=== 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: b314dfac77ca549718eae9848a54a5a4c852e9f1 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # 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 (25.05.2023) # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/01 job id: 46851284 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/02 job id: 46851285 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/03 job id: 46851286 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/04 job id: 46851287 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/05 job id: 46851288 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/06 job id: 46851289 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/07 job id: 46851290 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/08 job id: 46851291 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/09 job id: 46851292 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/10 job id: 46851293 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/11 job id: 46851294 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/12 job id: 46851295 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/13 job id: 46851296 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/14 job id: 46851297 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/15 job id: 46851298 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/16 job id: 46851299 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/17 job id: 46851300 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/18 job id: 46851301 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/19 job id: 46851302 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/20 job id: 46851303 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/21 job id: 46851305 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/22 job id: 46851306 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/23 job id: 46851307 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/24 job id: 46851308 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/25 job id: 46851309 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/26 job id: 46851310 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/27 job id: 46851311 --- 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/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 0.021 0.036 135.576 135.577 farming_run 1 2.0 135.028 135.038 135.549 135.552 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.450394E+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 0.006 0.024 116.341 116.342 qs_energies 1 2.0 0.000 0.000 116.116 116.119 mp2_main 1 3.0 0.000 0.000 113.767 113.771 mp2_gpw_main 1 4.0 0.019 0.026 112.723 112.726 mp2_ri_gpw_compute_in 1 5.0 0.172 0.178 93.740 94.184 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.492 55.936 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.170 41.809 47.177 get_2c_integrals 1 6.0 0.008 0.009 37.337 38.070 integrate_v_rspace 273 8.0 0.437 0.450 25.187 30.362 pw_transfer 6555 10.6 0.375 0.385 27.442 28.041 grid_integrate_task_list 273 9.0 21.000 26.660 21.000 26.660 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.094 26.507 fft_wrap_pw1pw2_100 2178 12.4 1.188 1.430 23.563 24.002 compute_2c_integrals 1 7.0 0.002 0.004 19.824 19.825 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.860 19.524 mp2_eri_2c_integrate_gpw 1 9.0 2.375 2.429 18.857 19.523 rpa_ri_compute_en 1 5.0 0.020 0.023 18.874 19.093 cp_fm_cholesky_decompose 12 8.2 17.442 18.161 17.442 18.161 cholesky_decomp 1 7.0 0.000 0.000 16.362 17.078 fft3d_s 5443 13.4 16.154 16.519 16.177 16.543 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.591 10.856 15.591 calculate_wavefunction 272 8.0 5.417 5.581 12.541 13.232 rpa_num_int 1 6.0 0.000 0.001 10.601 10.602 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.583 10.591 calc_mat_Q 8 8.0 0.000 0.000 9.460 9.546 contract_S_to_Q 8 9.0 0.000 0.000 8.882 8.972 calc_potential_gpw 544 9.5 0.005 0.006 8.266 8.698 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.204 8.600 parallel_gemm_fm 14 9.1 0.000 0.000 8.471 8.568 parallel_gemm_fm_cosma 14 10.1 8.471 8.568 8.471 8.568 potential_pw2rs 545 10.0 0.108 0.111 7.666 8.337 collocate_single_gaussian 272 10.0 0.040 0.042 7.460 7.828 create_integ_mat 1 6.0 0.021 0.028 7.734 7.735 array2fm 1 7.0 0.000 0.000 6.748 7.144 pw_scatter_s 2720 13.7 4.431 4.557 4.431 4.557 pw_gather_s 2722 13.2 3.856 4.180 3.856 4.180 array2fm_buffer_send 1 8.0 3.005 3.163 3.005 3.163 pw_poisson_solve 545 10.5 1.098 1.135 2.152 2.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.726242, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2793.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 0.029 0.038 396.136 396.137 farming_run 1 2.0 395.466 395.471 396.101 396.104 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.224761E+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 11627462. 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 0.009 0.029 209.475 209.476 qs_energies 1 2.0 0.000 0.000 209.240 209.253 scf_env_do_scf 1 3.0 0.000 0.000 106.173 106.173 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.307 105.316 rebuild_ks_matrix 4 6.0 0.000 0.000 105.306 105.315 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.066 105.306 105.315 hfx_ks_matrix 4 8.0 0.001 0.001 104.927 104.932 integrate_four_center 4 9.0 0.144 0.457 104.927 104.931 mp2_main 1 3.0 0.000 0.000 102.779 102.792 mp2_gpw_main 1 4.0 0.031 0.046 101.942 101.959 integrate_four_center_main 4 10.0 0.057 0.449 96.728 99.215 integrate_four_center_bin 265 11.0 96.671 99.181 96.671 99.181 init_scf_loop 1 4.0 0.000 0.000 91.983 91.983 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.879 75.891 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.500 55.512 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.161 42.192 47.207 integrate_v_rspace 95 8.0 0.399 0.568 28.572 33.391 pw_transfer 2240 10.6 0.146 0.161 29.873 30.253 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.897 29.278 grid_integrate_task_list 95 9.0 23.865 28.921 23.865 28.921 mp2_ri_gpw_compute_en 1 5.0 0.066 0.081 26.918 28.543 ao_to_mo_and_store_B_mult_1 91 7.0 10.629 28.002 10.629 28.002 fft_wrap_pw1pw2_100 730 12.4 1.248 1.388 26.603 26.953 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.881 25.252 25.263 get_2c_integrals 1 6.0 0.000 0.000 20.295 20.315 compute_2c_integrals 1 7.0 0.002 0.003 19.271 19.274 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.772 19.150 mp2_eri_2c_integrate_gpw 1 9.0 1.735 1.815 18.771 19.149 fft3d_s 1823 13.4 18.426 18.789 18.440 18.802 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.188 14.188 calculate_wavefunction 91 8.0 2.015 2.050 9.732 9.938 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.591 8.841 9.359 potential_pw2rs 186 10.0 0.033 0.035 8.599 9.148 local_gemm 172 8.0 8.283 8.816 8.283 8.816 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.519 8.186 8.709 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.197 8.528 calc_potential_gpw 182 9.5 0.002 0.002 7.901 8.191 collocate_single_gaussian 91 10.0 0.017 0.020 7.843 8.052 mp_sendrecv_dm3 2068 8.0 6.219 6.722 6.219 6.722 mp2_ri_gpw_compute_en_ener 172 7.0 6.348 6.413 6.348 6.413 mp_sync 37 10.5 2.894 5.882 2.894 5.882 pw_gather_s 912 13.2 4.908 5.406 4.908 5.406 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.929036, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1512.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 447.959040E+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 0.041 0.047 48.659 48.660 qs_mol_dyn_low 1 2.0 0.003 0.005 48.254 48.263 qs_forces 11 3.9 0.009 0.059 48.184 48.185 qs_energies 11 4.9 0.004 0.009 46.712 46.723 scf_env_do_scf 11 5.9 0.001 0.001 40.806 40.806 scf_env_do_scf_inner_loop 108 6.5 0.003 0.016 38.795 38.796 dbcsr_multiply_generic 2286 12.5 0.095 0.100 29.144 29.538 qs_scf_new_mos 108 7.5 0.000 0.001 28.809 29.071 qs_scf_loop_do_ot 108 8.5 0.000 0.001 28.808 29.071 ot_scf_mini 108 9.5 0.002 0.002 27.326 27.518 multiply_cannon 2286 13.5 0.181 0.194 24.070 25.315 multiply_cannon_loop 2286 14.5 1.437 1.547 23.412 24.693 velocity_verlet 10 3.0 0.001 0.002 23.372 23.372 ot_mini 108 10.5 0.001 0.001 15.949 16.198 qs_ot_get_derivative 108 11.5 0.001 0.001 13.191 13.371 mp_waitall_1 245248 16.5 6.379 12.587 6.379 12.587 multiply_cannon_metrocomm3 54864 15.5 0.069 0.075 4.731 11.583 multiply_cannon_multrec 54864 15.5 4.484 6.843 7.611 11.047 rebuild_ks_matrix 119 8.3 0.000 0.000 7.830 7.955 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.830 7.955 qs_ot_get_p 119 10.4 0.001 0.001 7.193 7.452 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.894 7.010 multiply_cannon_sync_h2d 54864 15.5 6.253 6.911 6.253 6.911 mp_sum_l 7287 12.8 3.209 5.037 3.209 5.037 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.577 4.999 qs_ot_p2m_diag 50 11.0 0.004 0.006 4.908 4.939 init_scf_run 11 5.9 0.000 0.001 4.553 4.553 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.552 4.553 sum_up_and_integrate 119 10.3 0.012 0.014 4.513 4.520 integrate_v_rspace 119 11.3 0.002 0.005 4.501 4.509 dbcsr_mm_accdrv_process 76910 16.1 1.054 1.822 3.049 4.422 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.189 4.285 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.280 4.280 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.123 4.191 calculate_rho_elec 119 8.7 0.012 0.017 4.123 4.190 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.094 4.094 cp_fm_redistribute_end 50 14.0 2.090 4.071 2.096 4.073 cp_fm_diag_elpa_base 50 14.0 1.972 3.969 1.976 3.978 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.852 3.041 apply_single 119 13.6 0.000 0.000 2.852 3.041 rs_pw_transfer 974 11.9 0.012 0.014 2.709 2.833 calculate_dm_sparse 119 9.5 0.000 0.001 2.686 2.797 ot_diis_step 108 11.5 0.006 0.006 2.576 2.576 jit_kernel_multiply 13 15.8 1.939 2.516 1.939 2.516 calculate_first_density_matrix 1 7.0 0.000 0.001 2.470 2.475 density_rs2pw 119 9.7 0.004 0.005 2.143 2.257 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.185 2.186 acc_transpose_blocks 54864 15.5 0.210 0.247 1.624 2.152 grid_integrate_task_list 119 12.3 2.046 2.133 2.046 2.133 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.980 2.028 wfi_extrapolate 11 7.9 0.001 0.001 2.019 2.019 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.970 2.018 init_scf_loop 11 6.9 0.000 0.000 1.994 1.995 potential_pw2rs 119 12.3 0.004 0.004 1.813 1.823 pw_transfer 1439 11.6 0.052 0.059 1.687 1.756 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.028 1.734 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.611 1.683 mp_sum_d 4135 12.0 1.081 1.661 1.081 1.661 make_m2s 4572 13.5 0.053 0.055 1.555 1.607 make_images 4572 14.5 0.133 0.139 1.473 1.525 mp_waitany 12084 13.8 1.300 1.490 1.300 1.490 grid_collocate_task_list 119 9.7 1.383 1.454 1.383 1.454 mp_alltoall_d11v 2130 13.8 1.187 1.386 1.187 1.386 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.358 1.373 fft3d_ps 1201 14.6 0.372 0.477 1.254 1.319 fft_wrap_pw1pw2_140 487 13.2 0.184 0.202 1.244 1.316 acc_transpose_blocks_kernels 54864 16.5 0.230 0.380 0.736 1.010 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.990 1.008 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.588 0.986 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=48.660000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=427.272727, yerr=0.749656 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 483.676160E+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 1702217. 