mp_sum_l 7287 12.8 1.068 1.789 1.068 1.789 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.769 1.770 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.684 1.735 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.684 1.685 fft3d_ps 1201 14.6 0.596 0.607 1.626 1.630 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.600 1.622 grid_collocate_task_list 119 9.7 1.515 1.533 1.515 1.533 cp_fm_cholesky_invert 11 10.9 1.515 1.519 1.515 1.519 rs_pw_transfer 974 11.9 0.009 0.009 1.473 1.514 wfi_extrapolate 11 7.9 0.001 0.001 1.413 1.413 potential_pw2rs 119 12.3 0.014 0.014 1.400 1.402 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.161 1.400 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.396 1.397 cp_fm_diag_elpa_base 50 14.0 1.248 1.302 1.394 1.395 make_images_data 4572 15.5 0.042 0.044 1.108 1.312 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.288 1.308 mp_alltoall_d11v 2130 13.8 1.135 1.178 1.135 1.178 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.146 1.167 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.104 1.119 multiply_cannon_sync_h2d 9144 15.5 1.038 1.042 1.038 1.042 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.990 1.032 jit_kernel_multiply 6 15.5 0.739 1.012 0.739 1.012 acc_transpose_blocks 9144 15.5 0.039 0.039 0.916 0.963 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.935 0.949 qs_create_task_list 11 7.9 0.000 0.001 0.933 0.945 generate_qs_task_list 11 8.9 0.367 0.386 0.932 0.945 mp_alltoall_z22v 1201 16.6 0.895 0.910 0.895 0.910 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.361000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=717.818182, yerr=13.776252 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.612544E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 3430608. 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 0.016 0.032 90.443 90.444 qs_mol_dyn_low 1 2.0 0.003 0.006 88.287 90.045 qs_forces 11 3.9 0.003 0.003 88.141 88.142 qs_energies 11 4.9 0.001 0.001 85.092 85.106 scf_env_do_scf 11 5.9 0.000 0.001 75.055 75.058 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 68.501 68.503 dbcsr_multiply_generic 2055 12.4 0.110 0.113 52.281 52.527 qs_scf_new_mos 99 7.5 0.000 0.001 49.420 49.541 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.420 49.540 ot_scf_mini 99 9.5 0.002 0.002 47.015 47.112 velocity_verlet 10 3.0 0.001 0.001 45.918 45.919 multiply_cannon 2055 13.4 0.178 0.181 42.896 43.982 multiply_cannon_loop 2055 14.4 1.516 1.552 41.636 42.773 ot_mini 99 10.5 0.001 0.001 27.332 27.453 qs_ot_get_derivative 99 11.5 0.001 0.001 20.490 20.612 multiply_cannon_multrec 49320 15.4 12.363 12.932 17.479 18.264 rebuild_ks_matrix 110 8.3 0.000 0.001 15.535 15.721 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.534 15.720 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.582 13.741 mp_waitall_1 220248 16.4 11.707 12.418 11.707 12.418 qs_ot_get_p 110 10.4 0.001 0.001 11.023 11.150 multiply_cannon_sync_h2d 49320 15.4 10.221 10.941 10.221 10.941 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.767 8.325 sum_up_and_integrate 110 10.3 0.036 0.043 7.867 7.880 init_scf_run 11 5.9 0.000 0.001 7.861 7.862 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.861 7.862 integrate_v_rspace 110 11.3 0.003 0.004 7.831 7.853 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.286 7.834 apply_single 110 13.6 0.000 0.001 7.286 7.834 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.685 7.828 calculate_rho_elec 110 8.6 0.021 0.026 7.684 7.827 multiply_cannon_metrocomm3 49320 15.4 0.077 0.081 6.506 7.744 qs_ot_p2m_diag 48 11.0 0.012 0.019 7.616 7.640 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.634 6.635 ot_diis_step 99 11.5 0.006 0.006 6.632 6.632 init_scf_loop 11 6.9 0.000 0.000 6.511 6.513 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.914 5.935 cp_fm_diag_elpa_base 48 14.0 5.895 5.919 5.911 5.933 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.838 5.910 rs_pw_transfer 902 11.9 0.011 0.013 4.703 5.392 dbcsr_mm_accdrv_process 87628 16.1 1.945 2.041 4.993 5.291 density_rs2pw 110 9.6 0.004 0.005 4.338 5.040 mp_sum_l 6594 12.7 4.246 5.018 4.246 5.018 make_m2s 4110 13.4 0.061 0.064 4.678 4.817 make_images 4110 14.4 0.178 0.193 4.580 4.721 wfi_extrapolate 11 7.9 0.001 0.001 4.422 4.423 prepare_preconditioner 11 7.9 0.000 0.000 4.244 4.263 make_preconditioner 11 8.9 0.000 0.000 4.244 4.263 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.018 4.067 multiply_cannon_metrocomm1 49320 15.4 0.062 0.064 2.628 4.036 pw_transfer 1331 11.6 0.055 0.065 3.864 3.989 calculate_dm_sparse 110 9.5 0.001 0.001 3.862 3.959 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.775 3.899 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 3.803 3.807 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.422 3.495 grid_integrate_task_list 110 12.3 3.256 3.440 3.256 3.440 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.352 3.405 potential_pw2rs 110 12.3 0.006 0.006 3.295 3.357 calculate_first_density_matrix 1 7.0 0.000 0.001 3.321 3.326 fft_wrap_pw1pw2_140 451 13.1 0.451 0.501 3.051 3.172 fft3d_ps 1111 14.6 0.795 0.891 2.971 3.091 jit_kernel_multiply 13 15.9 2.770 2.804 2.770 2.804 mp_waitany 14300 13.8 1.918 2.745 1.918 2.745 mp_alltoall_d11v 2046 13.8 2.362 2.655 2.362 2.655 grid_collocate_task_list 110 9.6 2.159 2.410 2.159 2.410 cp_fm_cholesky_invert 11 10.9 2.384 2.389 2.384 2.389 make_images_data 4110 15.4 0.042 0.046 2.047 2.291 acc_transpose_blocks 49320 15.4 0.210 0.219 2.154 2.211 hybrid_alltoall_any 4261 16.3 0.082 0.479 1.734 2.156 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.085 2.114 mp_alltoall_z22v 1111 16.6 1.797 1.995 1.797 1.995 parallel_gemm_fm 81 9.0 0.000 0.000 1.959 1.964 parallel_gemm_fm_cosma 81 10.0 1.959 1.964 1.959 1.964 mp_sum_d 3889 11.9 1.397 1.943 1.397 1.943 md_output 10 3.0 0.000 0.000 0.023 1.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=90.444000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.181818, yerr=2.328409 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 588.075008E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2359657. 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 0.020 0.061 76.960 76.960 qs_mol_dyn_low 1 2.0 0.003 0.004 76.624 76.634 qs_forces 11 3.9 0.003 0.009 74.444 74.445 qs_energies 11 4.9 0.001 0.002 70.994 70.998 scf_env_do_scf 11 5.9 0.000 0.001 61.658 61.661 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 52.983 52.984 velocity_verlet 10 3.0 0.001 0.001 40.829 40.874 dbcsr_multiply_generic 2055 12.4 0.118 0.122 38.819 38.989 qs_scf_new_mos 99 7.5 0.001 0.001 35.237 35.385 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.237 35.384 ot_scf_mini 99 9.5 0.003 0.003 33.538 33.671 multiply_cannon 2055 13.4 0.221 0.242 31.356 32.538 multiply_cannon_loop 2055 14.4 0.923 0.947 29.960 30.789 ot_mini 99 10.5 0.001 0.001 18.951 19.102 multiply_cannon_multrec 24660 15.4 7.639 9.540 13.863 15.998 rebuild_ks_matrix 110 8.3 0.000 0.001 14.716 14.845 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.716 14.844 qs_ot_get_derivative 99 11.5 0.001 0.001 13.102 13.238 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.964 13.081 mp_waitall_1 176588 16.5 8.541 10.975 8.541 10.975 init_scf_loop 11 6.9 0.000 0.000 8.624 8.625 multiply_cannon_sync_h2d 24660 15.4 7.089 8.269 7.089 8.269 qs_ot_get_p 110 10.4 0.001 0.001 7.799 7.946 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.191 7.845 sum_up_and_integrate 110 10.3 0.053 0.059 7.418 7.429 integrate_v_rspace 110 11.3 0.003 0.004 7.365 7.376 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.530 7.142 apply_single 110 13.6 0.000 0.001 6.529 7.142 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.942 6.951 calculate_rho_elec 110 8.6 0.040 0.048 6.941 6.951 init_scf_run 11 5.9 0.000 0.001 6.902 6.903 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.902 6.902 prepare_preconditioner 11 7.9 0.000 0.000 6.473 6.488 make_preconditioner 11 8.9 0.000 0.000 6.473 6.488 dbcsr_mm_accdrv_process 52282 16.1 4.387 5.386 6.062 6.362 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.017 6.210 ot_diis_step 99 11.5 0.010 0.011 5.798 5.799 qs_ot_p2m_diag 48 11.0 0.028 0.044 5.587 5.608 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.896 5.594 make_m2s 4110 13.4 0.057 0.060 4.952 5.350 make_images 4110 14.4 0.401 0.447 4.839 5.236 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.114 5.114 density_rs2pw 110 9.6 0.004 0.005 3.992 4.598 rs_pw_transfer 902 11.9 0.012 0.014 3.920 4.561 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.388 4.406 cp_fm_diag_elpa_base 48 14.0 4.334 4.353 4.385 4.402 pw_transfer 1331 11.6 0.067 0.075 4.161 4.336 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.053 4.232 wfi_extrapolate 11 7.9 0.001 0.001 3.772 3.772 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.468 3.545 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.427 3.430 fft_wrap_pw1pw2_140 451 13.1 0.516 0.530 3.198 3.374 grid_integrate_task_list 110 12.3 3.136 3.332 3.136 3.332 fft3d_ps 1111 14.6 1.111 1.339 3.032 3.211 potential_pw2rs 110 12.3 0.008 0.009 3.131 3.169 make_images_data 4110 15.4 0.046 0.050 2.680 3.164 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.056 3.104 cp_fm_cholesky_invert 11 10.9 3.051 3.059 3.051 3.059 calculate_dm_sparse 110 9.5 0.001 0.001 3.011 3.038 calculate_first_density_matrix 1 7.0 0.000 0.001 3.017 3.020 hybrid_alltoall_any 4261 16.3 0.102 0.446 2.244 2.836 mp_sum_l 6594 12.7 1.899 2.617 1.899 2.617 grid_collocate_task_list 110 9.6 2.098 2.540 2.098 2.540 mp_alltoall_d11v 2046 13.8 1.933 2.307 1.933 2.307 mp_sum_dm 438 4.9 2.175 2.236 2.175 2.236 md_write_output 11 3.9 0.045 2.125 0.047 