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 0.010 0.029 38.486 38.486 qs_mol_dyn_low 1 2.0 0.003 0.003 38.206 38.219 qs_forces 11 3.9 0.002 0.003 38.132 38.132 qs_energies 11 4.9 0.001 0.002 36.438 36.441 scf_env_do_scf 11 5.9 0.000 0.001 31.127 31.128 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.711 28.711 dbcsr_multiply_generic 2286 12.5 0.110 0.113 21.083 21.451 qs_scf_new_mos 108 7.5 0.001 0.001 19.623 19.869 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.622 19.868 ot_scf_mini 108 9.5 0.002 0.003 18.751 18.926 velocity_verlet 10 3.0 0.001 0.001 17.999 18.000 multiply_cannon 2286 13.5 0.227 0.237 16.235 17.833 multiply_cannon_loop 2286 14.5 0.897 0.968 15.119 16.472 ot_mini 108 10.5 0.001 0.001 11.541 11.780 mp_waitall_1 200699 16.5 5.529 10.804 5.529 10.804 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.119 9.462 qs_ot_get_derivative 108 11.5 0.001 0.001 9.100 9.279 multiply_cannon_multrec 27432 15.5 1.964 4.489 5.846 8.641 rebuild_ks_matrix 119 8.3 0.000 0.000 7.136 7.281 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.136 7.281 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.286 6.417 dbcsr_mm_accdrv_process 47894 16.0 3.099 5.023 3.814 5.654 qs_ot_get_p 119 10.4 0.001 0.001 4.334 4.563 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.509 4.353 sum_up_and_integrate 119 10.3 0.025 0.027 4.156 4.165 integrate_v_rspace 119 11.3 0.002 0.003 4.132 4.142 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.031 4.104 apply_single 119 13.6 0.000 0.000 3.031 4.104 init_scf_run 11 5.9 0.000 0.001 4.101 4.102 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.101 4.101 mp_sum_l 7287 12.8 1.970 3.915 1.970 3.915 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.760 3.795 calculate_rho_elec 119 8.7 0.021 0.024 3.759 3.795 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.932 2.950 rs_pw_transfer 974 11.9 0.010 0.011 2.454 2.899 multiply_cannon_sync_h2d 27432 15.5 2.192 2.785 2.192 2.785 make_m2s 4572 13.5 0.052 0.054 2.415 2.648 calculate_first_density_matrix 1 7.0 0.000 0.000 2.606 2.608 make_images 4572 14.5 0.200 0.237 2.326 2.556 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.507 2.508 density_rs2pw 119 9.7 0.004 0.004 2.033 2.495 init_scf_loop 11 6.9 0.000 0.000 2.397 2.398 ot_diis_step 108 11.5 0.011 0.011 2.394 2.395 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.200 2.201 cp_fm_redistribute_end 50 14.0 1.116 2.177 1.119 2.179 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.077 2.167 cp_fm_diag_elpa_base 50 14.0 1.029 2.087 1.057 2.129 calculate_dm_sparse 119 9.5 0.000 0.001 2.006 2.084 grid_integrate_task_list 119 12.3 1.876 1.977 1.876 1.977 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.949 1.952 pw_transfer 1439 11.6 0.066 0.070 1.881 1.910 potential_pw2rs 119 12.3 0.006 0.006 1.834 1.842 fft_wrap_pw1pw2 1201 12.6 0.008 0.009 1.788 1.820 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.761 1.801 jit_kernel_multiply 8 16.2 0.663 1.560 0.663 1.560 prepare_preconditioner 11 7.9 0.000 0.000 1.502 1.529 make_preconditioner 11 8.9 0.000 0.000 1.502 1.529 make_images_data 4572 15.5 0.045 0.051 1.128 1.524 acc_transpose_blocks 27432 15.5 0.107 0.112 1.189 1.482 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.405 1.462 fft_wrap_pw1pw2_140 487 13.2 0.201 0.211 1.423 1.454 wfi_extrapolate 11 7.9 0.001 0.001 1.445 1.445 grid_collocate_task_list 119 9.7 1.282 1.438 1.282 1.438 hybrid_alltoall_any 4725 16.4 0.051 0.112 0.978 1.393 fft3d_ps 1201 14.6 0.521 0.579 1.339 1.365 mp_alltoall_d11v 2130 13.8 1.159 1.326 1.159 1.326 mp_allgather_i34 2286 14.5 0.538 1.301 0.538 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.247 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.158 1.204 mp_sum_d 4135 12.0 0.552 1.005 0.552 1.005 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.532 0.977 mp_waitany 5720 13.7 0.513 0.961 0.513 0.961 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.929 0.930 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.907 0.921 acc_transpose_blocks_kernels 27432 16.5 0.181 0.270 0.658 0.865 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.486000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=460.181818, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 519.561216E+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 2521162. 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 0.036 0.056 32.438 32.439 qs_mol_dyn_low 1 2.0 0.003 0.003 32.108 32.118 qs_forces 11 3.9 0.003 0.011 32.041 32.042 qs_energies 11 4.9 0.006 0.025 30.456 30.459 scf_env_do_scf 11 5.9 0.014 0.025 25.786 25.791 scf_env_do_scf_inner_loop 108 6.5 0.006 0.018 22.938 22.940 dbcsr_multiply_generic 2286 12.5 0.096 0.098 16.187 16.296 velocity_verlet 10 3.0 0.002 0.006 15.133 15.135 qs_scf_new_mos 108 7.5 0.001 0.001 14.697 14.708 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.697 14.707 ot_scf_mini 108 9.5 0.002 0.003 13.984 13.992 multiply_cannon 2286 13.5 0.195 0.205 12.917 13.709 multiply_cannon_loop 2286 14.5 0.633 0.661 12.161 12.959 ot_mini 108 10.5 0.001 0.001 8.589 8.598 qs_ot_get_derivative 108 11.5 0.001 0.001 7.103 7.113 multiply_cannon_multrec 18288 15.5 1.932 2.926 6.738 7.053 rebuild_ks_matrix 119 8.3 0.000 0.000 6.436 6.449 qs_ks_build_kohn_sham_matrix 119 9.3 0.020 0.023 6.435 6.448 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.672 5.684 dbcsr_mm_accdrv_process 38222 16.0 4.143 5.401 4.722 5.526 sum_up_and_integrate 119 10.3 0.030 0.031 4.020 4.027 integrate_v_rspace 119 11.3 0.003 0.004 3.990 3.999 mp_waitall_1 158411 16.6 2.485 3.651 2.485 3.651 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.481 3.491 calculate_rho_elec 119 8.7 0.031 0.031 3.481 3.490 init_scf_run 11 5.9 0.000 0.001 3.448 3.449 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.448 3.448 qs_ot_get_p 119 10.4 0.002 0.015 3.289 3.310 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.653 3.282 init_scf_loop 11 6.9 0.019 0.090 2.818 2.835 rs_pw_transfer 974 11.9 0.010 0.011 2.272 2.533 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.400 2.392 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.035 2.362 apply_single 119 13.6 0.000 0.000 2.035 2.362 calculate_first_density_matrix 1 7.0 0.001 0.002 2.228 2.230 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.210 2.215 density_rs2pw 119 9.7 0.004 0.004 1.951 2.211 grid_integrate_task_list 119 12.3 1.850 1.956 1.850 1.956 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.953 1.954 pw_transfer 1439 11.6 0.066 0.070 1.889 1.897 make_m2s 4572 13.5 0.044 0.045 1.716 1.852 calculate_dm_sparse 119 9.5 0.000 0.001 1.799 1.808 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.795 1.806 prepare_preconditioner 11 7.9 0.005 0.037 1.779 1.783 make_preconditioner 11 8.9 0.010 0.055 1.774 1.783 make_images 4572 14.5 0.190 0.202 1.631 1.764 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.620 1.720 potential_pw2rs 119 12.3 0.007 0.008 1.714 1.718 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.700 1.716 cp_fm_diag_elpa_base 50 14.0 1.677 1.692 1.698 1.715 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.660 1.662 multiply_cannon_sync_h2d 18288 15.5 1.405 1.581 1.405 1.581 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.543 1.553 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.516 1.523 mp_sum_l 7287 12.8 1.168 1.519 1.168 1.519 jit_kernel_multiply 8 15.9 0.528 1.516 0.528 1.516 ot_diis_step 108 11.5 0.011 0.011 1.467 1.467 fft_wrap_pw1pw2_140 487 13.2 0.253 0.259 1.443 1.454 grid_collocate_task_list 119 9.7 1.263 1.423 1.263 1.423 fft3d_ps 1201 14.6 0.529 0.547 1.283 1.296 acc_transpose_blocks 18288 15.5 0.074 0.076 1.243 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.177 1.181 wfi_extrapolate 11 7.9 0.001 0.001 1.162 1.162 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.964 0.965 make_images_data 4572 15.5 0.045 0.049 0.769 0.925 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.857 0.882 mp_waitany 9880 13.7 0.551 0.842 0.551 0.842 hybrid_alltoall_any 4725 16.4 0.056 0.114 0.657 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.834 0.836 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.804 0.815 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.122 0.538 0.802 mp_alltoall_d11v 2130 13.8 0.657 0.800 0.657 0.800 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.653 0.723 cp_fm_cholesky_invert 11 10.9 0.698 0.701 0.698 0.701 mp_alltoall_z22v 1201 16.6 0.614 0.690 0.614 0.690 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.439000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=492.727273, yerr=2.377581 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 541.110272E+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 0.035 0.056 35.233 35.235 qs_mol_dyn_low 1 2.0 0.003 0.003 34.890 34.898 qs_forces 11 3.9 0.004 0.019 34.819 34.820 qs_energies 11 4.9 0.004 0.009 33.058 33.063 scf_env_do_scf 11 5.9 0.001 0.003 27.970 27.972 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.408 24.409 dbcsr_multiply_generic 2286 12.5 0.103 0.113 17.964 18.088 velocity_verlet 10 3.0 0.001 0.002 17.617 17.620 qs_scf_new_mos 108 7.5 0.001 0.001 16.012 16.070 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.011 16.069 ot_scf_mini 108 9.5 0.005 0.024 15.081 15.131 multiply_cannon 2286 13.5 0.230 0.269 14.389 14.876 multiply_cannon_loop 2286 14.5 0.937 0.969 13.478 13.912 ot_mini 108 10.5 0.001 0.001 9.195 9.265 multiply_cannon_multrec 27432 15.5 2.338 3.008 8.595 8.989 qs_ot_get_derivative 108 11.5 0.001 0.001 7.393 7.446 dbcsr_mm_accdrv_process 47916 15.9 5.316 6.952 6.165 7.396 rebuild_ks_matrix 119 8.3 0.000 0.000 6.614 6.691 qs_ks_build_kohn_sham_matrix 119 9.3 0.021 0.080 6.614 6.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.853 5.921 sum_up_and_integrate 119 10.3 0.035 0.038 3.849 3.860 integrate_v_rspace 119 11.3 0.003 0.005 3.814 3.826 init_scf_run 11 5.9 0.000 0.001 3.732 3.732 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.731 3.732 init_scf_loop 11 6.9 0.045 0.127 3.542 3.544 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.508 3.539 calculate_rho_elec 119 8.7 0.040 0.046 3.508 3.539 qs_ot_get_p 119 10.4 0.002 0.006 3.308 3.388 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.718 3.177 prepare_preconditioner 11 7.9 0.001 0.003 2.581 2.619 make_preconditioner 11 8.9 0.012 0.083 2.580 2.619 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.081 2.526 apply_single 119 13.6 0.000 0.000 2.081 2.