2.162 update_particle_set 20 4.0 0.000 0.000 2.114 2.159 md_output 10 3.0 0.000 0.000 0.048 2.147 jit_kernel_multiply 11 16.2 1.315 2.115 1.315 2.115 mp_waitany 10164 13.8 1.359 2.058 1.359 2.058 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.027 2.044 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.912 1.926 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.874 1.875 cp_fm_cholesky_decompose 22 10.9 1.814 1.821 1.814 1.821 mp_allgather_i34 2055 14.4 0.723 1.817 0.723 1.817 mp_alltoall_z22v 1111 16.6 1.633 1.748 1.633 1.748 acc_transpose_blocks 24660 15.4 0.113 0.117 1.568 1.689 rs_pw_transfer_RS2PW_140 121 11.5 0.204 0.215 1.012 1.652 multiply_cannon_metrocomm4 22605 15.4 0.077 0.080 0.766 1.587 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.559 1.571 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=76.960000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.181818, yerr=7.030042 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 660.525056E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.028 62.422 62.423 qs_mol_dyn_low 1 2.0 0.003 0.004 62.174 62.184 qs_forces 11 3.9 0.003 0.003 60.413 60.414 qs_energies 11 4.9 0.001 0.001 57.154 57.157 scf_env_do_scf 11 5.9 0.000 0.001 49.150 49.150 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.478 40.478 velocity_verlet 10 3.0 0.001 0.001 34.767 34.822 dbcsr_multiply_generic 2055 12.4 0.109 0.114 28.652 28.980 qs_scf_new_mos 99 7.5 0.001 0.001 25.185 25.284 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.185 25.283 ot_scf_mini 99 9.5 0.002 0.003 23.951 24.063 multiply_cannon 2055 13.4 0.213 0.229 22.283 23.443 multiply_cannon_loop 2055 14.4 0.617 0.631 21.052 22.347 ot_mini 99 10.5 0.001 0.001 13.656 13.778 rebuild_ks_matrix 110 8.3 0.000 0.000 12.545 12.712 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.544 12.711 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.023 11.172 multiply_cannon_multrec 16440 15.4 3.784 4.899 9.515 10.511 mp_waitall_1 139946 16.5 7.305 10.403 7.305 10.403 qs_ot_get_derivative 99 11.5 0.001 0.001 9.109 9.226 init_scf_loop 11 6.9 0.000 0.000 8.631 8.632 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.604 7.564 prepare_preconditioner 11 7.9 0.000 0.000 6.847 6.864 make_preconditioner 11 8.9 0.000 0.000 6.847 6.864 sum_up_and_integrate 110 10.3 0.060 0.061 6.617 6.633 integrate_v_rspace 110 11.3 0.003 0.003 6.556 6.572 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.187 6.540 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.124 6.138 calculate_rho_elec 110 8.6 0.059 0.059 6.123 6.137 dbcsr_mm_accdrv_process 34862 16.1 4.425 5.345 5.586 5.717 init_scf_run 11 5.9 0.000 0.001 5.619 5.619 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.618 5.619 qs_ot_get_p 110 10.4 0.001 0.001 5.426 5.594 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.064 5.548 apply_single 110 13.6 0.000 0.000 5.064 5.547 make_m2s 4110 13.4 0.049 0.051 4.191 4.573 ot_diis_step 99 11.5 0.011 0.011 4.515 4.515 make_images 4110 14.4 0.393 0.511 4.074 4.456 density_rs2pw 110 9.6 0.004 0.005 3.169 4.425 multiply_cannon_sync_h2d 16440 15.4 3.657 4.301 3.657 4.301 rs_pw_transfer 902 11.9 0.010 0.011 2.698 3.944 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.803 3.808 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.114 3.755 pw_transfer 1331 11.6 0.066 0.074 3.735 3.751 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.627 3.637 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.442 3.442 grid_integrate_task_list 110 12.3 3.188 3.409 3.188 3.409 fft_wrap_pw1pw2_140 451 13.1 0.636 0.643 3.101 3.113 wfi_extrapolate 11 7.9 0.001 0.001 3.037 3.037 make_images_data 4110 15.4 0.042 0.046 2.458 2.966 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.897 2.907 cp_fm_diag_elpa_base 48 14.0 2.831 2.862 2.895 2.905 hybrid_alltoall_any 4261 16.3 0.106 0.376 2.192 2.853 cp_fm_cholesky_invert 11 10.9 2.799 2.806 2.799 2.806 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.796 2.798 calculate_dm_sparse 110 9.5 0.001 0.001 2.507 2.545 multiply_cannon_metrocomm4 14385 15.4 0.044 0.049 0.916 2.514 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.511 grid_collocate_task_list 110 9.6 2.125 2.497 2.125 2.497 mp_waitany 17072 13.8 1.188 2.497 1.188 2.497 calculate_first_density_matrix 1 7.0 0.000 0.000 2.485 2.486 fft3d_ps 1111 14.6 1.095 1.106 2.451 2.469 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.348 2.411 mp_irecv_dv 48980 15.7 0.846 2.387 0.846 2.387 potential_pw2rs 110 12.3 0.011 0.012 2.316 2.324 mp_sum_l 6594 12.7 1.556 2.254 1.556 2.254 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.906 2.148 mp_alltoall_d11v 2046 13.8 1.767 2.085 1.767 2.085 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.953 1.954 jit_kernel_multiply 10 16.3 0.764 1.907 0.764 1.907 dbcsr_complete_redistribute 325 12.2 0.323 0.372 1.434 1.884 mp_sum_dm 438 4.9 1.732 1.804 1.732 1.804 cp_fm_upper_to_full 70 14.2 1.383 1.766 1.383 1.766 md_write_output 11 3.9 0.055 1.728 0.056 1.748 md_output 10 3.0 0.000 0.000 0.058 1.747 update_particle_set 20 4.0 0.000 0.000 1.689 1.744 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.716 1.734 mp_allgather_i34 2055 14.4 0.555 1.649 0.555 1.649 cp_fm_cholesky_decompose 22 10.9 1.616 1.633 1.616 1.633 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.366 1.471 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.457 1.470 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.975 1.430 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.274 1.282 acc_transpose_blocks 16440 15.4 0.074 0.077 1.229 1.249 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=62.423000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.818182, yerr=7.107207 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 733.437952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.033 67.540 67.542 qs_mol_dyn_low 1 2.0 0.003 0.004 67.274 67.284 qs_forces 11 3.9 0.003 0.003 65.744 65.746 qs_energies 11 4.9 0.001 0.005 62.206 62.210 scf_env_do_scf 11 5.9 0.001 0.001 53.680 53.683 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 41.818 41.819 velocity_verlet 10 3.0 0.001 0.001 38.957 39.020 dbcsr_multiply_generic 2055 12.4 0.139 0.144 29.589 29.815 qs_scf_new_mos 99 7.5 0.001 0.001 26.551 26.680 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.550 26.679 ot_scf_mini 99 9.5 0.003 0.003 24.927 25.026 multiply_cannon 2055 13.4 0.245 0.263 22.635 23.645 multiply_cannon_loop 2055 14.4 0.889 0.910 21.078 21.701 ot_mini 99 10.5 0.001 0.001 14.184 14.301 multiply_cannon_multrec 24660 15.4 4.256 7.040 12.725 13.928 rebuild_ks_matrix 110 8.3 0.000 0.000 12.418 12.521 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.417 12.521 init_scf_loop 11 6.9 0.001 0.005 11.822 11.825 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.878 10.971 qs_ot_get_derivative 99 11.5 0.001 0.001 10.023 10.127 prepare_preconditioner 11 7.9 0.000 0.001 10.043 10.060 make_preconditioner 11 8.9 0.000 0.001 10.043 10.060 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.273 9.731 dbcsr_mm_accdrv_process 52304 16.0 7.001 8.122 8.323 9.212 sum_up_and_integrate 110 10.3 0.067 0.070 6.680 6.705 integrate_v_rspace 110 11.3 0.003 0.003 6.613 6.640 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.255 6.267 calculate_rho_elec 110 8.6 0.078 0.081 6.254 6.266 mp_waitall_1 121746 16.5 4.151 6.091 4.151 6.091 qs_ot_get_p 110 10.4 0.001 0.002 5.751 5.890 make_m2s 4110 13.4 0.059 0.060 5.275 5.740 init_scf_run 11 5.9 0.000 0.001 5.723 5.723 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.723 5.723 make_images 4110 14.4 0.577 0.701 5.132 5.596 cp_fm_upper_to_full 70 14.2 3.339 4.839 3.339 4.839 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.024 4.124 apply_single 110 13.6 0.000 0.000 4.024 4.124 ot_diis_step 99 11.5 0.011 0.011 4.124 4.124 qs_ot_p2m_diag 48 11.0 0.054 0.062 3.944 3.957 pw_transfer 1331 11.6 0.066 0.075 3.915 3.952 dbcsr_complete_redistribute 325 12.2 0.459 0.503 2.726 3.874 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.808 3.850 density_rs2pw 110 9.6 0.004 0.004 3.138 3.654 grid_integrate_task_list 110 12.3 3.270 3.483 3.270 3.483 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.471 3.472 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.297 3.345 multiply_cannon_sync_h2d 24660 15.4 3.189 3.331 3.189 3.331 copy_fm_to_dbcsr 174 11.2 0.001 0.002 2.174 3.315 fft_wrap_pw1pw2_140 451 13.1 0.666 0.687 3.247 3.294 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.372 3.231 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.038 3.093 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.286 3.091 make_images_data 4110 15.4 0.045 0.049 2.609 3.080 wfi_extrapolate 11 7.9 0.001 0.001 3.059 3.059 rs_pw_transfer 902 11.9 0.010 0.011 2.492 3.021 calculate_dm_sparse 110 9.5 0.001 0.001 2.908 2.940 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.923 2.933 cp_fm_diag_elpa_base 48 14.0 2.775 2.835 2.921 2.931 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.915 2.917 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.761 2.873 cp_fm_cholesky_invert 11 10.9 2.826 2.836 2.826 2.836 mp_alltoall_i22 605 13.7 1.677 2.834 1.677 2.834 fft3d_ps 1111 14.6 1.088 1.121 2.573 2.597 calculate_first_density_matrix 1 7.0 0.000 0.002 2.563 2.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.500 2.536 grid_collocate_task_list 110 9.6 2.222 2.513 2.222 2.513 potential_pw2rs 110 12.3 0.012 0.013 2.298 2.318 qs_energies_init_hamiltonians 11 5.9 0.012 0.016 2.285 2.286 mp_alltoall_d11v 2046 13.8 1.798 2.140 1.798 2.140 jit_kernel_multiply 10 15.8 0.987 1.813 0.987 1.813 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.798 cp_fm_cholesky_decompose 22 10.9 1.684 1.744 1.684 1.744 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.696 1.724 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.702 mp_waitany 13376 13.8 1.099 1.583 1.099 1.583 mp_sum_dm 438 4.9 1.490 1.569 1.490 1.569 acc_transpose_blocks 24660 15.4 0.105 0.107 1.537 1.557 multiply_cannon_metrocomm4 20550 15.4 0.059 0.061 0.845 1.549 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.526 1.542 mp_sum_l 6594 12.7 0.972 1.530 0.972 1.530 mp_allgather_i34 2055 14.4 0.653 1.517 0.653 1.517 md_write_output 11 3.9 0.063 1.495 0.065 1.517 md_output 10 3.0 0.000 0.000 0.066 1.516 update_particle_set 20 4.0 0.000 0.000 1.451 1.514 mp_irecv_dv 62702 16.1 0.744 1.465 0.744 1.465 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.175 0.837 1.371 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.542000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.000000, yerr=10.862780 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 838.672384E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.029 63.302 63.303 qs_mol_dyn_low 1 2.0 0.003 0.003 63.052 63.061 qs_forces 11 3.9 0.003 0.003 60.646 60.647 qs_energies 11 4.9 0.001 0.001 56.839 56.844 scf_env_do_scf 11 5.9 0.000 0.001 48.152 48.152 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 38.840 38.841 velocity_verlet 10 3.0 0.001 0.001 36.678 36.827 dbcsr_multiply_generic 2055 12.4 0.106 0.110 24.144 24.347 qs_scf_new_mos 99 7.5 0.001 0.001 22.657 22.716 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.656 22.716 ot_scf_mini 99 9.5 0.002 0.002 21.344 21.374 multiply_cannon 2055 13.4 0.249 0.262 17.772 19.414 multiply_cannon_loop 2055 14.4 0.324 0.337 16.265 16.805 rebuild_ks_matrix 110 8.3 0.000 0.000 12.650 12.662 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.650 12.662 ot_mini 99 10.5 0.001 0.001 11.313 11.319 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.229 11.244 init_scf_loop 11 6.9 0.000 0.000 9.258 9.258 mp_waitall_1 103326 16.6 6.975 8.623 6.975 8.623 multiply_cannon_multrec 8220 15.4 3.209 4.413 7.430 8.327 prepare_preconditioner 11 7.9 0.000 0.000 7.506 7.508 make_preconditioner 11 8.9 0.000 0.000 7.506 7.508 qs_ot_get_derivative 99 11.5 0.001 0.001 7.415 7.441 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.020 7.133 sum_up_and_integrate 110 10.3 0.080 0.081 7.092 7.105 integrate_v_rspace 110 11.3 0.003 0.003 7.012 7.024 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.902 6.916 calculate_rho_elec 110 8.6 0.114 0.115 6.902 6.915 qs_ot_get_p 110 10.4 0.001 0.001 5.921 5.942 init_scf_run 11 5.9 0.000 0.001 5.460 5.460 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.460 5.460 make_m2s 4110 13.4 0.038 0.040 4.870 5.115 dbcsr_mm_accdrv_process 17442 15.9 2.852 3.770 4.092 5.002 make_images 4110 14.4 0.636 0.695 4.738 4.980 pw_transfer 1331 11.6 0.066 0.071 4.373 4.392 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.331 4.338 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.265 4.289 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.867 4.194 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.967 3.968 density_rs2pw 110 9.6 0.004 0.004 3.539 3.912 ot_diis_step 99 11.5 0.012 0.012 3.856 3.857 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.770 3.794 apply_single 110 13.6 0.000 0.000 3.770 3.794 cp_fm_cholesky_invert 11 10.9 3.764 3.768 3.764 3.768 fft_wrap_pw1pw2_140 451 13.1 0.829 0.843 3.583 3.619 grid_integrate_task_list 110 12.3 3.344 3.538 3.344 3.538 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.326 3.334 cp_fm_diag_elpa_base 48 14.0 3.258 3.291 3.325 3.332 make_images_data 4110 15.4 0.037 0.043 2.734 3.267 hybrid_alltoall_any 4261 16.3 0.199 0.857 2.577 3.175 rs_pw_transfer 902 11.9 0.010 0.010 2.782 3.163 wfi_extrapolate 11 7.9 0.001 0.001 3.072 3.072 multiply_cannon_sync_h2d 8220 15.4 2.899 2.955 2.899 2.955 fft3d_ps 1111 14.6 1.147 1.179 2.816 2.833 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.694 2.694 calculate_dm_sparse 110 9.5 0.001 0.001 2.564 2.610 grid_collocate_task_list 110 9.6 2.317 2.581 2.317 2.581 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.579 2.581 potential_pw2rs 110 12.3 0.014 0.015 2.515 2.531 mp_sum_dm 438 4.9 2.291 2.443 2.291 2.443 md_write_output 11 3.9 0.083 1.320 0.151 2.390 update_particle_set 20 4.0 0.000 0.000 2.239 2.389 md_output 10 3.0 0.000 0.000 0.152 2.388 calculate_first_density_matrix 1 7.0 0.000 0.000 2.280 2.281 mp_alltoall_d11v 2046 13.8 1.989 2.241 1.989 2.241 cp_fm_cholesky_decompose 22 10.9 2.199 2.210 2.199 2.210 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.191 2.196 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.161 2.174 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.125 2.141 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.771 1.996 mp_allgather_i34 2055 14.4 0.683 1.898 0.683 1.898 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.267 1.818 dbcsr_complete_redistribute 325 12.2 0.545 0.562 1.649 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.752 1.755 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.527 1.651 mp_waitany 9240 13.8 1.163 1.590 1.163 1.590 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.467 1.478 mp_alltoall_z22v 1111 16.6 1.433 1.458 1.433 1.458 qs_create_task_list 11 7.9 0.000 0.001 1.231 1.338 generate_qs_task_list 11 8.9 0.374 0.443 1.231 1.338 rs_gather_matrices 110 12.3 0.325 0.368 1.057 1.272 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=63.303000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=787.454545, yerr=11.765617 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 1.339798E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 89.352 89.353 qs_mol_dyn_low 1 2.0 0.003 0.003 88.341 89.053 qs_forces 11 3.9 0.003 0.003 86.787 86.788 qs_energies 11 4.9 0.001 0.001 82.576 82.579 scf_env_do_scf 11 5.9 0.000 0.001 72.418 72.419 velocity_verlet 10 3.0 0.001 0.001 57.046 57.226 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.160 44.163 dbcsr_multiply_generic 2055 12.4 0.119 0.125 28.444 28.509 init_scf_loop 11 6.9 0.000 0.000 28.187 28.192 prepare_preconditioner 11 7.9 0.000 0.000 26.174 26.180 make_preconditioner 11 8.9 0.000 0.000 26.174 26.180 qs_scf_new_mos 99 7.5 0.001 0.001 26.018 26.067 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.017 26.066 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.426 25.654 ot_scf_mini 99 9.5 0.002 0.002 24.237 24.263 multiply_cannon 2055 13.4 0.348 0.390 21.452 22.272 multiply_cannon_loop 2055 14.4 0.342 0.345 19.654 19.975 cp_fm_upper_to_full 70 14.2 12.500 17.977 12.500 17.977 rebuild_ks_matrix 110 8.3 0.000 0.001 14.030 14.059 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.030 14.058 ot_mini 99 10.5 0.001 0.001 13.490 13.516 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.685 12.711 dbcsr_complete_redistribute 325 12.2 1.046 1.089 7.356 10.494 multiply_cannon_multrec 8220 15.4 4.373 4.547 9.554 9.641 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.328 9.463 qs_ot_get_derivative 99 11.5 0.001 0.001 8.964 8.990 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.732 8.846 mp_alltoall_i22 605 13.7 5.368 8.498 5.368 8.498 mp_waitall_1 84994 16.7 7.436 8.216 7.436 8.216 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.684 7.723 calculate_rho_elec 110 8.6 0.225 0.225 7.684 7.722 sum_up_and_integrate 110 10.3 0.151 0.152 7.292 7.308 integrate_v_rspace 110 11.3 0.004 0.004 7.141 7.157 init_scf_run 11 5.9 0.000 0.001 5.830 5.830 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.830 5.830 make_m2s 4110 13.4 0.042 0.043 5.247 5.750 qs_ot_get_p 110 10.4 0.001 0.001 5.552 5.598 make_images 4110 14.4 0.878 0.932 5.056 5.560 cp_fm_cholesky_invert 11 10.9 5.354 5.358 5.354 5.358 dbcsr_mm_accdrv_process 11614 15.7 3.272 3.675 5.039 5.293 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.640 5.113 apply_single 110 13.6 0.000 0.000 4.640 5.112 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.738 5.059 pw_transfer 1331 11.6 0.075 0.076 4.971 4.985 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.853 4.867 ot_diis_step 99 11.5 0.016 0.021 4.501 4.501 fft_wrap_pw1pw2_140 451 13.1 1.337 1.349 4.214 4.238 qs_ot_p2m_diag 48 11.0 0.150 0.155 4.048 4.056 multiply_cannon_sync_h2d 8220 15.4 3.943 3.947 3.943 3.947 density_rs2pw 110 9.6 0.004 0.004 3.694 3.733 grid_integrate_task_list 110 12.3 3.674 3.725 3.674 3.725 qs_energies_init_hamiltonians 11 5.9 0.023 0.033 3.709 3.709 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.604 3.605 hybrid_alltoall_any 4261 16.3 0.257 0.559 2.892 3.556 make_images_data 4110 15.4 0.040 0.043 2.804 3.454 wfi_extrapolate 11 7.9 0.001 0.001 3.362 3.362 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.867 3.314 calculate_dm_sparse 110 9.5 0.001 0.001 3.148 3.169 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.016 3.016 cp_fm_diag_elpa_base 48 14.0 2.480 2.680 3.013 3.014 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.940 2.941 fft3d_ps 1111 14.6 1.304 1.330 2.783 2.790 grid_collocate_task_list 110 9.6 2.669 2.679 2.669 2.679 potential_pw2rs 110 12.3 0.021 0.021 2.410 2.412 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.394 2.411 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.399 2.404 calculate_first_density_matrix 1 7.0 0.000 0.000 2.344 2.347 rs_pw_transfer 902 11.9 0.010 0.011 2.301 2.337 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.190 2.254 md_output 10 3.0 0.000 0.000 0.284 2.244 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.095 2.192 cp_fm_cholesky_decompose 22 10.9 2.042 2.062 2.042 2.062 mp_alltoall_d11v 2046 13.8 1.918 2.021 1.918 2.021 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.945 1.949 qs_create_task_list 11 7.9 0.000 0.001 1.882 1.933 generate_qs_task_list 11 8.9 0.729 0.783 1.881 1.932 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.778 1.790 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=89.353000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1204.000000, yerr=54.757316 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 631.787520E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 57767. MP_Allreduce 11084 796. MP_Sync 87 MP_Alltoall 2226 2529087. 