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.180 2.482 calculate_first_density_matrix 1 7.0 0.001 0.003 2.370 2.373 mp_waitall_1 137007 16.6 1.699 2.343 1.699 2.343 make_m2s 4572 13.5 0.054 0.056 2.092 2.201 rs_pw_transfer 974 11.9 0.009 0.010 1.933 2.133 make_images 4572 14.5 0.269 0.328 1.985 2.093 density_rs2pw 119 9.7 0.004 0.004 1.889 2.084 qs_ot_p2m_diag 50 11.0 0.016 0.026 2.067 2.076 calculate_dm_sparse 119 9.5 0.000 0.000 2.018 2.065 pw_transfer 1439 11.6 0.066 0.070 1.981 2.014 grid_integrate_task_list 119 12.3 1.887 1.971 1.887 1.971 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.888 1.926 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.874 1.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.829 1.831 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.759 1.761 ot_diis_step 108 11.5 0.012 0.012 1.760 1.761 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.656 1.669 fft_wrap_pw1pw2_140 487 13.2 0.287 0.299 1.563 1.601 mp_sum_l 7287 12.8 1.060 1.596 1.060 1.596 potential_pw2rs 119 12.3 0.009 0.010 1.525 1.530 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.518 1.529 cp_fm_diag_elpa_base 50 14.0 1.485 1.503 1.516 1.527 acc_transpose_blocks 27432 15.5 0.108 0.111 1.479 1.521 jit_kernel_multiply 8 15.8 0.788 1.477 0.788 1.477 grid_collocate_task_list 119 9.7 1.261 1.372 1.261 1.372 multiply_cannon_metrocomm3 27432 15.5 0.039 0.040 0.758 1.365 fft3d_ps 1201 14.6 0.557 0.606 1.328 1.353 wfi_extrapolate 11 7.9 0.001 0.001 1.312 1.312 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.205 1.213 cp_fm_upper_to_full 72 14.2 0.809 1.154 0.809 1.154 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.089 1.110 qs_energies_init_hamiltonians 11 5.9 0.004 0.008 1.099 1.100 multiply_cannon_sync_h2d 27432 15.5 0.997 1.090 0.997 1.090 dbcsr_complete_redistribute 329 12.2 0.120 0.142 0.762 1.037 make_images_data 4572 15.5 0.044 0.048 0.803 0.914 acc_transpose_blocks_kernels 27432 16.5 0.268 0.275 0.857 0.878 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.872 mp_alltoall_d11v 2130 13.8 0.712 0.871 0.712 0.871 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.696 0.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.855 0.860 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.571 0.838 cp_fm_cholesky_invert 11 10.9 0.734 0.737 0.734 0.737 mp_alltoall_i22 627 13.8 0.450 0.735 0.450 0.735 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.235000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=513.454545, yerr=4.030873 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 594.948096E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 30.138 30.139 qs_mol_dyn_low 1 2.0 0.003 0.003 29.955 29.962 qs_forces 11 3.9 0.003 0.003 29.582 29.582 qs_energies 11 4.9 0.001 0.001 27.800 27.803 scf_env_do_scf 11 5.9 0.004 0.015 22.726 22.726 scf_env_do_scf_inner_loop 108 6.5 0.005 0.017 19.908 19.908 velocity_verlet 10 3.0 0.001 0.001 16.228 16.250 dbcsr_multiply_generic 2286 12.5 0.093 0.099 12.660 12.771 qs_scf_new_mos 108 7.5 0.001 0.001 11.513 11.545 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.512 11.545 ot_scf_mini 108 9.5 0.002 0.002 10.805 10.835 multiply_cannon 2286 13.5 0.232 0.237 9.889 10.359 multiply_cannon_loop 2286 14.5 0.330 0.342 8.925 9.205 rebuild_ks_matrix 119 8.3 0.000 0.000 6.346 6.374 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.345 6.373 ot_mini 108 10.5 0.001 0.001 6.164 6.206 multiply_cannon_multrec 9144 15.5 1.715 1.955 5.883 6.169 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.629 5.655 qs_ot_get_derivative 108 11.5 0.001 0.001 4.805 4.840 dbcsr_mm_accdrv_process 12550 15.8 2.948 3.887 4.066 4.144 sum_up_and_integrate 119 10.3 0.039 0.044 3.927 3.935 integrate_v_rspace 119 11.3 0.003 0.003 3.888 3.898 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.792 3.801 calculate_rho_elec 119 8.7 0.060 0.061 3.792 3.801 init_scf_run 11 5.9 0.000 0.001 3.500 3.500 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.500 3.500 qs_ot_get_p 119 10.4 0.001 0.001 2.786 2.823 init_scf_loop 11 6.9 0.000 0.000 2.792 2.795 mp_waitall_1 115863 16.7 1.811 2.372 1.811 2.372 calculate_first_density_matrix 1 7.0 0.000 0.000 2.246 2.247 pw_transfer 1439 11.6 0.067 0.070 2.223 2.236 density_rs2pw 119 9.7 0.004 0.004 2.069 2.208 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.128 2.144 make_m2s 4572 13.5 0.035 0.035 1.883 2.083 rs_pw_transfer 974 11.9 0.008 0.009 1.952 2.073 grid_integrate_task_list 119 12.3 1.911 1.995 1.911 1.995 make_images 4572 14.5 0.269 0.301 1.792 1.992 prepare_preconditioner 11 7.9 0.000 0.000 1.939 1.949 make_preconditioner 11 8.9 0.000 0.000 1.939 1.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.789 1.826 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.799 1.801 calculate_dm_sparse 119 9.5 0.000 0.000 1.747 1.770 jit_kernel_multiply 10 15.8 1.080 1.751 1.080 1.751 fft_wrap_pw1pw2_140 487 13.2 0.364 0.373 1.719 1.736 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.687 1.713 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.608 1.608 potential_pw2rs 119 12.3 0.010 0.011 1.559 1.569 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.560 1.561 fft3d_ps 1201 14.6 0.562 0.574 1.463 1.476 grid_collocate_task_list 119 9.7 1.307 1.425 1.307 1.425 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.389 1.398 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.348 1.349 ot_diis_step 108 11.5 0.013 0.013 1.341 1.342 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.332 1.340 cp_fm_diag_elpa_base 50 14.0 1.305 1.321 1.330 1.339 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.208 1.235 apply_single 119 13.6 0.000 0.000 1.208 1.234 wfi_extrapolate 11 7.9 0.001 0.001 1.203 1.203 make_images_data 4572 15.5 0.039 0.042 0.915 1.179 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.891 1.178 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.137 1.147 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.129 1.144 mp_alltoall_d11v 2130 13.8 0.878 0.970 0.878 0.970 cp_fm_cholesky_invert 11 10.9 0.926 0.928 0.926 0.928 acc_transpose_blocks 9144 15.5 0.037 0.038 0.915 0.925 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.868 0.920 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.392 0.881 qs_env_update_s_mstruct 11 6.9 0.001 0.003 0.802 0.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.845 mp_alltoall_z22v 1201 16.6 0.774 0.811 0.774 0.811 multiply_cannon_sync_h2d 9144 15.5 0.719 0.806 0.719 0.806 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.761 0.770 mp_allgather_i34 2286 14.5 0.291 0.749 0.291 0.749 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.675 0.681 qs_create_task_list 11 7.9 0.001 0.001 0.649 0.675 generate_qs_task_list 11 8.9 0.192 0.268 0.648 0.674 dbcsr_complete_redistribute 329 12.2 0.162 0.172 0.593 0.622 mp_waitany 5200 13.7 0.491 0.608 0.491 0.608 mp_sum_l 7287 12.8 0.454 0.604 0.454 0.604 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=30.139000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=565.090909, yerr=3.728381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/08/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 235.585836E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1388964 0.0% 0.0% 100.0% average stack size 0.0 0.0 105.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 763.813888E+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 0.033 0.037 44.251 44.252 qs_mol_dyn_low 1 2.0 0.003 0.003 43.883 43.891 qs_forces 11 3.9 0.003 0.003 43.816 43.817 qs_energies 11 4.9 0.001 0.001 41.748 41.750 scf_env_do_scf 11 5.9 0.001 0.001 35.494 35.495 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.064 27.067 velocity_verlet 10 3.0 0.001 0.001 25.872 25.879 dbcsr_multiply_generic 2286 12.5 0.101 0.103 18.536 18.786 qs_scf_new_mos 108 7.5 0.001 0.001 16.629 16.717 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.628 16.717 ot_scf_mini 108 9.5 0.002 0.002 15.508 15.606 multiply_cannon 2286 13.5 0.303 0.322 14.188 15.291 multiply_cannon_loop 2286 14.5 0.344 0.349 12.829 13.822 ot_mini 108 10.5 0.001 0.001 9.311 9.421 multiply_cannon_multrec 9144 15.5 3.358 4.724 8.655 8.770 init_scf_loop 11 6.9 0.000 0.000 8.400 8.404 rebuild_ks_matrix 119 8.3 0.000 0.000 7.784 7.921 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.784 7.920 prepare_preconditioner 11 7.9 0.000 0.000 7.349 7.363 make_preconditioner 11 8.9 0.000 0.000 7.349 7.363 qs_ot_get_derivative 108 11.5 0.001 0.001 7.172 7.269 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.862 7.210 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.014 7.135 dbcsr_mm_accdrv_process 12550 15.8 4.270 5.869 5.174 6.487 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.645 4.655 calculate_rho_elec 119 8.7 0.118 0.121 4.645 4.655 cp_fm_upper_to_full 72 14.2 3.190 4.589 3.190 4.589 sum_up_and_integrate 119 10.3 0.065 0.066 4.324 4.331 integrate_v_rspace 119 11.3 0.003 0.003 4.259 4.265 init_scf_run 11 5.9 0.000 0.001 4.137 4.137 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.136 4.137 mp_waitall_1 94719 16.7 2.925 4.000 2.925 4.000 qs_ot_get_p 119 10.4 0.001 0.001 3.243 3.383 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.698 3.148 pw_transfer 1439 11.6 0.069 0.070 2.974 2.979 dbcsr_complete_redistribute 329 12.2 0.283 0.289 2.098 2.925 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.874 2.880 make_m2s 4572 13.5 0.038 0.038 2.624 2.863 make_images 4572 14.5 0.354 0.384 2.503 2.742 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.778 2.604 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.302 2.517 apply_single 119 13.6 0.000 0.000 2.302 2.517 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.553 2.495 density_rs2pw 119 9.7 0.004 0.004 2.457 2.485 calculate_first_density_matrix 1 7.0 0.000 0.000 2.431 2.433 fft_wrap_pw1pw2_140 487 13.2 0.615 0.619 2.340 2.346 mp_alltoall_i22 627 13.8 1.457 2.325 1.457 2.325 calculate_dm_sparse 119 9.5 0.000 0.000 2.296 2.324 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.482 2.299 grid_integrate_task_list 119 12.3 2.124 2.145 2.124 2.145 ot_diis_step 108 11.5 0.014 0.015 2.107 2.108 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.052 2.053 fft3d_ps 1201 14.6 0.597 0.606 1.893 1.897 qs_ot_p2m_diag 50 11.0 0.042 0.043 1.888 1.889 qs_energies_init_hamiltonians 11 5.9 0.003 0.004 1.839 1.839 mp_sum_l 7287 12.8 1.175 1.829 1.175 1.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.759 1.807 rs_pw_transfer 974 11.9 0.009 0.009 1.701 1.742 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.686 1.737 cp_fm_cholesky_invert 11 10.9 1.691 1.694 1.691 1.694 wfi_extrapolate 11 7.9 0.001 0.001 1.623 1.624 make_images_data 4572 15.5 0.043 0.046 1.301 1.622 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.619 1.619 hybrid_alltoall_any 4725 16.4 0.089 0.150 1.303 1.585 potential_pw2rs 119 12.3 0.014 0.015 1.578 1.580 grid_collocate_task_list 119 9.7 1.518 1.539 1.518 1.539 mp_alltoall_d11v 2130 13.8 1.321 1.340 1.321 1.