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 0.032 0.081 220.748 220.750 qs_mol_dyn_low 1 2.0 0.003 0.004 220.181 220.195 qs_forces 11 3.9 0.013 0.013 220.095 220.096 qs_energies 11 4.9 0.001 0.001 214.308 214.324 scf_env_do_scf 11 5.9 0.001 0.001 195.332 195.336 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 171.853 171.856 velocity_verlet 10 3.0 0.001 0.002 131.251 131.252 qs_scf_new_mos 117 7.6 0.001 0.001 129.622 129.851 qs_scf_loop_do_ot 117 8.6 0.001 0.001 129.621 129.850 dbcsr_multiply_generic 2507 12.6 0.183 0.187 128.423 129.248 ot_scf_mini 117 9.6 0.003 0.003 122.917 123.149 multiply_cannon 2507 13.6 0.239 0.248 103.739 105.697 multiply_cannon_loop 2507 14.6 2.091 2.147 101.322 103.224 ot_mini 117 10.6 0.001 0.001 67.565 67.795 multiply_cannon_multrec 60168 15.6 33.254 36.213 42.021 45.386 qs_ot_get_derivative 117 11.6 0.001 0.001 42.527 42.744 rebuild_ks_matrix 128 8.3 0.001 0.001 34.985 35.319 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.985 35.318 mp_waitall_1 267128 16.5 31.557 35.113 31.557 35.113 qs_ot_get_p 128 10.4 0.001 0.001 33.185 33.517 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.388 31.687 multiply_cannon_sync_h2d 60168 15.6 27.357 29.637 27.357 29.637 qs_ot_p2m_diag 83 11.4 0.079 0.091 26.266 26.342 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 24.481 25.587 apply_single 128 13.6 0.001 0.001 24.480 25.587 ot_diis_step 117 11.6 0.008 0.008 24.728 24.729 cp_dbcsr_syevd 83 12.4 0.005 0.005 23.533 23.535 init_scf_loop 11 6.9 0.000 0.001 23.398 23.399 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.293 20.467 cp_fm_diag_elpa 83 13.4 0.000 0.000 20.103 20.137 cp_fm_diag_elpa_base 83 14.4 20.007 20.044 20.099 20.130 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 16.679 19.212 prepare_preconditioner 11 7.9 0.000 0.000 18.608 18.669 make_preconditioner 11 8.9 0.000 0.000 18.608 18.669 make_full_inverse_cholesky 11 9.9 0.000 0.000 17.820 18.026 sum_up_and_integrate 128 10.3 0.088 0.105 15.396 15.415 integrate_v_rspace 128 11.3 0.004 0.005 15.308 15.330 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.027 15.146 calculate_rho_elec 128 8.7 0.046 0.065 15.027 15.146 make_m2s 5014 13.6 0.105 0.113 14.615 15.013 make_images 5014 14.6 0.401 0.421 14.430 14.842 init_scf_run 11 5.9 0.000 0.001 14.765 14.765 scf_env_initial_rho_setup 11 6.9 0.001 0.001 14.765 14.765 density_rs2pw 128 9.7 0.006 0.008 8.285 11.698 rs_pw_transfer 1046 11.9 0.016 0.019 7.344 10.848 cp_fm_cholesky_invert 11 10.9 10.777 10.786 10.777 10.786 calculate_dm_sparse 128 9.5 0.001 0.001 10.429 10.530 dbcsr_mm_accdrv_process 124484 16.2 3.186 3.340 8.335 9.929 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 7.268 9.905 mp_sum_l 7950 12.9 8.806 9.882 8.806 9.882 wfi_extrapolate 11 7.9 0.001 0.001 9.552 9.552 pw_transfer 1547 11.6 0.075 0.088 8.519 8.731 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 8.315 8.528 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.308 8.466 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.051 8.151 make_images_data 5014 15.6 0.065 0.071 7.107 8.066 grid_integrate_task_list 128 12.3 7.022 7.687 7.022 7.687 hybrid_alltoall_any 5200 16.5 0.289 2.256 6.146 7.342 fft_wrap_pw1pw2_140 523 13.2 1.261 1.315 6.819 7.044 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.753 6.765 fft3d_ps 1291 14.7 2.150 2.790 6.096 6.445 mp_waitany 16020 13.9 2.741 6.275 2.741 6.275 mp_alltoall_d11v 2415 14.1 4.464 6.225 4.464 6.225 grid_collocate_task_list 128 9.7 4.678 5.975 4.678 5.975 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.781 5.881 potential_pw2rs 128 12.3 0.009 0.010 5.717 5.786 cp_fm_cholesky_decompose 22 10.9 5.609 5.623 5.609 5.623 rs_pw_transfer_RS2PW_140 139 11.5 0.280 0.301 2.154 5.595 calculate_first_density_matrix 1 7.0 0.000 0.001 5.009 5.017 jit_kernel_multiply 13 15.9 3.059 4.497 3.059 4.497 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.443 4.481 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=220.750000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.909091, yerr=6.556808 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 831.184896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58220. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6356612. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.123 200.918 200.919 qs_mol_dyn_low 1 2.0 0.003 0.004 200.235 200.249 qs_forces 11 3.9 0.009 0.052 200.096 200.098 qs_energies 11 4.9 0.002 0.007 193.051 193.063 scf_env_do_scf 11 5.9 0.001 0.002 175.563 175.572 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 140.061 140.063 velocity_verlet 10 3.0 0.001 0.001 125.681 125.684 dbcsr_multiply_generic 2507 12.6 0.193 0.206 98.834 99.977 qs_scf_new_mos 117 7.6 0.001 0.001 98.105 98.550 qs_scf_loop_do_ot 117 8.6 0.001 0.001 98.104 98.549 ot_scf_mini 117 9.6 0.004 0.005 93.282 93.828 multiply_cannon 2507 13.6 0.478 0.533 78.017 82.073 multiply_cannon_loop 2507 14.6 1.244 1.276 74.514 76.859 ot_mini 117 10.6 0.001 0.001 50.607 51.123 mp_waitall_1 214728 16.6 25.623 39.371 25.623 39.371 multiply_cannon_multrec 30084 15.6 22.102 26.960 31.745 37.075 init_scf_loop 11 6.9 0.005 0.025 35.405 35.407 rebuild_ks_matrix 128 8.3 0.001 0.001 34.596 35.118 qs_ks_build_kohn_sham_matrix 128 9.3 0.021 0.048 34.596 35.117 qs_ks_update_qs_env 128 7.6 0.001 0.001 31.085 31.564 prepare_preconditioner 11 7.9 0.000 0.000 30.820 30.877 make_preconditioner 11 8.9 0.000 0.001 30.820 30.877 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.447 30.038 qs_ot_get_derivative 117 11.6 0.001 0.002 28.659 29.192 multiply_cannon_metrocomm3 30084 15.6 0.089 0.093 15.683 28.328 qs_ot_get_p 128 10.4 0.001 0.001 24.203 24.963 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.006 23.090 apply_single 128 13.6 0.001 0.001 22.006 23.089 multiply_cannon_sync_h2d 30084 15.6 19.458 22.095 19.458 22.095 ot_diis_step 117 11.6 0.015 0.015 21.780 21.782 qs_ot_p2m_diag 83 11.4 0.187 0.215 19.088 19.122 cp_fm_cholesky_invert 11 10.9 18.018 18.031 18.018 18.031 cp_dbcsr_syevd 83 12.4 0.006 0.006 17.870 17.872 make_m2s 5014 13.6 0.090 0.097 15.084 16.751 make_images 5014 14.6 1.164 1.350 14.867 16.530 sum_up_and_integrate 128 10.3 0.116 0.133 15.708 15.738 integrate_v_rspace 128 11.3 0.004 0.004 15.592 15.628 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.331 15.381 calculate_rho_elec 128 8.7 0.087 0.104 15.331 15.380 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.413 14.442 cp_fm_diag_elpa_base 83 14.4 14.114 14.236 14.406 14.437 init_scf_run 11 5.9 0.000 0.001 12.467 12.469 scf_env_initial_rho_setup 11 6.9 0.017 0.021 12.467 12.468 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.377 11.798 density_rs2pw 128 9.7 0.006 0.007 8.558 11.264 make_images_data 5014 15.6 0.063 0.070 9.060 11.125 multiply_cannon_metrocomm4 27577 15.6 0.102 0.116 3.699 10.290 hybrid_alltoall_any 5200 16.5 0.341 1.508 7.720 10.248 rs_pw_transfer 1046 11.9 0.014 0.016 7.260 10.011 mp_irecv_dv 69486 16.3 3.502 9.902 3.502 9.902 dbcsr_mm_accdrv_process 62242 16.2 4.426 5.217 9.095 9.712 pw_transfer 1547 11.6 0.086 0.095 9.451 9.495 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 9.225 9.272 wfi_extrapolate 11 7.9 0.001 0.001 8.766 8.766 cp_fm_cholesky_decompose 22 10.9 7.877 7.958 7.877 7.958 fft_wrap_pw1pw2_140 523 13.2 1.333 1.356 7.876 7.950 grid_integrate_task_list 128 12.3 7.147 7.701 7.147 7.701 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.408 7.089 calculate_dm_sparse 128 9.5 0.001 0.001 6.527 6.677 fft3d_ps 1291 14.7 2.815 2.991 6.538 6.577 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.380 6.390 grid_collocate_task_list 128 9.7 4.793 5.997 4.793 5.997 potential_pw2rs 128 12.3 0.016 0.018 5.871 5.903 mp_sum_l 7950 12.9 3.991 5.713 3.991 5.713 mp_alltoall_d11v 2415 14.1 4.463 5.645 4.463 5.645 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.001 0.008 5.444 5.605 mp_waitany 11748 13.9 2.724 5.520 2.724 5.520 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.420 5.483 rs_pw_transfer_RS2PW_140 139 11.5 0.351 0.382 2.261 4.979 mp_allgather_i34 2507 14.6 2.069 4.972 2.069 4.972 dbcsr_complete_redistribute 395 12.7 0.791 0.888 3.367 4.270 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=200.919000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=792.909091, yerr=0.287480 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 941.326336E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172558800 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58284. MP_Allreduce 11070 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.183 193.569 193.575 qs_mol_dyn_low 1 2.0 0.003 0.004 192.757 192.771 qs_forces 11 3.9 0.005 0.006 192.654 192.659 qs_energies 11 4.9 0.019 0.142 185.749 185.759 scf_env_do_scf 11 5.9 0.001 0.002 167.076 167.077 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 127.528 127.529 velocity_verlet 10 3.0 0.002 0.006 121.849 121.851 qs_scf_new_mos 116 7.6 0.001 0.001 87.445 87.764 qs_scf_loop_do_ot 116 8.6 0.001 0.001 87.444 87.763 dbcsr_multiply_generic 2485 12.5 0.182 0.187 85.672 86.972 ot_scf_mini 116 9.6 0.003 0.004 83.270 83.647 multiply_cannon 2485 13.5 0.499 0.516 64.288 68.479 multiply_cannon_loop 2485 14.5 0.852 0.878 60.540 