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.338 1.338 cp_fm_diag_elpa_base 50 14.0 1.191 1.244 1.336 1.336 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.326 1.333 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.242 1.260 qs_env_update_s_mstruct 11 6.9 0.010 0.011 1.171 1.189 mp_alltoall_z22v 1201 16.6 1.159 1.181 1.159 1.181 jit_kernel_multiply 6 15.7 0.876 1.131 0.876 1.131 acc_transpose_blocks 9144 15.5 0.038 0.039 1.059 1.064 multiply_cannon_sync_h2d 9144 15.5 1.042 1.046 1.042 1.046 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.993 1.038 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.991 1.005 qs_create_task_list 11 7.9 0.006 0.007 0.982 0.994 generate_qs_task_list 11 8.9 0.368 0.387 0.976 0.987 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=44.252000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=714.636364, yerr=16.255196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 498.675712E+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 1821501. 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 0.025 0.043 82.166 82.168 qs_mol_dyn_low 1 2.0 0.003 0.003 81.894 81.904 qs_forces 11 3.9 0.003 0.003 81.826 81.827 qs_energies 11 4.9 0.001 0.002 78.933 78.948 scf_env_do_scf 11 5.9 0.001 0.001 69.903 69.906 scf_env_do_scf_inner_loop 99 6.5 0.003 0.038 64.496 64.497 dbcsr_multiply_generic 2055 12.4 0.109 0.112 49.715 49.995 qs_scf_new_mos 99 7.5 0.001 0.001 46.428 46.555 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.428 46.554 ot_scf_mini 99 9.5 0.002 0.002 44.118 44.197 multiply_cannon 2055 13.4 0.178 0.183 41.615 42.664 multiply_cannon_loop 2055 14.4 1.489 1.520 40.573 41.467 velocity_verlet 10 3.0 0.001 0.002 40.882 40.883 ot_mini 99 10.5 0.001 0.001 25.408 25.496 qs_ot_get_derivative 99 11.5 0.001 0.001 18.584 18.665 multiply_cannon_multrec 49320 15.4 12.664 13.238 17.416 18.142 rebuild_ks_matrix 110 8.3 0.000 0.001 15.099 15.213 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.098 15.213 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.277 13.388 multiply_cannon_sync_h2d 49320 15.4 10.544 11.215 10.544 11.215 mp_waitall_1 220248 16.4 10.067 10.841 10.067 10.841 qs_ot_get_p 110 10.4 0.001 0.001 10.273 10.390 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.198 7.794 apply_single 110 13.6 0.000 0.001 7.198 7.794 sum_up_and_integrate 110 10.3 0.036 0.043 7.501 7.516 integrate_v_rspace 110 11.3 0.003 0.004 7.465 7.487 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.117 7.130 init_scf_run 11 5.9 0.000 0.001 6.965 6.965 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.964 6.965 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.420 6.935 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.781 6.900 calculate_rho_elec 110 8.6 0.021 0.026 6.781 6.899 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 5.640 6.666 ot_diis_step 99 11.5 0.006 0.006 6.660 6.661 cp_dbcsr_syevd 48 12.0 0.002 0.003 6.279 6.280 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.683 5.731 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.562 5.595 cp_fm_diag_elpa_base 48 14.0 5.544 5.577 5.560 5.593 init_scf_loop 11 6.9 0.000 0.000 5.381 5.382 dbcsr_mm_accdrv_process 87628 16.1 1.820 1.899 4.627 4.916 rs_pw_transfer 902 11.9 0.012 0.013 3.892 4.450 make_m2s 4110 13.4 0.060 0.064 4.268 4.370 make_images 4110 14.4 0.177 0.191 4.172 4.277 mp_sum_l 6594 12.7 3.363 4.206 3.363 4.206 wfi_extrapolate 11 7.9 0.001 0.001 4.015 4.015 density_rs2pw 110 9.6 0.004 0.005 3.471 3.968 calculate_dm_sparse 110 9.5 0.001 0.001 3.682 3.807 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.575 3.579 grid_integrate_task_list 110 12.3 3.327 3.498 3.327 3.498 pw_transfer 1331 11.6 0.055 0.064 3.293 3.366 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.256 3.301 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.205 3.281 prepare_preconditioner 11 7.9 0.000 0.000 3.243 3.264 make_preconditioner 11 8.9 0.000 0.000 3.243 3.264 multiply_cannon_metrocomm1 49320 15.4 0.060 0.062 2.345 3.187 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.064 3.142 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.026 3.064 calculate_first_density_matrix 1 7.0 0.000 0.000 2.860 2.865 potential_pw2rs 110 12.3 0.006 0.007 2.839 2.862 fft_wrap_pw1pw2_140 451 13.1 0.453 0.499 2.726 2.808 jit_kernel_multiply 13 15.9 2.537 2.584 2.537 2.584 mp_alltoall_d11v 2046 13.8 2.200 2.577 2.200 2.577 fft3d_ps 1111 14.6 0.794 0.880 2.398 2.449 grid_collocate_task_list 110 9.6 2.180 2.425 2.180 2.425 mp_waitany 14300 13.8 1.898 2.381 1.898 2.381 acc_transpose_blocks 49320 15.4 0.198 0.206 2.028 2.104 make_images_data 4110 15.4 0.042 0.045 1.889 2.036 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.965 1.984 mp_sum_d 3889 11.9 1.377 1.940 1.377 1.940 hybrid_alltoall_any 4261 16.3 0.082 0.484 1.607 1.897 cp_fm_cholesky_invert 11 10.9 1.785 1.789 1.785 1.789 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.652 1.677 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.168000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=473.000000, yerr=2.696799 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 582.197248E+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 1824256. 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 0.011 0.026 75.263 75.264 qs_mol_dyn_low 1 2.0 0.003 0.003 75.047 75.057 qs_forces 11 3.9 0.003 0.003 74.974 74.975 qs_energies 11 4.9 0.001 0.001 71.507 71.509 scf_env_do_scf 11 5.9 0.000 0.001 62.633 62.636 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 54.447 54.448 dbcsr_multiply_generic 2055 12.4 0.118 0.123 39.447 39.628 velocity_verlet 10 3.0 0.001 0.001 38.106 38.107 qs_scf_new_mos 99 7.5 0.001 0.001 36.297 36.469 qs_scf_loop_do_ot 99 8.5 0.001 0.001 36.297 36.468 ot_scf_mini 99 9.5 0.003 0.003 34.471 34.633 multiply_cannon 2055 13.4 0.221 0.243 31.485 32.731 multiply_cannon_loop 2055 14.4 0.919 0.946 29.988 30.807 ot_mini 99 10.5 0.001 0.001 19.221 19.387 multiply_cannon_multrec 24660 15.4 7.548 9.112 13.753 15.106 rebuild_ks_matrix 110 8.3 0.000 0.001 14.880 14.984 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.880 14.983 qs_ot_get_derivative 99 11.5 0.001 0.001 13.273 13.436 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.126 13.215 mp_waitall_1 176588 16.5 8.674 11.223 8.674 11.223 qs_ot_get_p 110 10.4 0.001 0.001 8.311 8.503 init_scf_loop 11 6.9 0.000 0.000 8.144 8.145 multiply_cannon_sync_h2d 24660 15.4 6.986 7.856 6.986 7.856 multiply_cannon_metrocomm3 24660 15.4 0.067 0.072 5.273 7.808 sum_up_and_integrate 110 10.3 0.053 0.059 7.462 7.476 integrate_v_rspace 110 11.3 0.003 0.003 7.409 7.422 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.557 7.215 apply_single 110 13.6 0.000 0.001 6.557 7.215 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.137 7.148 calculate_rho_elec 110 8.6 0.040 0.048 7.136 7.147 init_scf_run 11 5.9 0.000 0.001 6.458 6.458 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.457 6.458 dbcsr_mm_accdrv_process 52282 16.1 4.513 5.612 6.046 6.346 prepare_preconditioner 11 7.9 0.000 0.000 6.023 6.043 make_preconditioner 11 8.9 0.000 0.000 6.023 6.043 qs_ot_p2m_diag 48 11.0 0.028 0.043 5.901 5.925 ot_diis_step 99 11.5 0.010 0.010 5.879 5.881 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.577 5.755 make_m2s 4110 13.4 0.057 0.060 5.088 5.630 make_images 4110 14.4 0.397 0.441 4.977 5.515 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.789 5.507 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.327 5.327 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.588 4.602 cp_fm_diag_elpa_base 48 14.0 4.508 4.551 4.585 4.598 density_rs2pw 110 9.6 0.004 0.005 4.002 4.560 pw_transfer 1331 11.6 0.068 0.075 4.317 4.479 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.209 4.373 rs_pw_transfer 902 11.9 0.012 0.013 3.699 4.286 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.709 3.798 wfi_extrapolate 11 7.9 0.001 0.001 3.732 3.732 fft_wrap_pw1pw2_140 451 13.1 0.518 0.534 3.366 3.532 grid_integrate_task_list 110 12.3 3.230 3.461 3.230 3.461 fft3d_ps 1111 14.6 1.108 1.335 3.186 3.366 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.325 3.326 make_images_data 4110 15.4 0.046 0.051 2.689 3.246 calculate_dm_sparse 110 9.5 0.001 0.001 3.132 3.162 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.345 3.097 potential_pw2rs 110 12.3 0.008 0.009 3.004 3.030 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.942 2.971 mp_sum_l 6594 12.7 2.242 2.883 2.242 2.883 cp_fm_cholesky_invert 11 10.9 2.869 2.878 2.869 2.878 mp_alltoall_d11v 2046 13.8 2.271 2.697 2.271 2.697 calculate_first_density_matrix 1 7.0 0.000 0.000 2.632 2.635 grid_collocate_task_list 110 9.6 2.127 2.547 2.127 2.547 jit_kernel_multiply 11 16.3 1.182 2.115 1.182 2.115 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.088 2.106 mp_allgather_i34 2055 14.4 0.824 1.975 0.824 1.975 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.925 1.940 mp_alltoall_z22v 1111 16.6 1.793 1.890 1.793 1.890 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.877 1.878 mp_waitany 10164 13.8 1.350 1.875 1.350 1.875 multiply_cannon_metrocomm4 22605 15.4 0.077 0.081 0.799 1.716 rs_pw_transfer_RS2PW_140 121 11.5 0.206 0.216 1.033 1.593 mp_irecv_dv 57340 16.2 0.673 1.592 0.673 1.592 dbcsr_complete_redistribute 325 12.2 0.235 0.291 1.315 1.585 acc_transpose_blocks 24660 15.4 0.108 0.111 1.545 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.546 1.557 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=75.264000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=551.181818, yerr=6.293116 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 659.636224E+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 0.024 0.052 63.618 63.630 qs_mol_dyn_low 1 2.0 0.003 0.004 63.210 63.223 qs_forces 11 3.9 0.003 0.004 63.138 63.139 qs_energies 11 4.9 0.005 0.018 59.891 59.894 scf_env_do_scf 11 5.9 0.001 0.002 51.362 51.362 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 42.859 42.860 velocity_verlet 10 3.0 0.001 0.002 32.559 32.561 dbcsr_multiply_generic 2055 12.4 0.113 0.117 29.759 30.043 qs_scf_new_mos 99 7.5 0.001 0.001 26.738 26.834 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.737 26.833 ot_scf_mini 99 9.5 0.002 0.003 25.429 25.554 multiply_cannon 2055 13.4 0.213 0.225 22.772 24.245 multiply_cannon_loop 2055 14.4 0.615 0.630 21.530 22.574 ot_mini 99 10.5 0.001 0.001 14.037 14.158 rebuild_ks_matrix 110 8.3 0.000 0.000 13.099 13.259 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.099 13.258 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.585 11.729 mp_waitall_1 139946 16.5 7.750 10.845 7.750 10.845 multiply_cannon_multrec 16440 15.4 3.806 4.568 9.644 10.425 qs_ot_get_derivative 99 11.5 0.001 0.001 9.445 9.562 init_scf_loop 11 6.9 0.001 0.006 8.470 8.470 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.532 