63.965 ot_mini 116 10.6 0.001 0.001 44.270 44.679 init_scf_loop 11 6.9 0.001 0.003 39.434 39.435 mp_waitall_1 169034 16.6 27.822 39.420 27.822 39.420 prepare_preconditioner 11 7.9 0.000 0.000 35.212 35.261 make_preconditioner 11 8.9 0.000 0.001 35.212 35.261 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.803 34.238 rebuild_ks_matrix 127 8.3 0.001 0.001 31.943 32.457 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 31.943 32.456 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.677 29.148 multiply_cannon_metrocomm3 19880 15.5 0.057 0.062 16.156 28.130 multiply_cannon_multrec 19880 15.5 13.354 16.150 22.150 25.340 qs_ot_get_derivative 116 11.6 0.002 0.002 24.523 24.906 qs_ot_get_p 127 10.4 0.001 0.001 24.071 24.548 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.864 20.746 apply_single 127 13.6 0.001 0.001 19.863 20.745 ot_diis_step 116 11.6 0.018 0.018 19.618 19.618 qs_ot_p2m_diag 82 11.4 0.263 0.270 19.394 19.432 cp_dbcsr_syevd 82 12.4 0.005 0.006 18.252 18.254 cp_fm_cholesky_invert 11 10.9 17.230 17.243 17.230 17.243 make_m2s 4970 13.5 0.079 0.083 15.709 16.863 make_images 4970 14.5 1.182 1.274 15.476 16.627 multiply_cannon_sync_h2d 19880 15.5 14.109 15.795 14.109 15.795 sum_up_and_integrate 127 10.3 0.131 0.146 15.545 15.570 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.490 15.549 calculate_rho_elec 127 8.7 0.129 0.144 15.490 15.549 integrate_v_rspace 127 11.3 0.004 0.004 15.414 15.439 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.543 14.564 cp_fm_diag_elpa_base 82 14.4 14.095 14.290 14.538 14.560 init_scf_run 11 5.9 0.000 0.001 13.262 13.263 scf_env_initial_rho_setup 11 6.9 0.012 0.084 13.262 13.263 make_images_data 4970 15.5 0.057 0.066 9.750 11.346 hybrid_alltoall_any 5155 16.4 0.432 1.985 8.484 10.526 density_rs2pw 127 9.7 0.006 0.006 8.222 10.311 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.820 10.124 dbcsr_mm_accdrv_process 41158 16.2 4.529 6.261 8.267 9.548 pw_transfer 1535 11.6 0.087 0.102 9.400 9.512 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.176 9.291 multiply_cannon_metrocomm4 17395 15.5 0.061 0.070 3.380 9.132 mp_irecv_dv 49801 16.2 3.261 8.885 3.261 8.885 rs_pw_transfer 1038 11.9 0.013 0.014 6.649 8.805 cp_fm_cholesky_decompose 22 10.9 8.726 8.744 8.726 8.744 wfi_extrapolate 11 7.9 0.001 0.001 8.001 8.001 fft_wrap_pw1pw2_140 519 13.2 1.397 1.433 7.860 7.988 grid_integrate_task_list 127 12.3 7.213 7.711 7.213 7.711 calculate_dm_sparse 127 9.5 0.001 0.001 7.625 7.709 cp_fm_upper_to_full 104 14.8 5.861 7.410 5.861 7.410 dbcsr_complete_redistribute 393 12.7 1.176 1.212 4.883 6.759 fft3d_ps 1281 14.7 2.719 2.938 6.327 6.410 grid_collocate_task_list 127 9.7 4.943 5.825 4.943 5.825 mp_sum_l 7884 12.9 3.939 5.752 3.939 5.752 potential_pw2rs 127 12.3 0.020 0.023 5.601 5.632 mp_alltoall_d11v 2401 14.1 4.751 5.611 4.751 5.611 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.600 5.605 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.879 5.575 copy_fm_to_dbcsr 208 11.6 0.002 0.003 3.604 5.481 mp_allgather_i34 2485 14.5 2.270 5.288 2.270 5.288 calculate_first_density_matrix 1 7.0 0.008 0.060 5.021 5.038 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.605 4.745 mp_waitany 11660 13.9 2.565 4.698 2.565 4.698 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.217 4.256 transfer_fm_to_dbcsr 11 9.9 0.020 0.029 2.390 4.204 rs_pw_transfer_RS2PW_140 138 11.5 0.329 0.347 2.000 4.119 mp_alltoall_i22 712 14.1 2.100 4.084 2.100 4.084 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 4.012 4.040 qs_energies_init_hamiltonians 11 5.9 0.063 0.086 3.926 3.928 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=193.575000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=893.909091, yerr=7.925271 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 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 11444706349056 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 92.796578E+12 0.0% 0.0% 100.0% flops max/rank 4.320337E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500608 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 1.153556E+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 660837763760 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58195. MP_Allreduce 11085 1083. MP_Sync 86 MP_Alltoall 1700 12496381. 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 0.043 0.078 200.297 200.298 qs_mol_dyn_low 1 2.0 0.003 0.004 199.757 199.788 qs_forces 11 3.9 0.004 0.005 199.557 199.562 qs_energies 11 4.9 0.011 0.040 192.244 192.255 scf_env_do_scf 11 5.9 0.001 0.001 174.920 174.927 velocity_verlet 10 3.0 0.001 0.002 132.509 132.514 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 123.641 123.643 qs_scf_new_mos 116 7.6 0.001 0.001 85.430 85.693 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.430 85.692 dbcsr_multiply_generic 2485 12.5 0.187 0.192 80.367 81.354 ot_scf_mini 116 9.6 0.003 0.004 80.907 81.195 multiply_cannon 2485 13.5 0.551 0.583 54.819 58.528 multiply_cannon_loop 2485 14.5 1.178 1.205 50.734 53.242 init_scf_loop 11 6.9 0.002 0.016 51.145 51.146 prepare_preconditioner 11 7.9 0.000 0.000 46.873 46.906 make_preconditioner 11 8.9 0.001 0.011 46.873 46.906 make_full_inverse_cholesky 11 9.9 0.013 0.024 40.385 45.506 ot_mini 116 10.6 0.001 0.001 43.476 43.740 rebuild_ks_matrix 127 8.3 0.001 0.001 30.321 30.683 qs_ks_build_kohn_sham_matrix 127 9.3 0.023 0.046 30.320 30.682 multiply_cannon_multrec 29820 15.5 13.820 19.269 25.670 30.550 mp_waitall_1 146592 16.7 18.197 27.727 18.197 27.727 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.260 27.590 qs_ot_get_derivative 116 11.6 0.002 0.002 23.800 24.083 qs_ot_get_p 127 10.4 0.001 0.001 23.574 23.897 make_m2s 4970 13.5 0.095 0.098 20.821 21.934 make_images 4970 14.5 1.941 2.244 20.512 21.625 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.032 19.702 apply_single 127 13.6 0.001 0.001 19.032 19.702 ot_diis_step 116 11.6 0.018 0.019 19.546 19.547 qs_ot_p2m_diag 82 11.4 0.338 0.385 19.148 19.209 cp_fm_cholesky_invert 11 10.9 19.116 19.126 19.116 19.126 cp_dbcsr_syevd 82 12.4 0.005 0.006 17.734 17.737 cp_fm_upper_to_full 104 14.8 11.622 17.191 11.622 17.191 multiply_cannon_metrocomm3 29820 15.5 0.047 0.049 6.542 16.073 sum_up_and_integrate 127 10.3 0.139 0.152 15.231 15.253 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.179 15.204 calculate_rho_elec 127 8.7 0.171 0.187 15.178 15.204 integrate_v_rspace 127 11.3 0.004 0.004 15.092 15.118 cp_fm_diag_elpa 82 13.4 0.000 0.000 14.217 14.228 cp_fm_diag_elpa_base 82 14.4 13.165 13.531 14.210 14.222 dbcsr_complete_redistribute 393 12.7 1.507 1.633 9.258 13.171 multiply_cannon_sync_h2d 29820 15.5 11.586 13.108 11.586 13.108 make_images_data 4970 15.5 0.060 0.064 11.268 13.082 dbcsr_mm_accdrv_process 61748 16.2 7.261 8.044 11.430 11.958 copy_fm_to_dbcsr 208 11.6 0.001 0.002 7.788 11.691 hybrid_alltoall_any 5155 16.4 0.524 2.198 10.089 11.681 init_scf_run 11 5.9 0.000 0.001 11.310 11.311 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.310 11.311 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.463 10.200 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.881 10.092 mp_alltoall_i22 712 14.1 5.757 9.581 5.757 9.581 pw_transfer 1535 11.6 0.086 0.102 9.304 9.388 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.079 9.169 density_rs2pw 127 9.7 0.006 0.006 7.551 9.119 cp_fm_cholesky_decompose 22 10.9 9.006 9.094 9.006 9.094 grid_integrate_task_list 127 12.3 7.432 8.096 7.432 8.096 wfi_extrapolate 11 7.9 0.001 0.001 8.090 8.090 fft_wrap_pw1pw2_140 519 13.2 1.546 1.578 7.905 8.007 multiply_cannon_metrocomm4 24850 15.5 0.078 0.088 2.860 7.884 mp_irecv_dv 75445 16.2 2.713 7.603 2.713 7.603 rs_pw_transfer 1038 11.9 0.013 0.014 5.634 7.244 calculate_dm_sparse 127 9.5 0.001 0.001 6.218 6.338 fft3d_ps 1281 14.7 2.811 2.889 6.033 6.094 grid_collocate_task_list 127 9.7 5.096 5.895 5.096 5.895 mp_alltoall_d11v 2401 14.1 5.075 5.746 5.075 5.746 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.566 5.637 potential_pw2rs 127 12.3 0.022 0.023 5.068 5.094 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.625 4.702 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.421 4.521 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.473 4.474 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.356 4.418 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=200.298000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1083.818182, yerr=22.192620 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 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 11613083000832 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 94.228655E+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 6806314816 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 1.509773E+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 726591807936 ------------------------------------------------------------------------------- - - - 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 4030 57813. MP_Allreduce 11152 1164. MP_Sync 87 MP_Alltoall 1724 18848449. 