7.549 sum_up_and_integrate 110 10.3 0.060 0.062 7.016 7.037 integrate_v_rspace 110 11.3 0.003 0.003 6.955 6.976 prepare_preconditioner 11 7.9 0.003 0.023 6.683 6.704 make_preconditioner 11 8.9 0.001 0.007 6.680 6.704 qs_ot_get_p 110 10.4 0.001 0.001 6.394 6.540 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.385 6.397 calculate_rho_elec 110 8.6 0.059 0.059 6.384 6.396 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.962 6.350 init_scf_run 11 5.9 0.000 0.001 6.078 6.079 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.078 6.078 dbcsr_mm_accdrv_process 34862 16.1 4.634 5.454 5.692 5.850 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.006 5.458 apply_single 110 13.6 0.000 0.000 5.006 5.458 make_m2s 4110 13.4 0.050 0.051 4.605 4.964 make_images 4110 14.4 0.389 0.507 4.489 4.847 density_rs2pw 110 9.6 0.004 0.005 3.372 4.620 ot_diis_step 99 11.5 0.011 0.011 4.557 4.559 qs_ot_p2m_diag 48 11.0 0.042 0.046 4.472 4.475 multiply_cannon_sync_h2d 16440 15.4 3.656 4.260 3.656 4.260 rs_pw_transfer 902 11.9 0.011 0.012 2.923 4.160 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.073 4.074 pw_transfer 1331 11.6 0.066 0.073 3.981 3.995 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.874 3.889 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.067 3.727 grid_integrate_task_list 110 12.3 3.263 3.592 3.263 3.592 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.455 3.466 cp_fm_diag_elpa_base 48 14.0 3.377 3.418 3.452 3.463 fft_wrap_pw1pw2_140 451 13.1 0.636 0.643 3.240 3.256 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.246 3.248 make_images_data 4110 15.4 0.043 0.047 2.648 3.101 calculate_first_density_matrix 1 7.0 0.001 0.005 3.011 3.013 wfi_extrapolate 11 7.9 0.001 0.001 2.972 2.972 hybrid_alltoall_any 4261 16.3 0.106 0.375 2.304 2.968 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.788 2.843 fft3d_ps 1111 14.6 1.098 1.106 2.697 2.711 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.621 2.691 calculate_dm_sparse 110 9.5 0.001 0.001 2.610 2.640 mp_waitany 17072 13.8 1.307 2.595 1.307 2.595 cp_fm_cholesky_invert 11 10.9 2.575 2.580 2.575 2.580 potential_pw2rs 110 12.3 0.011 0.011 2.520 2.531 grid_collocate_task_list 110 9.6 2.144 2.517 2.144 2.517 mp_alltoall_d11v 2046 13.8 1.962 2.462 1.962 2.462 mp_sum_l 6594 12.7 1.771 2.450 1.771 2.450 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.903 2.439 mp_irecv_dv 48980 15.7 0.833 2.316 0.833 2.316 rs_pw_transfer_RS2PW_140 121 11.5 0.174 0.178 0.960 2.193 dbcsr_complete_redistribute 325 12.2 0.425 0.485 1.626 2.089 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 2.004 2.005 mp_allgather_i34 2055 14.4 0.565 1.850 0.565 1.850 cp_fm_upper_to_full 70 14.2 1.397 1.811 1.397 1.811 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.722 1.742 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.148 1.604 jit_kernel_multiply 8 16.6 0.670 1.579 0.670 1.579 cp_fm_cholesky_decompose 22 10.9 1.542 1.561 1.542 1.561 rs_gather_matrices 110 12.3 0.233 0.258 1.095 1.558 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.481 1.495 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.359 1.465 mp_alltoall_z22v 1111 16.6 1.390 1.413 1.390 1.413 acc_transpose_blocks 16440 15.4 0.073 0.075 1.375 1.404 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.333 1.345 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=63.630000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=623.000000, yerr=9.155227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 719.540224E+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 0.014 0.028 67.484 67.485 qs_mol_dyn_low 1 2.0 0.003 0.003 67.219 67.230 qs_forces 11 3.9 0.003 0.004 67.145 67.148 qs_energies 11 4.9 0.001 0.001 63.666 63.672 scf_env_do_scf 11 5.9 0.001 0.001 55.423 55.426 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 43.644 43.645 velocity_verlet 10 3.0 0.001 0.002 37.054 37.056 dbcsr_multiply_generic 2055 12.4 0.115 0.121 30.136 30.375 qs_scf_new_mos 99 7.5 0.001 0.001 27.898 28.013 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.898 28.013 ot_scf_mini 99 9.5 0.002 0.003 26.223 26.318 multiply_cannon 2055 13.4 0.247 0.270 22.622 23.691 multiply_cannon_loop 2055 14.4 0.883 0.903 21.133 21.763 ot_mini 99 10.5 0.001 0.001 14.448 14.560 multiply_cannon_multrec 24660 15.4 4.213 6.818 12.641 14.016 rebuild_ks_matrix 110 8.3 0.000 0.000 12.635 12.735 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.016 12.635 12.735 init_scf_loop 11 6.9 0.000 0.000 11.739 11.739 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.160 11.249 qs_ot_get_derivative 99 11.5 0.001 0.001 10.238 10.333 prepare_preconditioner 11 7.9 0.000 0.000 9.948 9.966 make_preconditioner 11 8.9 0.000 0.000 9.948 9.966 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.146 9.643 dbcsr_mm_accdrv_process 52304 16.0 7.011 8.528 8.283 9.219 sum_up_and_integrate 110 10.3 0.068 0.071 6.823 6.836 qs_ot_get_p 110 10.4 0.001 0.001 6.660 6.790 integrate_v_rspace 110 11.3 0.003 0.003 6.755 6.771 mp_waitall_1 121746 16.5 4.678 6.695 4.678 6.695 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.471 6.480 calculate_rho_elec 110 8.6 0.078 0.081 6.470 6.480 make_m2s 4110 13.4 0.059 0.062 5.753 6.111 make_images 4110 14.4 0.574 0.693 5.613 5.968 init_scf_run 11 5.9 0.000 0.001 5.503 5.503 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.502 5.503 cp_fm_upper_to_full 70 14.2 3.404 4.957 3.404 4.957 qs_ot_p2m_diag 48 11.0 0.054 0.062 4.644 4.659 ot_diis_step 99 11.5 0.011 0.011 4.171 4.171 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.042 4.144 apply_single 110 13.6 0.000 0.000 4.041 4.144 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.085 4.085 pw_transfer 1331 11.6 0.066 0.076 3.982 4.023 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.875 3.919 density_rs2pw 110 9.6 0.004 0.004 3.321 3.841 dbcsr_complete_redistribute 325 12.2 0.416 0.455 2.673 3.840 grid_integrate_task_list 110 12.3 3.381 3.611 3.381 3.611 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.504 3.514 cp_fm_diag_elpa_base 48 14.0 3.336 3.408 3.499 3.511 multiply_cannon_sync_h2d 24660 15.4 3.197 3.370 3.197 3.370 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.301 3.359 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.284 3.328 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.146 3.301 fft_wrap_pw1pw2_140 451 13.1 0.667 0.689 3.250 3.297 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.412 3.283 make_images_data 4110 15.4 0.046 0.050 2.914 3.282 hybrid_alltoall_any 4261 16.3 0.121 0.459 2.434 3.249 rs_pw_transfer 902 11.9 0.010 0.012 2.633 3.170 calculate_dm_sparse 110 9.5 0.001 0.001 2.965 2.995 wfi_extrapolate 11 7.9 0.001 0.001 2.990 2.990 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.794 2.935 mp_alltoall_i22 605 13.7 1.680 2.884 1.680 2.884 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.796 2.798 cp_fm_cholesky_invert 11 10.9 2.664 2.672 2.664 2.672 fft3d_ps 1111 14.6 1.090 1.125 2.638 2.656 grid_collocate_task_list 110 9.6 2.239 2.561 2.239 2.561 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.442 2.487 calculate_first_density_matrix 1 7.0 0.000 0.000 2.419 2.421 potential_pw2rs 110 12.3 0.013 0.013 2.314 2.325 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.253 2.256 mp_alltoall_d11v 2046 13.8 1.789 2.125 1.789 2.125 jit_kernel_multiply 10 15.7 0.943 1.872 0.943 1.872 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.759 1.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.726 1.738 cp_fm_cholesky_decompose 22 10.9 1.690 1.729 1.690 1.729 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.710 mp_sum_l 6594 12.7 1.084 1.689 1.084 1.689 multiply_cannon_metrocomm4 20550 15.4 0.060 0.063 0.851 1.629 mp_waitany 13376 13.8 1.127 1.604 1.127 1.604 mp_allgather_i34 2055 14.4 0.593 1.599 0.593 1.599 mp_irecv_dv 62702 16.1 0.750 1.552 0.750 1.552 acc_transpose_blocks 24660 15.4 0.101 0.103 1.514 1.550 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.525 1.540 mp_alltoall_z22v 1111 16.6 1.334 1.395 1.334 1.395 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.176 0.837 1.370 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=67.485000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=684.727273, yerr=5.690357 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 830.058496E+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 0.012 0.028 55.138 55.138 qs_mol_dyn_low 1 2.0 0.003 0.003 54.882 54.914 qs_forces 11 3.9 0.003 0.003 54.786 54.786 qs_energies 11 4.9 0.006 0.006 51.059 51.064 scf_env_do_scf 11 5.9 0.001 0.005 42.811 42.812 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 35.091 35.092 velocity_verlet 10 3.0 0.001 0.002 31.115 31.120 dbcsr_multiply_generic 2055 12.4 0.104 0.110 22.603 22.777 qs_scf_new_mos 99 7.5 0.001 0.001 19.954 19.995 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.954 19.995 ot_scf_mini 99 9.5 0.002 0.002 18.717 18.744 multiply_cannon 2055 13.4 0.247 0.257 17.209 18.682 multiply_cannon_loop 2055 14.4 0.321 0.333 15.866 16.235 rebuild_ks_matrix 110 8.3 0.000 0.000 12.066 12.090 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.024 12.066 12.090 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.679 10.700 ot_mini 99 10.5 0.001 0.001 10.258 10.278 multiply_cannon_multrec 8220 15.4 3.210 4.569 7.418 8.348 init_scf_loop 11 6.9 0.000 0.000 7.675 7.677 mp_waitall_1 103326 16.6 5.960 7.504 5.960 7.504 sum_up_and_integrate 110 10.3 0.080 0.082 6.641 6.655 integrate_v_rspace 110 11.3 0.003 0.003 6.561 6.573 qs_ot_get_derivative 99 11.5 0.001 0.001 6.492 6.517 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.301 6.313 calculate_rho_elec 110 8.6 0.114 0.115 6.301 6.313 prepare_preconditioner 11 7.9 0.000 0.000 5.950 5.953 make_preconditioner 11 8.9 0.000 0.000 5.950 5.953 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.527 5.597 init_scf_run 11 5.9 0.000 0.001 5.130 5.130 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.130 5.130 dbcsr_mm_accdrv_process 17442 15.9 2.825 3.944 4.078 4.976 qs_ot_get_p 110 10.4 0.001 0.001 4.592 4.611 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.048 4.488 make_m2s 4110 13.4 0.039 0.040 4.152 4.435 make_images 4110 14.4 0.635 0.686 4.021 4.304 pw_transfer 1331 11.6 0.067 0.072 3.909 3.919 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.801 3.815 ot_diis_step 99 11.5 0.012 0.012 3.746 3.747 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.669 3.700 apply_single 110 13.6 0.000 0.000 3.668 3.700 grid_integrate_task_list 110 12.3 3.480 3.664 3.480 3.664 density_rs2pw 110 9.6 0.004 0.004 2.994 3.370 fft_wrap_pw1pw2_140 451 13.1 0.830 0.843 3.291 3.308 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.247 3.250 multiply_cannon_sync_h2d 8220 15.4 2.919 3.018 2.919 3.018 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.954 2.955 cp_fm_cholesky_invert 11 10.9 2.842 2.845 2.842 2.845 make_images_data 4110 15.4 0.037 0.043 2.388 2.839 wfi_extrapolate 11 7.9 0.001 0.001 2.738 2.738 hybrid_alltoall_any 4261 16.3 0.200 0.861 2.300 2.