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 0.034 0.060 191.739 191.739 qs_mol_dyn_low 1 2.0 0.003 0.003 191.267 191.280 qs_forces 11 3.9 0.004 0.004 190.442 190.445 qs_energies 11 4.9 0.001 0.001 182.611 182.619 scf_env_do_scf 11 5.9 0.001 0.001 164.082 164.085 velocity_verlet 10 3.0 0.001 0.001 125.573 125.622 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 124.028 124.029 qs_scf_new_mos 118 7.6 0.001 0.001 83.152 83.255 qs_scf_loop_do_ot 118 8.6 0.001 0.001 83.151 83.254 dbcsr_multiply_generic 2527 12.6 0.196 0.209 78.529 79.095 ot_scf_mini 118 9.6 0.003 0.004 78.477 78.560 multiply_cannon 2527 13.6 0.590 0.627 56.717 61.964 multiply_cannon_loop 2527 14.6 0.449 0.459 51.381 52.581 ot_mini 118 10.6 0.001 0.001 42.200 42.293 init_scf_loop 11 6.9 0.000 0.000 39.890 39.891 mp_waitall_1 126780 16.7 28.540 37.235 28.540 37.235 prepare_preconditioner 11 7.9 0.000 0.000 35.823 35.856 make_preconditioner 11 8.9 0.000 0.000 35.823 35.856 make_full_inverse_cholesky 11 9.9 0.016 0.026 33.618 33.936 rebuild_ks_matrix 129 8.3 0.001 0.001 31.363 31.490 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 31.362 31.489 qs_ks_update_qs_env 129 7.6 0.001 0.001 28.412 28.522 qs_ot_get_p 129 10.4 0.001 0.001 22.722 22.823 qs_ot_get_derivative 118 11.6 0.002 0.002 21.728 21.817 cp_fm_cholesky_invert 11 10.9 20.943 20.950 20.943 20.950 make_m2s 5054 13.6 0.066 0.072 17.752 20.750 multiply_cannon_multrec 10108 15.6 10.430 14.764 17.814 20.665 make_images 5054 14.6 2.320 2.670 17.441 20.442 ot_diis_step 118 11.6 0.020 0.021 20.370 20.370 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.925 20.217 apply_single 129 13.6 0.001 0.001 19.925 20.216 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 12.390 20.193 qs_ot_p2m_diag 83 11.4 0.495 0.500 18.450 18.467 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.157 17.159 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.506 16.554 calculate_rho_elec 129 8.7 0.258 0.269 16.505 16.553 sum_up_and_integrate 129 10.3 0.183 0.193 16.015 16.062 integrate_v_rspace 129 11.3 0.004 0.004 15.832 15.886 make_images_data 5054 15.6 0.050 0.060 10.916 13.996 cp_fm_diag_elpa 83 13.4 0.000 0.000 13.668 13.676 cp_fm_diag_elpa_base 83 14.4 13.400 13.502 13.661 13.669 hybrid_alltoall_any 5240 16.5 0.831 3.708 10.547 13.099 multiply_cannon_sync_h2d 10108 15.6 11.635 12.298 11.635 12.298 init_scf_run 11 5.9 0.000 0.001 11.538 11.538 scf_env_initial_rho_setup 11 6.9 0.013 0.013 11.538 11.538 pw_transfer 1559 11.6 0.087 0.097 10.297 10.322 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 10.071 10.104 cp_fm_cholesky_decompose 22 10.9 9.058 9.171 9.058 9.171 mp_allgather_i34 2527 14.6 3.517 8.971 3.517 8.971 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 8.870 8.934 multiply_cannon_metrocomm1 10108 15.6 0.028 0.029 5.610 8.799 fft_wrap_pw1pw2_140 527 13.2 1.922 1.964 8.702 8.745 density_rs2pw 129 9.7 0.006 0.006 7.773 8.685 grid_integrate_task_list 129 12.3 7.778 8.369 7.778 8.369 wfi_extrapolate 11 7.9 0.001 0.001 8.129 8.129 dbcsr_mm_accdrv_process 20926 16.1 2.703 3.504 7.018 7.678 mp_alltoall_d11v 2423 14.1 5.662 6.734 5.662 6.734 fft3d_ps 1301 14.7 2.784 2.853 6.529 6.562 calculate_dm_sparse 129 9.5 0.001 0.001 6.397 6.460 grid_collocate_task_list 129 9.7 5.510 6.186 5.510 6.186 rs_pw_transfer 1054 12.0 0.013 0.014 5.150 6.092 dbcsr_complete_redistribute 395 12.7 2.159 2.219 5.576 6.009 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.703 5.723 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.311 5.312 potential_pw2rs 129 12.3 0.026 0.026 5.177 5.197 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.488 4.506 multiply_cannon_metrocomm4 7581 15.6 0.023 0.026 1.891 4.126 mp_irecv_dv 29086 15.9 1.857 4.049 1.857 4.049 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.965 4.034 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.638 4.004 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.910 3.934 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.551 3.881 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.861 3.875 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=191.739000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1402.909091, yerr=41.068578 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1379368673280 0.0% 0.0% 100.0% flops 32 x 32 x 32 1879048192000 0.0% 0.0% 100.0% flops 22 x 9 x 32 1915308527616 0.0% 0.0% 100.0% flops 9 x 22 x 32 1920862715904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2655609651200 0.0% 0.0% 100.0% flops 32 x 32 x 9 4261681299456 0.0% 0.0% 100.0% flops 32 x 32 x 22 5208721588224 0.0% 0.0% 100.0% flops 9 x 32 x 32 5250698182656 0.0% 0.0% 100.0% flops 22 x 32 x 32 6417520001024 0.0% 0.0% 100.0% flops 9 x 32 x 9 11192137887744 0.0% 0.0% 100.0% flops 22 x 32 x 9 14689248583680 0.0% 0.0% 100.0% flops 9 x 32 x 22 14689248583680 0.0% 0.0% 100.0% flops 22 x 32 x 22 19195467382784 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 90.654921E+12 0.0% 0.0% 100.0% flops max/rank 11.338554E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6554377184 0.0% 0.0% 100.0% number of processed stacks 1901296 0.0% 0.0% 100.0% average stack size 0.0 0.0 3447.3 marketing flops 140.305844E+12 ------------------------------------------------------------------------------- # multiplications 2425 max memory usage/rank 3.130671E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 97000 MPI messages size (bytes): total size 1.101303E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.353643E+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 43688 33915142144 4194304 < size <= 16777216 42880 367001600000 16777216 < size 9816 700384012944 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3930 60185. MP_Allreduce 10903 1510. MP_Sync 85 MP_Alltoall 1664 36893295. MP_SendRecv 1736 218624. MP_ISendRecv 1736 218624. MP_Wait 9506 MP_ISend 6216 1080958. MP_IRecv 6216 1080958. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.043 322.381 322.382 qs_mol_dyn_low 1 2.0 0.003 0.003 321.757 321.768 qs_forces 11 3.9 0.004 0.004 321.653 321.658 qs_energies 11 4.9 0.001 0.001 311.976 311.986 scf_env_do_scf 11 5.9 0.001 0.001 286.473 286.484 velocity_verlet 10 3.0 0.001 0.001 228.754 228.763 scf_env_do_scf_inner_loop 113 6.6 0.003 0.008 150.563 150.565 init_scf_loop 11 6.9 0.000 0.000 135.643 135.647 prepare_preconditioner 11 7.9 0.000 0.000 130.355 130.384 make_preconditioner 11 8.9 0.000 0.000 130.355 130.384 make_full_inverse_cholesky 11 9.9 0.036 0.039 105.901 127.338 qs_scf_new_mos 113 7.6 0.001 0.001 98.419 98.573 qs_scf_loop_do_ot 113 8.6 0.001 0.001 98.418 98.572 ot_scf_mini 113 9.6 0.003 0.003 93.685 93.780 dbcsr_multiply_generic 2425 12.5 0.215 0.229 86.925 87.923 cp_fm_upper_to_full 103 14.7 53.710 76.994 53.710 76.994 multiply_cannon 2425 13.5 0.687 0.726 60.691 62.027 multiply_cannon_loop 2425 14.5 0.457 0.464 56.849 59.266 ot_mini 113 10.6 0.001 0.001 46.344 46.439 dbcsr_complete_redistribute 391 12.7 4.081 4.143 30.537 43.880 cp_fm_cholesky_invert 11 10.9 40.624 40.632 40.624 40.632 copy_fm_to_dbcsr 207 11.6 0.001 0.001 26.979 40.265 rebuild_ks_matrix 124 8.3 0.001 0.001 39.029 39.190 qs_ks_build_kohn_sham_matrix 124 9.3 0.017 0.017 39.028 39.190 transfer_fm_to_dbcsr 11 9.9 0.028 0.032 24.408 37.511 qs_ks_update_qs_env 124 7.6 0.001 0.001 35.748 35.897 mp_alltoall_i22 704 14.0 22.141 35.463 22.141 35.463 mp_waitall_1 100290 16.8 31.190 34.669 31.190 34.669 qs_ot_get_p 124 10.4 0.001 0.001 31.755 31.860 qs_ot_p2m_diag 81 11.4 0.857 0.863 27.203 27.240 qs_ot_get_derivative 113 11.6 0.002 0.002 26.012 26.117 cp_dbcsr_syevd 81 12.4 0.006 0.006 25.321 25.324 make_m2s 4850 13.5 0.072 0.075 20.997 22.030 multiply_cannon_metrocomm3 9700 15.5 0.022 0.024 20.369 22.016 make_images 4850 14.5 3.650 3.763 20.530 21.567 cp_fm_diag_elpa 81 13.4 0.000 0.000 21.531 21.531 cp_fm_diag_elpa_base 81 14.4 17.060 18.687 21.525 21.526 qs_rho_update_rho_low 124 7.7 0.001 0.001 21.290 21.308 calculate_rho_elec 124 8.7 0.465 0.466 21.290 21.307 sum_up_and_integrate 124 10.3 0.311 0.313 20.422 20.496 ot_diis_step 113 11.6 0.021 0.022 20.266 20.267 integrate_v_rspace 124 11.3 0.004 0.004 20.110 20.185 apply_preconditioner_dbcsr 124 12.6 0.000 0.000 19.406 19.900 apply_single 124 13.6 0.001 0.001 19.406 19.900 multiply_cannon_multrec 9700 15.5 10.184 11.889 17.616 18.824 init_scf_run 11 5.9 0.000 0.001 15.289 15.290 scf_env_initial_rho_setup 11 6.9 0.001 0.002 15.289 15.289 multiply_cannon_sync_h2d 9700 15.5 15.177 15.202 15.177 15.202 make_images_data 4850 15.5 0.055 0.060 11.573 13.655 hybrid_alltoall_any 5034 16.4 1.260 2.943 11.431 13.384 pw_transfer 1499 11.6 0.091 0.092 12.521 12.528 fft_wrap_pw1pw2 1251 12.7 0.011 0.011 12.292 12.301 cp_fm_cholesky_decompose 22 10.9 11.342 11.379 11.342 11.379 fft_wrap_pw1pw2_140 507 13.2 2.960 2.993 10.718 10.731 qs_ot_get_derivative_diag 75 12.3 0.002 0.002 10.374 10.462 mp_alltoall_d11v 2371 14.1 9.443 10.343 9.443 10.343 dbcsr_mm_accdrv_process 20098 16.0 3.669 5.398 7.200 9.949 wfi_extrapolate 11 7.9 0.001 0.001 9.828 9.828 density_rs2pw 124 9.7 0.005 0.005 8.834 8.973 calculate_dm_sparse 124 9.5 0.001 0.001 8.582 8.661 grid_integrate_task_list 124 12.3 8.353 8.530 8.353 8.530 qs_energies_init_hamiltonians 11 5.9 0.008 0.017 7.973 7.974 fft3d_ps 1251 14.7 2.757 2.764 7.542 7.555 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.617 6.692 rs_scatter_matrices 135 9.7 3.588 4.466 6.228 6.455 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=322.382000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2830.090909, yerr=162.310283 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 1.259246E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.048 85.838 85.840 qs_energies 1 2.0 0.000 0.000 85.315 85.325 ls_scf 1 3.0 0.000 0.000 84.428 84.438 dbcsr_multiply_generic 111 6.7 0.015 0.016 73.232 73.417 multiply_cannon 111 7.7 0.017 0.021 56.398 57.465 multiply_cannon_loop 111 8.7 0.208 0.222 52.971 54.616 ls_scf_main 1 4.0 0.000 0.000 52.502 52.514 density_matrix_trs4 2 5.0 0.002 0.003 46.910 46.990 ls_scf_init_scf 1 4.0 0.000 0.000 28.874 28.875 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.706 27.765 mp_waitall_1 11031 10.9 22.655 25.849 22.655 25.849 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.621 25.638 multiply_cannon_multrec 2664 9.7 8.161 8.989 15.563 17.309 multiply_cannon_sync_h2d 2664 9.7 13.486 14.719 13.486 14.719 make_m2s 222 7.7 0.008 0.011 13.084 13.581 make_images 222 8.7 0.099 0.109 13.063 13.562 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.860 12.346 make_images_data 222 9.7 0.004 0.005 7.649 8.219 dbcsr_mm_accdrv_process 4760 10.4 0.514 0.619 7.020 8.019 hybrid_alltoall_any 227 10.6 0.215 1.845 6.539 7.914 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.467 7.628 dbcsr_mm_accdrv_process_sort 4760 11.4 6.307 7.199 6.307 7.199 calculate_norms 4752 9.8 5.506 6.137 5.506 6.137 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.017 5.159 mp_sum_l 887 5.1 3.222 4.714 3.222 4.714 make_images_sizes 222 9.7 0.000 0.000 0.821 3.656 mp_alltoall_i44 222 10.7 0.821 3.656 0.821 3.656 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.422 3.645 arnoldi_extremal 4 6.8 0.000 0.000 3.340 3.367 arnoldi_normal_ev 4 7.8 0.001 0.003 3.340 3.367 build_subspace 16 8.4 0.009 0.012 3.245 3.248 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.047 3.244 mp_irecv_dv 6231 10.9 2.030 3.233 2.030 3.233 ls_scf_post 1 4.0 0.000 0.000 3.052 3.061 ls_scf_store_result 1 5.0 0.000 0.000 2.866 2.912 dbcsr_special_finalize 555 9.7 0.005 0.006 2.265 2.711 dbcsr_merge_single_wm 555 10.7 0.459 0.595 2.257 2.703 make_images_pack 222 9.7 2.204 2.626 2.206 2.628 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.357 2.591 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.370 2.459 dbcsr_matrix_vector_mult_local 304 10.0 2.061 2.448 2.063 2.450 dbcsr_sort_data 658 11.4 2.054 2.436 2.054 2.436 buffer_matrices_ensure_size 222 8.7 1.772 2.058 1.772 2.058 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.831 1.832 rebuild_ks_matrix 3 7.3 0.000 0.000 1.821 1.822 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.821 1.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.840000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 2.102047E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.045 91.249 91.250 qs_energies 1 2.0 0.000 0.000 90.678 90.681 ls_scf 1 3.0 0.000 0.000 89.354 89.357 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.216 75.532 multiply_cannon 111 7.7 0.028 0.040 53.429 56.634 ls_scf_main 1 4.0 0.000 0.000 54.731 54.736 multiply_cannon_loop 111 8.7 0.116 0.122 50.142 53.030 density_matrix_trs4 2 5.0 0.002 0.003 48.942 49.134 ls_scf_init_scf 1 4.0 0.000 0.002 31.083 31.084 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.842 29.964 mp_waitall_1 9105 10.9 21.188 29.801 21.188 29.801 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.235 27.243 multiply_cannon_multrec 1332 9.7 13.045 16.647 22.431 27.134 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.813 20.786 make_m2s 222 7.7 0.006 0.007 15.252 15.879 make_images 222 8.7 1.577 1.969 15.222 15.850 dbcsr_mm_accdrv_process 4041 10.4 0.307 0.652 8.986 10.487 dbcsr_mm_accdrv_process_sort 4041 11.4 8.507 10.057 8.507 10.057 make_images_data 222 9.7 0.004 0.004 8.750 9.631 hybrid_alltoall_any 227 10.6 0.521 2.482 8.203 9.028 mp_sum_l 887 5.1 5.272 8.184 5.272 8.184 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.219 7.818 mp_irecv_dv 3311 11.0 3.199 7.758 3.199 7.758 calculate_norms 2376 9.8 6.006 6.920 6.006 6.920 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.993 6.280 multiply_cannon_sync_h2d 1332 9.7 4.830 5.891 4.830 5.891 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.240 5.499 arnoldi_extremal 4 6.8 0.000 0.000 4.695 4.709 arnoldi_normal_ev 4 7.8 0.001 0.005 4.695 4.709 build_subspace 16 8.4 0.014 0.021 4.445 4.448 ls_scf_post 1 4.0 0.000 0.000 3.539 3.543 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.174 3.414 ls_scf_store_result 1 5.0 0.000 0.000 3.232 3.363 dbcsr_matrix_vector_mult_local 304 10.0 2.747 3.227 2.749 3.229 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.257 2.863 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.664 2.780 make_images_pack 222 9.7 2.019 2.428 2.021 2.430 mp_allgather_i34 111 8.7 0.982 2.232 0.982 2.232 dbcsr_sort_data 436 11.2 1.831 2.057 1.831 2.057 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.876 1.878 rebuild_ks_matrix 3 7.3 0.000 0.000 1.863 1.866 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.863 1.866 dbcsr_data_new 4174 10.1 1.616 1.832 1.616 1.832 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.250000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1726.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 2.757157E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.152 94.104 94.135 qs_energies 1 2.0 0.000 0.000 93.405 93.454 ls_scf 1 3.0 0.000 0.001 91.972 92.020 dbcsr_multiply_generic 111 6.7 0.016 0.019 76.516 76.791 ls_scf_main 1 4.0 0.001 0.018 57.641 57.646 multiply_cannon 111 7.7 0.040 0.099 52.562 56.570 multiply_cannon_loop 111 8.7 0.100 0.115 49.025 53.112 density_matrix_trs4 2 5.0 0.002 0.003 51.597 51.778 mp_waitall_1 7281 11.0 23.524 33.672 23.524 33.672 ls_scf_init_scf 1 4.0 0.000 0.002 30.821 30.823 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.551 29.625 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 27.155 27.168 multiply_cannon_multrec 888 9.7 12.634 15.529 21.267 24.440 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.632 22.648 make_m2s 222 7.7 0.006 0.007 17.198 18.397 make_images 222 8.7 1.972 2.317 17.160 18.358 hybrid_alltoall_any 227 10.6 0.619 2.846 9.433 10.949 make_images_data 222 9.7 0.003 0.004 9.847 10.831 dbcsr_mm_accdrv_process 3754 10.4 0.240 0.417 8.162 9.394 dbcsr_mm_accdrv_process_sort 3754 11.4 7.792 8.977 7.792 8.977 mp_sum_l 887 5.1 5.249 8.648 5.249 8.648 multiply_cannon_sync_h2d 888 9.7 6.042 7.275 6.042 7.275 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.426 7.044 mp_irecv_dv 2335 11.1 2.412 7.001 2.412 7.001 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.981 6.795 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.667 6.611 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.919 5.165 arnoldi_extremal 4 6.8 0.000 0.000 5.127 5.142 arnoldi_normal_ev 4 7.8 0.001 0.005 5.127 5.142 build_subspace 16 8.4 0.014 0.020 4.821 4.830 calculate_norms 1584 9.8 4.257 4.604 4.257 4.604 mp_allgather_i34 111 8.7 1.369 3.797 1.369 3.797 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.444 3.793 dbcsr_matrix_vector_mult_local 304 10.0 3.002 3.579 3.004 3.580 ls_scf_post 1 4.0 0.002 0.015 3.509 3.558 ls_scf_store_result 1 5.0 0.000 0.000 3.249 3.331 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.909 3.011 make_images_sizes 222 9.7 0.000 0.000 1.030 2.335 mp_alltoall_i44 222 10.7 1.030 2.335 1.030 2.335 dbcsr_sort_data 325 11.1 1.887 2.187 1.887 2.187 make_images_pack 222 9.7 1.812 2.108 1.815 2.111 dbcsr_data_release 9322 10.9 1.301 2.013 1.301 2.013 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.933 1.935 rebuild_ks_matrix 3 7.3 0.000 0.000 1.915 1.917 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.915 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.135000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2181.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 3.325301E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.047 97.562 97.562 qs_energies 1 2.0 0.000 0.000 96.958 96.963 ls_scf 1 3.0 0.000 0.000 95.319 95.326 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.885 79.140 ls_scf_main 1 4.0 0.000 0.000 59.193 59.196 multiply_cannon 111 7.7 0.044 0.080 51.929 55.750 density_matrix_trs4 2 5.0 0.002 0.003 52.976 53.124 multiply_cannon_loop 111 8.7 0.114 0.124 46.940 49.713 ls_scf_init_scf 1 4.0 0.000 0.000 32.864 32.866 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.609 31.681 mp_waitall_1 6369 11.0 23.299 30.539 23.299 30.539 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.120 29.130 multiply_cannon_multrec 1332 9.7 14.150 17.272 22.097 24.971 make_m2s 222 7.7 0.007 0.007 21.380 22.714 make_images 222 8.7 3.135 3.592 21.329 22.665 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.567 16.563 make_images_data 222 9.7 0.004 0.004 11.984 13.629 hybrid_alltoall_any 227 10.6 0.796 3.751 11.402 13.150 dbcsr_mm_accdrv_process 3641 10.4 0.220 0.411 7.591 9.120 dbcsr_mm_accdrv_process_sort 3641 11.4 7.218 8.704 7.218 8.704 mp_sum_l 887 5.1 4.005 7.321 4.005 7.321 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.079 6.228 mp_irecv_dv 3229 10.9 2.058 6.152 2.058 6.152 multiply_cannon_sync_h2d 1332 9.7 5.493 6.152 5.493 6.152 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.083 5.727 arnoldi_extremal 4 6.8 0.000 0.000 5.322 5.335 arnoldi_normal_ev 4 7.8 0.001 0.005 5.322 5.335 build_subspace 16 8.4 0.014 0.021 4.979 4.986 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.576 4.906 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.577 4.787 calculate_norms 2376 9.8 4.185 4.530 4.185 4.530 mp_allgather_i34 111 8.7 2.097 4.220 2.097 4.220 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.577 3.867 dbcsr_matrix_vector_mult_local 304 10.0 3.160 3.653 3.162 3.654 dbcsr_sort_data 658 11.4 3.062 3.341 3.062 3.341 ls_scf_post 1 4.0 0.000 0.000 3.262 3.267 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.083 3.150 dbcsr_special_finalize 555 9.7 0.006 0.007 2.808 3.117 dbcsr_merge_single_wm 555 10.7 0.530 0.654 2.799 3.109 ls_scf_store_result 1 5.0 0.000 0.000 3.012 3.064 dbcsr_data_release 10477 10.7 1.577 2.403 1.577 2.403 dbcsr_finalize 304 7.8 0.049 0.061 1.798 1.984 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.960 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.562000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 4.675863E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.067 99.650 99.651 qs_energies 1 2.0 0.000 0.000 98.832 98.837 ls_scf 1 3.0 0.000 0.000 96.903 96.914 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.452 