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.659 2.660 grid_collocate_task_list 110 9.6 2.335 2.592 2.335 2.592 rs_pw_transfer 902 11.9 0.010 0.012 2.148 2.547 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.500 2.501 calculate_dm_sparse 110 9.5 0.001 0.001 2.459 2.498 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.445 2.452 cp_fm_diag_elpa_base 48 14.0 2.390 2.410 2.443 2.451 fft3d_ps 1111 14.6 1.132 1.142 2.350 2.366 calculate_first_density_matrix 1 7.0 0.000 0.000 2.297 2.298 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.146 2.161 potential_pw2rs 110 12.3 0.016 0.016 2.053 2.056 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.772 1.980 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.864 1.884 mp_alltoall_d11v 2046 13.8 1.597 1.863 1.597 1.863 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.747 1.758 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.716 1.719 cp_fm_cholesky_decompose 22 10.9 1.669 1.691 1.669 1.691 mp_allgather_i34 2055 14.4 0.528 1.643 0.528 1.643 qs_env_update_s_mstruct 11 6.9 0.001 0.009 1.498 1.626 dbcsr_complete_redistribute 325 12.2 0.569 0.601 1.446 1.545 jit_kernel_multiply 9 15.8 0.946 1.524 0.946 1.524 mp_waitany 9240 13.8 1.064 1.483 1.064 1.483 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.424 1.433 qs_create_task_list 11 7.9 0.000 0.001 1.218 1.325 generate_qs_task_list 11 8.9 0.374 0.442 1.218 1.325 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.166 0.796 1.200 rs_gather_matrices 110 12.3 0.327 0.373 0.938 1.158 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.112 1.136 multiply_cannon_metrocomm4 6165 15.4 0.018 0.019 0.485 1.106 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.138000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.909091, yerr=11.081964 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 1.335861E+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 0.017 0.039 108.078 108.079 qs_mol_dyn_low 1 2.0 0.003 0.003 107.702 107.730 qs_forces 11 3.9 0.003 0.004 107.632 107.633 qs_energies 11 4.9 0.007 0.024 103.156 103.160 scf_env_do_scf 11 5.9 0.000 0.001 92.195 92.195 velocity_verlet 10 3.0 0.001 0.001 63.068 63.074 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 59.045 59.047 qs_scf_new_mos 99 7.5 0.001 0.001 34.649 34.688 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.648 34.687 dbcsr_multiply_generic 2055 12.4 0.121 0.126 33.023 33.085 init_scf_loop 11 6.9 0.000 0.000 33.055 33.060 ot_scf_mini 99 9.5 0.002 0.002 32.662 32.687 prepare_preconditioner 11 7.9 0.000 0.000 30.663 30.670 make_preconditioner 11 8.9 0.000 0.000 30.663 30.670 make_full_inverse_cholesky 11 9.9 0.000 0.000 24.473 30.066 multiply_cannon 2055 13.4 0.347 0.368 22.433 23.653 multiply_cannon_loop 2055 14.4 0.343 0.346 20.392 20.749 cp_fm_upper_to_full 70 14.2 13.212 19.189 13.212 19.189 rebuild_ks_matrix 110 8.3 0.000 0.001 17.892 17.921 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 17.892 17.921 ot_mini 99 10.5 0.001 0.001 16.571 16.598 qs_ks_update_qs_env 110 7.6 0.001 0.001 16.268 16.295 mp_waitall_1 84994 16.7 11.659 13.077 11.659 13.077 qs_ot_get_derivative 99 11.5 0.001 0.001 11.909 11.932 dbcsr_complete_redistribute 325 12.2 1.012 1.049 8.185 11.739 qs_rho_update_rho_low 110 7.6 0.001 0.001 10.890 10.928 calculate_rho_elec 110 8.6 0.225 0.225 10.889 10.927 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.923 10.483 qs_ot_get_p 110 10.4 0.001 0.001 10.355 10.393 sum_up_and_integrate 110 10.3 0.150 0.152 10.306 10.324 integrate_v_rspace 110 11.3 0.004 0.004 10.155 10.173 multiply_cannon_multrec 8220 15.4 4.389 4.601 9.629 9.725 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.175 9.700 mp_alltoall_i22 605 13.7 5.782 9.341 5.782 9.341 make_m2s 4110 13.4 0.043 0.043 8.627 9.233 make_images 4110 14.4 0.887 0.942 8.436 9.043 qs_ot_p2m_diag 48 11.0 0.151 0.155 8.069 8.075 pw_transfer 1331 11.6 0.076 0.077 8.052 8.068 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 7.933 7.948 cp_dbcsr_syevd 48 12.0 0.003 0.003 7.314 7.315 cp_fm_cholesky_invert 11 10.9 7.257 7.262 7.257 7.262 fft_wrap_pw1pw2_140 451 13.1 1.334 1.341 6.621 6.637 make_images_data 4110 15.4 0.040 0.044 5.205 6.583 init_scf_run 11 5.9 0.000 0.001 6.460 6.460 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.459 6.460 cp_fm_diag_elpa 48 13.0 0.000 0.000 6.267 6.268 cp_fm_diag_elpa_base 48 14.0 5.566 5.922 6.263 6.264 density_rs2pw 110 9.6 0.004 0.004 6.226 6.253 hybrid_alltoall_any 4261 16.3 0.257 0.551 4.803 6.094 fft3d_ps 1111 14.6 1.314 1.333 5.867 5.879 dbcsr_mm_accdrv_process 11614 15.7 3.340 3.689 5.095 5.351 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.779 5.271 apply_single 110 13.6 0.000 0.000 4.778 5.270 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.728 5.071 potential_pw2rs 110 12.3 0.021 0.022 4.662 4.680 rs_pw_transfer 902 11.9 0.010 0.010 4.607 4.668 ot_diis_step 99 11.5 0.015 0.015 4.643 4.645 mp_alltoall_z22v 1111 16.6 4.299 4.321 4.299 4.321 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.002 4.019 multiply_cannon_sync_h2d 8220 15.4 3.957 3.964 3.957 3.964 grid_integrate_task_list 110 12.3 3.901 3.962 3.901 3.962 wfi_extrapolate 11 7.9 0.001 0.001 3.877 3.877 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.726 3.727 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.251 3.715 mp_alltoall_d11v 2046 13.8 3.524 3.650 3.524 3.650 cp_fm_cholesky_decompose 22 10.9 3.483 3.512 3.483 3.512 calculate_dm_sparse 110 9.5 0.001 0.001 3.432 3.449 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.012 3.014 make_images_sizes 4110 15.4 0.004 0.005 1.402 2.986 mp_alltoall_i44 4110 16.4 1.397 2.982 1.397 2.982 grid_collocate_task_list 110 9.6 2.777 2.797 2.777 2.797 yz_to_x 561 15.1 0.110 0.112 2.520 2.537 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.425 2.428 calculate_first_density_matrix 1 7.0 0.000 0.000 2.417 2.418 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.254 2.310 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.228 2.231 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.178 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=108.079000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1196.272727, yerr=60.122739 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 627.408896E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456716E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459062E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762635464 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 57722. MP_Allreduce 11092 796. MP_Sync 86 MP_Alltoall 2210 3181886. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.067 205.605 205.606 qs_mol_dyn_low 1 2.0 0.003 0.006 205.081 205.111 qs_forces 11 3.9 0.005 0.010 204.987 204.989 qs_energies 11 4.9 0.004 0.021 199.312 199.328 scf_env_do_scf 11 5.9 0.001 0.002 182.687 182.691 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 161.915 161.917 dbcsr_multiply_generic 2485 12.5 0.179 0.183 123.221 123.808 qs_scf_new_mos 116 7.6 0.001 0.001 122.018 122.241 qs_scf_loop_do_ot 116 8.6 0.001 0.001 122.017 122.240 velocity_verlet 10 3.0 0.001 0.002 122.148 122.150 ot_scf_mini 116 9.6 0.003 0.004 115.469 115.707 multiply_cannon 2485 13.5 0.249 0.340 100.068 101.234 multiply_cannon_loop 2485 14.5 2.061 2.108 97.922 99.144 ot_mini 116 10.6 0.001 0.006 64.691 64.977 multiply_cannon_multrec 59640 15.5 32.990 34.717 41.170 42.972 qs_ot_get_derivative 116 11.6 0.001 0.003 40.293 40.526 rebuild_ks_matrix 127 8.3 0.001 0.001 33.713 33.924 qs_ks_build_kohn_sham_matrix 127 9.3 0.021 0.064 33.712 33.923 mp_waitall_1 264810 16.5 28.076 30.971 28.076 30.971 qs_ks_update_qs_env 127 7.6 0.001 0.001 30.184 30.360 qs_ot_get_p 127 10.4 0.001 0.001 28.835 29.144 multiply_cannon_sync_h2d 59640 15.5 27.173 28.843 27.173 28.843 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.059 24.517 apply_single 127 13.6 0.001 0.001 24.059 24.517 ot_diis_step 116 11.6 0.010 0.027 24.146 24.152 qs_ot_p2m_diag 82 11.4 0.078 0.091 22.451 22.522 init_scf_loop 11 6.9 0.000 0.001 20.698 20.699 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.855 19.856 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.085 19.280 multiply_cannon_metrocomm3 59640 15.5 0.109 0.113 15.648 17.548 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.971 17.006 cp_fm_diag_elpa_base 82 14.4 16.907 16.942 16.967 17.000 prepare_preconditioner 11 7.9 0.000 0.000 16.010 16.038 make_preconditioner 11 8.9 0.000 0.000 16.010 16.038 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.266 15.456 sum_up_and_integrate 127 10.3 0.090 0.115 14.354 14.371 integrate_v_rspace 127 11.3 0.004 0.005 14.264 14.286 make_m2s 4970 13.5 0.103 0.110 13.588 14.025 make_images 4970 14.5 0.396 0.414 13.408 13.853 qs_rho_update_rho_low 127 7.7 0.001 0.003 13.742 13.845 calculate_rho_elec 127 8.7 0.046 0.067 13.741 13.842 init_scf_run 11 5.9 0.000 0.001 12.396 12.396 scf_env_initial_rho_setup 11 6.9 0.001 0.013 12.395 12.396 density_rs2pw 127 9.7 0.006 0.007 6.960 10.359 mp_sum_l 7884 12.9 8.323 9.654 8.323 9.654 cp_fm_cholesky_invert 11 10.9 9.244 9.252 9.244 9.252 wfi_extrapolate 11 7.9 0.003 0.018 9.092 9.092 rs_pw_transfer 1038 11.9 0.016 0.019 5.674 9.047 calculate_dm_sparse 127 9.5 0.001 0.001 8.378 8.480 dbcsr_mm_accdrv_process 123452 16.2 3.156 3.254 7.750 8.262 qs_ot_get_orbitals 116 10.6 0.001 0.001 7.977 8.050 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.928 8.039 pw_transfer 1535 11.6 0.074 0.090 7.668 7.842 grid_integrate_task_list 127 12.3 7.205 7.685 7.205 7.685 make_images_data 4970 15.5 0.065 0.071 6.693 7.637 fft_wrap_pw1pw2 1281 12.7 0.011 0.013 7.467 7.635 multiply_cannon_metrocomm1 59640 15.5 0.084 0.089 6.021 7.324 hybrid_alltoall_any 5155 16.4 0.289 2.241 5.862 7.252 fft_wrap_pw1pw2_140 519 13.2 1.264 1.316 6.582 6.750 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.684 6.695 mp_waitany 15900 13.9 2.735 6.193 2.735 6.193 grid_collocate_task_list 127 9.7 4.686 5.940 4.686 5.940 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.811 5.881 mp_alltoall_d11v 2401 14.1 4.304 5.685 4.304 5.685 fft3d_ps 1281 14.7 2.127 2.760 5.251 5.560 rs_pw_transfer_RS2PW_140 138 11.5 0.278 0.297 2.175 5.550 cp_fm_cholesky_decompose 22 10.9 4.596 4.608 4.596 4.608 potential_pw2rs 127 12.3 0.009 0.010 4.522 4.545 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.606000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=594.636364, yerr=6.623275 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 2.200017E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6024768 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.7 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 836.427776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2427840 MPI messages size (bytes): total size 4.132588E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.702167E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 71532 2339635200 32768 < size <= 131072 729952 56049532928 131072 < size <= 4194304 1387568 1410045313024 4194304 < size <= 16777216 155760 1473828901424 16777216 < size 68112 1190343475200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4045 57613. MP_Allreduce 11190 956. MP_Sync 88 MP_Alltoall 1983 5620986. MP_SendRecv 12126 47072. MP_ISendRecv 12126 47072. MP_Wait 26114 MP_ISend 11836 212447. MP_IRecv 11836 212447. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.044 192.084 192.085 qs_mol_dyn_low 1 2.0 0.003 0.004 191.599 191.716 qs_forces 11 3.9 0.004 0.005 191.491 191.492 qs_energies 11 4.9 0.002 0.005 184.590 184.600 scf_env_do_scf 11 5.9 0.002 0.005 168.086 168.096 scf_env_do_scf_inner_loop 118 6.6 0.004 0.013 135.537 135.539 velocity_verlet 10 3.0 0.002 0.003 120.475 120.477 dbcsr_multiply_generic 2529 12.6 0.191 0.199 98.375 99.569 qs_scf_new_mos 118 7.6 0.001 0.001 94.513 94.918 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.512 94.917 ot_scf_mini 118 9.6 0.004 0.005 89.667 90.080 multiply_cannon 2529 13.6 0.481 0.530 78.264 82.611 multiply_cannon_loop 2529 14.6 1.258 1.293 74.868 77.224 ot_mini 118 10.6 0.001 0.001 50.153 50.624 mp_waitall_1 216598 16.6 24.649 37.715 24.649 37.715 multiply_cannon_multrec 30348 15.6 22.345 26.960 32.017 36.873 rebuild_ks_matrix 129 8.3 0.001 0.001 34.111 34.500 qs_ks_build_kohn_sham_matrix 129 9.3 0.019 0.023 34.111 34.499 init_scf_loop 11 6.9 0.001 0.002 32.461 32.463 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.664 31.039 multiply_cannon_metrocomm3 30348 15.6 0.091 0.095 15.716 28.743 qs_ot_get_derivative 118 11.6 0.001 0.002 28.100 28.531 prepare_preconditioner 11 7.9 0.000 0.000 27.976 28.034 make_preconditioner 11 8.9 0.000 0.003 27.976 28.034 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.682 27.202 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 22.101 23.348 apply_single 129 13.6 0.001 0.001 22.101 23.348 ot_diis_step 118 11.6 0.015 0.016 21.874 21.876 qs_ot_get_p 129 10.4 0.001 0.001 21.076 21.608 multiply_cannon_sync_h2d 30348 15.6 19.538 21.590 19.538 21.590 cp_fm_cholesky_invert 11 10.9 16.318 16.330 16.318 16.330 qs_ot_p2m_diag 84 11.4 0.189 0.218 16.219 16.252 make_m2s 5058 13.6 0.090 0.096 14.420 16.086 make_images 5058 14.6 1.152 1.349 14.207 15.872 sum_up_and_integrate 129 10.3 0.123 0.143 15.167 15.199 integrate_v_rspace 129 11.3 0.004 0.004 15.044 15.082 cp_dbcsr_syevd 84 12.4 0.006 0.006 15.053 15.054 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.413 14.441 calculate_rho_elec 129 8.7 0.088 0.105 14.412 14.440 cp_fm_diag_elpa 84 13.4 0.000 0.000 11.977 12.009 cp_fm_diag_elpa_base 84 14.4 11.713 11.786 11.971 11.998 init_scf_run 11 5.9 0.000 0.001 11.684 11.686 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.684 11.685 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.193 11.487 make_images_data 5058 15.6 0.065 0.074 8.755 10.760 multiply_cannon_metrocomm4 27819 15.6 0.104 0.118 3.756 10.308 hybrid_alltoall_any 5245 16.5 0.343 1.504 7.476 9.999 mp_irecv_dv 70084 16.3 3.554 9.918 3.554 9.918 density_rs2pw 129 9.7 0.006 0.007 7.587 9.808 dbcsr_mm_accdrv_process 62758 16.2 4.633 5.251 9.124 9.707 pw_transfer 1559 11.6 0.088 0.104 8.903 8.965 fft_wrap_pw1pw2 1301 12.7 0.011 0.012 8.676 8.737 wfi_extrapolate 11 7.9 0.001 0.001 8.416 8.416 rs_pw_transfer 1054 12.0 0.015 0.018 5.863 8.110 grid_integrate_task_list 129 12.3 7.416 7.874 7.416 7.874 fft_wrap_pw1pw2_140 527 13.2 1.344 1.364 7.661 7.746 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.267 7.031 cp_fm_cholesky_decompose 22 10.9 6.829 6.916 6.829 6.916 calculate_dm_sparse 129 9.5 0.001 0.001 6.514 6.649 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.172 6.181 grid_collocate_task_list 129 9.7 4.867 6.046 4.867 6.046 fft3d_ps 1301 14.7 2.830 3.019 5.966 6.025 mp_sum_l 8016 12.9 3.894 5.886 3.894 5.886 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.338 5.450 qs_ot_get_orbitals 118 10.6 0.001 0.001 5.366 5.435 mp_alltoall_d11v 2429 14.1 4.125 5.417 4.125 5.417 mp_allgather_i34 2529 14.6 1.943 5.148 1.943 5.148 potential_pw2rs 129 12.3 0.016 0.019 5.022 5.043 mp_waitany 11836 13.9 2.597 4.879 2.597 4.879 rs_pw_transfer_RS2PW_140 140 11.5 0.352 0.379 2.175 4.403 mp_sum_d 4510 12.1 2.630 4.024 2.630 4.024 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.847 3.880 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=192.085000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.090909, yerr=3.895325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 938.135552E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.049 175.505 175.524 qs_mol_dyn_low 1 2.0 0.003 0.003 174.931 174.944 qs_forces 11 3.9 0.004 0.005 174.833 174.838 qs_energies 11 4.9 0.004 0.007 168.128 168.136 scf_env_do_scf 11 5.9 0.001 0.002 152.301 152.302 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 117.515 117.515 velocity_verlet 10 3.0 0.001 0.002 112.717 112.719 dbcsr_multiply_generic 2507 12.6 0.181 0.188 81.304 82.438 qs_scf_new_mos 117 7.6 0.001 0.001 79.071 79.404 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.070 79.403 ot_scf_mini 117 9.6 0.003 0.004 74.913 75.295 multiply_cannon 2507 13.6 0.562 0.582 61.827 66.402 multiply_cannon_loop 2507 14.6 0.856 0.885 58.672 61.170 ot_mini 117 10.6 0.001 0.001 42.210 42.598 init_scf_loop 11 6.9 0.000 0.000 34.688 34.689 mp_waitall_1 170520 16.6 25.141 34.531 25.141 34.531 rebuild_ks_matrix 128 8.3 0.001 0.001 31.030 31.474 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.029 31.473 prepare_preconditioner 11 7.9 0.000 0.000 30.631 30.692 make_preconditioner 11 8.9 0.000 0.000 30.630 30.692 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.186 29.681 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.874 28.288 multiply_cannon_multrec 20056 15.6 13.380 16.287 22.046 24.961 multiply_cannon_metrocomm3 20056 15.6 0.057 0.061 15.494 24.481 qs_ot_get_derivative 117 11.6 0.001 0.002 22.375 22.754 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.989 21.092 apply_single 128 13.6 0.001 0.001 19.989 21.091 ot_diis_step 117 11.6 0.018 0.018 19.731 19.731 qs_ot_get_p 128 10.4 0.001 0.001 18.071 18.526 multiply_cannon_sync_h2d 20056 15.6 14.193 15.917 14.193 15.917 make_m2s 5014 13.6 0.080 0.087 14.635 15.653 make_images 5014 14.6 1.166 1.264 14.401 15.418 sum_up_and_integrate 128 10.3 0.133 0.144 14.853 14.885 integrate_v_rspace 128 11.3 0.004 0.005 14.720 14.749 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.399 14.434 calculate_rho_elec 128 8.7 0.130 0.145 14.398 14.434 cp_fm_cholesky_invert 11 10.9 14.210 14.218 14.210 14.218 qs_ot_p2m_diag 83 11.4 0.265 0.272 13.994 14.001 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.001 13.003 init_scf_run 11 5.9 0.000 0.001 10.808 10.808 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.808 10.808 make_images_data 5014 15.6 0.058 0.067 9.070 10.432 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.862 9.879 cp_fm_diag_elpa_base 83 14.4 9.453 9.603 9.859 9.876 hybrid_alltoall_any 5200 16.5 0.435 1.985 7.778 9.576 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.490 9.450 density_rs2pw 128 9.7 0.006 0.006 7.153 9.333 mp_irecv_dv 50230 16.2 3.368 9.201 3.368 9.201 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.749 9.037 pw_transfer 1547 11.6 0.086 0.103 8.836 8.935 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.612 8.717 dbcsr_mm_accdrv_process 41502 16.2 4.477 5.284 8.120 8.241 grid_integrate_task_list 128 12.3 7.504 8.064 7.504 8.064 fft_wrap_pw1pw2_140 523 13.2 1.412 1.431 7.653 7.765 wfi_extrapolate 11 7.9 0.001 0.001 7.445 7.445 cp_fm_upper_to_full 105 14.8 5.740 7.339 5.740 7.339 rs_pw_transfer 1046 11.9 0.014 0.015 5.117 7.292 cp_fm_cholesky_decompose 22 10.9 7.200 7.230 7.200 7.230 dbcsr_complete_redistribute 395 12.7 1.205 1.233 4.691 6.482 grid_collocate_task_list 128 9.7 5.013 5.923 5.013 5.923 calculate_dm_sparse 128 9.5 0.001 0.001 5.763 5.876 fft3d_ps 1291 14.7 2.741 2.965 5.736 5.824 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.541 5.550 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.525 5.317 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.470 5.164 mp_alltoall_d11v 2415 14.1 4.395 5.099 4.395 5.099 mp_allgather_i34 2507 14.6 1.595 4.857 1.595 4.857 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.644 4.772 potential_pw2rs 128 12.3 0.021 0.024 4.678 4.692 mp_waitany 11748 13.9 2.370 4.634 2.370 4.634 mp_sum_l 7950 12.9 3.020 4.464 3.020 4.464 transfer_fm_to_dbcsr 11 9.9 0.020 0.033 2.425 4.202 rs_pw_transfer_RS2PW_140 139 11.5 0.328 0.352 1.937 4.106 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.009 4.031 mp_alltoall_i22 716 14.1 1.989 3.912 1.989 3.912 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.745 3.745 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.654 3.699 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.524000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.181818, yerr=8.962733 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 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 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 4.320341E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 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 1.151062E+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 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3979 58540. MP_Allreduce 11025 1085. MP_Sync 86 MP_Alltoall 1700 12496367. 