78.690 ls_scf_main 1 4.0 0.000 0.000 62.466 62.468 multiply_cannon 111 7.7 0.090 0.157 55.520 60.412 density_matrix_trs4 2 5.0 0.002 0.003 55.252 55.326 multiply_cannon_loop 111 8.7 0.069 0.076 51.097 53.203 mp_waitall_1 5436 11.0 26.924 31.648 26.924 31.648 ls_scf_init_scf 1 4.0 0.000 0.000 30.803 30.809 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.547 29.579 matrix_sqrt_Newton_Schulz 2 6.5 0.003 0.013 27.376 27.386 multiply_cannon_multrec 444 9.7 14.125 16.515 21.246 23.717 make_m2s 222 7.7 0.004 0.005 17.902 20.319 make_images 222 8.7 3.723 4.403 17.841 20.259 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.383 16.437 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.105 14.344 make_images_data 222 9.7 0.003 0.004 10.235 12.501 hybrid_alltoall_any 227 10.6 0.788 3.768 9.968 12.373 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.345 6.826 7.973 multiply_cannon_sync_h2d 444 9.7 6.515 7.946 6.515 7.946 dbcsr_mm_accdrv_process_sort 3003 11.4 6.511 7.626 6.511 7.626 mp_allgather_i34 111 8.7 2.658 6.979 2.658 6.979 arnoldi_extremal 4 6.8 0.000 0.000 5.790 5.804 arnoldi_normal_ev 4 7.8 0.002 0.005 5.790 5.804 build_subspace 16 8.4 0.015 0.020 5.395 5.407 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.611 4.747 mp_sum_l 887 5.1 3.076 4.743 3.076 4.743 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.125 4.324 dbcsr_matrix_vector_mult_local 304 10.0 3.635 4.108 3.638 4.110 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.660 4.055 mp_irecv_dv 1241 11.2 1.642 4.025 1.642 4.025 calculate_norms 792 9.8 3.554 3.727 3.554 3.727 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.588 3.672 ls_scf_post 1 4.0 0.000 0.000 3.634 3.638 make_images_sizes 222 9.7 0.000 0.000 1.018 3.617 mp_alltoall_i44 222 10.7 1.018 3.617 1.018 3.617 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.093 3.535 ls_scf_store_result 1 5.0 0.000 0.000 3.423 3.463 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.274 dbcsr_merge_all 275 8.9 0.473 0.523 2.049 2.117 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.070 2.071 rebuild_ks_matrix 3 7.3 0.000 0.000 2.037 2.038 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.037 2.038 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.651000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3732.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_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 8.676606E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.096 0.118 108.158 108.159 qs_energies 1 2.0 0.000 0.000 106.692 106.705 ls_scf 1 3.0 0.000 0.000 103.739 103.752 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.364 77.482 ls_scf_main 1 4.0 0.000 0.000 65.991 65.992 density_matrix_trs4 2 5.0 0.002 0.003 56.710 56.773 multiply_cannon 111 7.7 0.149 0.207 49.781 51.597 multiply_cannon_loop 111 8.7 0.068 0.070 46.304 47.277 ls_scf_init_scf 1 4.0 0.000 0.000 34.056 34.057 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.489 32.502 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.664 29.675 mp_waitall_1 4527 11.1 22.077 25.905 22.077 25.905 make_m2s 222 7.7 0.005 0.005 23.882 24.904 make_images 222 8.7 4.587 4.982 23.776 24.796 multiply_cannon_multrec 444 9.7 17.840 18.667 22.491 23.146 hybrid_alltoall_any 227 10.6 1.660 3.608 12.933 15.655 make_images_data 222 9.7 0.003 0.003 13.126 15.596 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.378 10.822 multiply_cannon_sync_h2d 444 9.7 8.845 8.886 8.845 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.284 7.300 arnoldi_normal_ev 4 7.8 0.003 0.009 7.284 7.300 build_subspace 16 8.4 0.026 0.036 6.734 6.744 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.348 5.492 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.234 5.323 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.029 5.278 dbcsr_matrix_vector_mult_local 304 10.0 4.903 5.183 4.906 5.185 dbcsr_mm_accdrv_process 1814 10.4 0.253 0.316 4.488 4.618 dbcsr_mm_accdrv_process_sort 1814 11.4 4.189 4.323 4.189 4.323 ls_scf_post 1 4.0 0.000 0.000 3.692 3.704 mp_allgather_i34 111 8.7 1.130 3.555 1.130 3.555 make_images_sizes 222 9.7 0.000 0.000 1.415 3.487 mp_alltoall_i44 222 10.7 1.414 3.486 1.414 3.486 ls_scf_store_result 1 5.0 0.000 0.000 3.409 3.420 calculate_norms 792 9.8 3.243 3.276 3.243 3.276 dbcsr_finalize 304 7.8 0.082 0.089 3.090 3.168 dbcsr_merge_all 275 8.9 0.890 0.919 2.875 2.948 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.922 2.922 dbcsr_complete_redistribute 5 7.6 1.443 1.478 2.778 2.896 dbcsr_data_release 12724 10.6 2.337 2.845 2.337 2.845 matrix_ls_to_qs 2 6.0 0.000 0.000 2.441 2.576 dbcsr_sort_data 325 11.1 2.442 2.502 2.442 2.502 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.470 2.472 rebuild_ks_matrix 3 7.3 0.000 0.000 2.404 2.406 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.404 2.406 dbcsr_new_transposed 4 7.5 0.243 0.252 2.262 2.274 dbcsr_frobenius_norm 74 6.6 2.058 2.140 2.194 2.233 dbcsr_add_d 103 6.2 0.000 0.000 2.131 2.204 dbcsr_add_anytype 103 7.2 0.859 0.892 2.131 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.159000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6844.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/e790daffb71be1cee1ff37897efe550b44207090_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 589.598720E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 812892694. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.092 260.889 260.893 qs_mol_dyn_low 1 2.0 0.004 0.029 259.741 259.772 qs_forces 5 3.8 0.005 0.021 259.588 259.593 qs_energies 5 4.8 0.002 0.036 256.192 256.231 scf_env_do_scf 5 5.8 0.001 0.021 238.080 238.085 scf_env_do_scf_inner_loop 105 6.6 0.003 0.032 205.544 205.574 qs_scf_new_mos 105 7.6 0.001 0.001 157.336 157.603 qs_scf_loop_do_ot 105 8.6 0.001 0.001 157.336 157.603 ot_scf_mini 105 9.6 0.003 0.006 147.536 147.725 dbcsr_multiply_generic 1445 12.2 0.143 0.152 136.809 137.577 multiply_cannon 1445 13.2 0.274 0.289 116.513 118.580 multiply_cannon_loop 1445 14.2 2.393 2.513 114.148 116.361 velocity_verlet 4 3.0 0.001 0.010 115.202 115.205 ot_mini 105 10.6 0.003 0.048 61.150 61.282 qs_ot_get_p 112 10.4 0.001 0.001 56.351 56.661 multiply_cannon_multrec 69360 15.2 31.657 36.927 39.498 44.838 qs_ot_p2m_diag 40 11.0 0.020 0.031 44.722 44.825 mp_waitall_1 488190 16.1 36.213 44.672 36.213 44.672 cp_dbcsr_syevd 40 12.0 0.002 0.002 41.366 41.368 qs_ot_get_derivative 55 11.6 0.001 0.017 39.361 39.494 cp_fm_syevd 40 13.0 0.000 0.001 35.360 35.516 multiply_cannon_metrocomm3 69360 15.2 0.192 0.205 24.983 35.204 multiply_cannon_sync_h2d 69360 15.2 30.550 35.087 30.550 35.087 rebuild_ks_matrix 110 8.4 0.000 0.000 32.346 32.529 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.023 32.346 32.529 init_scf_loop 7 6.6 0.000 0.001 32.483 32.487 cp_fm_redistribute_end 40 14.0 15.117 30.147 15.124 30.151 cp_fm_syevd_base 40 14.0 15.011 30.039 15.011 30.039 qs_ks_update_qs_env 112 7.6 0.001 0.001 29.640 29.809 prepare_preconditioner 7 7.6 0.000 0.000 27.014 27.041 make_preconditioner 7 8.6 0.000 0.000 27.014 27.041 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.053 23.294 apply_single 62 13.6 0.000 0.000 23.053 23.294 qs_rho_update_rho_low 110 7.6 0.001 0.001 22.682 23.014 calculate_rho_elec 110 8.6 0.030 0.033 22.681 23.014 ot_new_cg_direction 55 11.6 0.003 0.057 21.087 21.090 rs_pw_transfer 690 11.5 0.010 0.013 18.075 19.428 make_full_inverse_cholesky 7 9.6 0.000 0.000 18.897 18.958 density_rs2pw 110 9.6 0.006 0.007 16.306 17.602 init_scf_run 5 5.8 0.000 0.001 15.254 15.255 scf_env_initial_rho_setup 5 6.8 0.002 0.003 15.254 15.255 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.762 14.999 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.209 14.325 pw_transfer 1645 12.4 0.085 0.115 13.202 13.466 fft_wrap_pw1pw2 1425 13.5 0.013 0.018 13.058 13.325 cp_fm_cholesky_invert 7 10.6 13.110 13.125 13.110 13.125 mp_sum_l 4764 12.2 11.985 12.744 11.985 12.744 calculate_dm_sparse 110 9.5 0.000 0.001 11.393 11.538 sum_up_and_integrate 60 10.3 0.028 0.042 11.313 11.331 integrate_v_rspace 60 11.3 0.001 0.003 11.285 11.303 fft_wrap_pw1pw2_240 915 15.0 1.213 1.304 10.921 11.150 qs_vxc_create 110 10.4 0.002 0.004 10.909 10.983 check_diag 80 13.5 8.588 8.858 10.544 10.693 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.555 10.638 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 10.224 10.240 calculate_first_density_matrix 1 7.0 0.000 0.003 9.942 9.954 acc_transpose_blocks 69360 15.2 0.365 0.381 9.396 9.920 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.673 9.726 multiply_cannon_metrocomm1 69360 15.2 0.091 0.097 4.470 9.290 fft3d_pb 915 16.0 2.391 2.716 8.956 9.271 make_m2s 2890 13.2 0.079 0.087 7.841 8.508 dbcsr_mm_accdrv_process 154766 15.8 4.077 4.255 7.710 8.506 make_images 2890 14.2 0.239 0.259 7.733 8.401 acc_transpose_blocks_kernels 69360 16.2 0.855 0.903 7.456 7.894 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 7.554 7.804 make_full_single_inverse 7 9.6 0.001 0.001 7.742 7.771 xc_vxc_pw_create 60 11.3 0.039 0.049 7.363 7.437 potential_pw2rs 60 12.3 0.003 0.003 7.197 7.253 jit_kernel_transpose 5 15.0 6.601 7.027 6.601 7.027 xc_pw_derive 510 13.4 0.006 0.007 6.590 6.659 rs_pw_transfer_RS2PW_30 110 11.6 1.599 1.685 6.257 6.575 mp_alltoall_z22v 2340 17.7 6.044 6.482 6.044 6.482 mp_sendrecv_dv 168740 12.6 5.858 6.009 5.858 6.009 mp_waitany 7680 13.5 4.584 5.830 4.584 5.830 mp_alltoall_d11v 1300 13.8 4.335 5.250 4.335 5.250 cp_fm_cholesky_decompose 14 10.2 5.214 5.221 5.214 5.221 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=260.893000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: e790daffb71be1cee1ff37897efe550b44207090 Summary: empty Status: OK