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 0.061 0.132 187.889 187.891 qs_mol_dyn_low 1 2.0 0.003 0.004 186.867 186.889 qs_forces 11 3.9 0.007 0.015 186.751 186.756 qs_energies 11 4.9 0.021 0.077 179.471 179.481 scf_env_do_scf 11 5.9 0.001 0.002 161.543 161.551 velocity_verlet 10 3.0 0.001 0.002 123.575 123.578 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 115.424 115.425 dbcsr_multiply_generic 2485 12.5 0.187 0.193 78.525 79.211 qs_scf_new_mos 116 7.6 0.001 0.001 78.091 78.389 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.091 78.388 ot_scf_mini 116 9.6 0.003 0.005 73.653 73.957 multiply_cannon 2485 13.5 0.558 0.597 54.159 57.106 multiply_cannon_loop 2485 14.5 1.171 1.201 50.571 52.139 init_scf_loop 11 6.9 0.000 0.002 46.000 46.001 ot_mini 116 10.6 0.001 0.001 41.627 41.934 prepare_preconditioner 11 7.9 0.000 0.000 41.887 41.909 make_preconditioner 11 8.9 0.001 0.010 41.887 41.909 make_full_inverse_cholesky 11 9.9 0.013 0.021 35.488 40.495 multiply_cannon_multrec 29820 15.5 14.045 19.583 25.911 30.900 rebuild_ks_matrix 127 8.3 0.001 0.001 29.929 30.163 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 29.928 30.162 mp_waitall_1 146592 16.7 17.101 27.756 17.101 27.756 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.872 27.079 qs_ot_get_derivative 116 11.6 0.001 0.002 22.146 22.451 make_m2s 4970 13.5 0.095 0.103 20.014 21.083 make_images 4970 14.5 1.921 2.244 19.710 20.776 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.929 19.439 apply_single 127 13.6 0.001 0.001 18.929 19.439 ot_diis_step 116 11.6 0.017 0.018 19.359 19.361 qs_ot_get_p 127 10.4 0.002 0.006 18.370 18.688 cp_fm_upper_to_full 104 14.8 11.421 16.820 11.421 16.820 cp_fm_cholesky_invert 11 10.9 15.916 15.925 15.916 15.925 multiply_cannon_metrocomm3 29820 15.5 0.047 0.050 6.434 15.312 sum_up_and_integrate 127 10.3 0.139 0.152 14.848 14.876 integrate_v_rspace 127 11.3 0.004 0.005 14.709 14.740 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.520 14.553 calculate_rho_elec 127 8.7 0.172 0.188 14.519 14.552 qs_ot_p2m_diag 82 11.4 0.338 0.385 14.431 14.483 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.185 13.186 dbcsr_complete_redistribute 393 12.7 1.571 1.722 9.077 12.808 make_images_data 4970 15.5 0.061 0.066 10.804 12.596 multiply_cannon_sync_h2d 29820 15.5 11.594 12.460 11.594 12.460 dbcsr_mm_accdrv_process 61748 16.2 7.495 8.242 11.441 11.990 hybrid_alltoall_any 5155 16.4 0.522 2.197 9.757 11.642 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.715 11.473 init_scf_run 11 5.9 0.000 0.001 11.242 11.243 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.242 11.243 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.159 10.171 cp_fm_diag_elpa_base 82 14.4 9.215 9.523 10.153 10.164 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.374 10.083 qs_ot_get_derivative_diag 76 12.4 0.002 0.005 9.097 9.315 mp_alltoall_i22 712 14.1 5.575 9.312 5.575 9.312 pw_transfer 1535 11.6 0.088 0.110 9.100 9.191 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.873 8.967 density_rs2pw 127 9.7 0.006 0.006 6.902 8.189 grid_integrate_task_list 127 12.3 7.669 8.065 7.669 8.065 fft_wrap_pw1pw2_140 519 13.2 1.548 1.580 7.913 8.016 wfi_extrapolate 11 7.9 0.001 0.001 7.826 7.826 cp_fm_cholesky_decompose 22 10.9 7.416 7.521 7.416 7.521 multiply_cannon_metrocomm4 24850 15.5 0.079 0.091 2.748 7.178 mp_irecv_dv 75445 16.2 2.602 6.908 2.602 6.908 rs_pw_transfer 1038 11.9 0.014 0.017 5.048 6.478 calculate_dm_sparse 127 9.5 0.001 0.001 6.209 6.296 fft3d_ps 1281 14.7 2.830 2.935 5.817 5.904 grid_collocate_task_list 127 9.7 5.159 5.843 5.159 5.843 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.697 5.750 mp_alltoall_d11v 2401 14.1 4.871 5.619 4.871 5.619 qs_energies_init_hamiltonians 11 5.9 0.004 0.007 5.122 5.126 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.509 4.605 potential_pw2rs 127 12.3 0.023 0.023 4.464 4.481 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.304 4.398 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.127 4.174 mp_waitany 14840 13.9 2.682 4.128 2.682 4.128 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=187.891000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1086.363636, yerr=19.891856 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 5.910926E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1979136 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.2 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 1.525584E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 243360 MPI messages size (bytes): total size 1.342108E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.514910E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 115584 60599304192 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592540656 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9028 51. MP_Alltoall 9736 793528. MP_ISend 40516 2095952. MP_IRecv 40516 2095058. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 57658. MP_Allreduce 11179 1163. MP_Sync 88 MP_Alltoall 1724 18848050. 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 0.015 0.031 175.941 175.942 qs_mol_dyn_low 1 2.0 0.003 0.003 175.435 175.448 qs_forces 11 3.9 0.006 0.007 175.110 175.116 qs_energies 11 4.9 0.001 0.002 167.429 167.436 scf_env_do_scf 11 5.9 0.001 0.001 149.514 149.521 velocity_verlet 10 3.0 0.001 0.001 115.355 115.373 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 113.900 113.901 dbcsr_multiply_generic 2535 12.6 0.209 0.224 75.246 75.645 qs_scf_new_mos 118 7.6 0.001 0.001 74.617 74.693 qs_scf_loop_do_ot 118 8.6 0.001 0.001 74.616 74.693 ot_scf_mini 118 9.6 0.004 0.004 70.129 70.194 multiply_cannon 2535 13.6 0.617 0.655 55.026 59.576 multiply_cannon_loop 2535 14.6 0.449 0.460 50.326 51.126 ot_mini 118 10.6 0.001 0.001 39.856 39.916 init_scf_loop 11 6.9 0.000 0.000 35.469 35.470 mp_waitall_1 127164 16.7 26.456 33.743 26.456 33.743 prepare_preconditioner 11 7.9 0.000 0.000 31.530 31.549 make_preconditioner 11 8.9 0.000 0.000 31.530 31.549 rebuild_ks_matrix 129 8.3 0.001 0.001 30.412 30.496 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 30.412 30.496 make_full_inverse_cholesky 11 9.9 0.017 0.025 29.402 29.690 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.565 27.640 multiply_cannon_multrec 10140 15.6 10.477 14.202 17.803 20.621 ot_diis_step 118 11.6 0.020 0.021 20.082 20.083 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.644 19.899 apply_single 129 13.6 0.001 0.001 19.643 19.898 qs_ot_get_derivative 118 11.6 0.001 0.002 19.702 19.759 multiply_cannon_metrocomm3 10140 15.6 0.024 0.025 12.407 19.455 make_m2s 5070 13.6 0.068 0.071 16.595 18.823 make_images 5070 14.6 2.316 2.798 16.284 18.517 cp_fm_cholesky_invert 11 10.9 17.928 17.934 17.928 17.934 qs_ot_get_p 129 10.4 0.001 0.001 17.120 17.193 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.614 15.654 calculate_rho_elec 129 8.7 0.258 0.269 15.614 15.654 sum_up_and_integrate 129 10.3 0.183 0.193 15.364 15.417 integrate_v_rspace 129 11.3 0.004 0.004 15.181 15.242 qs_ot_p2m_diag 84 11.4 0.501 0.506 13.288 13.308 make_images_data 5070 15.6 0.051 0.060 10.178 12.857 multiply_cannon_sync_h2d 10140 15.6 11.744 12.180 11.744 12.180 cp_dbcsr_syevd 84 12.4 0.005 0.006 12.129 12.131 hybrid_alltoall_any 5257 16.5 0.848 3.831 9.977 12.098 init_scf_run 11 5.9 0.000 0.001 10.936 10.937 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.936 10.936 pw_transfer 1559 11.6 0.088 0.099 9.566 9.590 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.338 9.369 cp_fm_diag_elpa 84 13.4 0.000 0.000 9.089 9.103 cp_fm_diag_elpa_base 84 14.4 8.831 8.925 9.079 9.092 grid_integrate_task_list 129 12.3 8.068 8.432 8.068 8.432 fft_wrap_pw1pw2_140 527 13.2 1.926 1.956 8.264 8.297 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 7.997 8.049 cp_fm_cholesky_decompose 22 10.9 7.874 8.001 7.874 8.001 density_rs2pw 129 9.7 0.006 0.006 6.927 7.857 multiply_cannon_metrocomm1 10140 15.6 0.028 0.029 4.658 7.629 wfi_extrapolate 11 7.9 0.001 0.001 7.574 7.574 dbcsr_mm_accdrv_process 20958 16.1 2.623 3.497 6.950 7.556 mp_allgather_i34 2535 14.6 2.855 7.407 2.855 7.407 calculate_dm_sparse 129 9.5 0.001 0.001 6.164 6.258 grid_collocate_task_list 129 9.7 5.554 6.258 5.554 6.258 mp_alltoall_d11v 2429 14.1 5.143 5.979 5.143 5.979 fft3d_ps 1301 14.7 2.780 2.857 5.790 5.821 dbcsr_complete_redistribute 397 12.7 2.128 2.182 5.198 5.478 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.431 5.452 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.420 5.420 rs_pw_transfer 1054 12.0 0.013 0.014 4.121 5.066 potential_pw2rs 129 12.3 0.027 0.028 4.432 4.444 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.350 4.380 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.563 3.880 multiply_cannon_metrocomm4 7605 15.6 0.024 0.027 1.868 3.843 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.743 3.789 mp_irecv_dv 29150 15.9 1.830 3.771 1.830 3.771 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.762 3.771 copy_fm_to_dbcsr 210 11.7 0.001 0.002 3.404 3.716 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.533 3.590 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=175.942000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1425.181818, yerr=52.356849 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 11.696233E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 1964048 0.0% 0.0% 100.0% average stack size 0.0 0.0 3439.8 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 3.139662E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 100280 MPI messages size (bytes): total size 1.136195E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.330227E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45208 35089547264 4194304 < size <= 16777216 44352 379752284160 16777216 < size 10104 721350092304 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 59285. MP_Allreduce 11057 1504. MP_Sync 87 MP_Alltoall 1712 36974159. MP_SendRecv 1792 218624. MP_ISendRecv 1792 218624. MP_Wait 9802 MP_ISend 6408 1080322. MP_IRecv 6408 1080322. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.077 311.060 311.060 qs_mol_dyn_low 1 2.0 0.003 0.003 310.426 310.436 qs_forces 11 3.9 0.004 0.005 310.318 310.324 qs_energies 11 4.9 0.002 0.002 300.778 300.787 scf_env_do_scf 11 5.9 0.001 0.001 277.881 277.890 velocity_verlet 10 3.0 0.001 0.001 220.303 220.311 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 149.902 149.904 init_scf_loop 11 6.9 0.000 0.000 127.723 127.724 prepare_preconditioner 11 7.9 0.000 0.000 122.548 122.576 make_preconditioner 11 8.9 0.000 0.000 122.548 122.576 make_full_inverse_cholesky 11 9.9 0.036 0.039 98.514 119.644 qs_scf_new_mos 117 7.6 0.001 0.001 97.491 97.604 qs_scf_loop_do_ot 117 8.6 0.001 0.001 97.490 97.604 ot_scf_mini 117 9.6 0.004 0.004 92.520 92.578 dbcsr_multiply_generic 2507 12.6 0.213 0.226 87.229 87.915 cp_fm_upper_to_full 105 14.8 53.139 76.150 53.139 76.150 multiply_cannon 2507 13.6 0.715 0.769 61.117 61.772 multiply_cannon_loop 2507 14.6 0.472 0.478 57.162 58.907 ot_mini 117 10.6 0.001 0.001 46.993 47.065 dbcsr_complete_redistribute 395 12.7 3.986 4.025 29.860 42.901 rebuild_ks_matrix 128 8.3 0.001 0.001 39.501 39.611 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 39.501 39.611 copy_fm_to_dbcsr 209 11.7 0.001 0.002 26.356 39.448 transfer_fm_to_dbcsr 11 9.9 0.026 0.033 23.988 36.897 qs_ks_update_qs_env 128 7.6 0.001 0.001 36.422 36.528 cp_fm_cholesky_invert 11 10.9 35.749 35.756 35.749 35.756 mp_waitall_1 103674 16.8 31.288 35.480 31.288 35.480 mp_alltoall_i22 716 14.1 21.820 35.098 21.820 35.098 qs_ot_get_p 128 10.4 0.001 0.001 29.368 29.423 qs_ot_get_derivative 117 11.6 0.002 0.002 25.672 25.734 qs_ot_p2m_diag 83 11.4 0.877 0.882 24.870 24.909 cp_dbcsr_syevd 83 12.4 0.006 0.006 23.080 23.081 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 20.847 22.485 make_m2s 5014 13.6 0.076 0.079 20.993 22.125 make_images 5014 14.6 3.731 3.869 20.511 21.649 ot_diis_step 117 11.6 0.022 0.023 21.273 21.274 qs_rho_update_rho_low 128 7.7 0.001 0.001 21.200 21.211 calculate_rho_elec 128 8.7 0.481 0.482 21.199 21.210 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.424 20.708 apply_single 128 13.6 0.001 0.001 20.424 20.708 sum_up_and_integrate 128 10.3 0.321 0.323 20.247 20.333 integrate_v_rspace 128 11.3 0.004 0.004 19.926 20.010 cp_fm_diag_elpa 83 13.4 0.000 0.000 19.679 19.680 cp_fm_diag_elpa_base 83 14.4 15.210 16.898 19.671 19.672 multiply_cannon_multrec 10028 15.6 10.539 12.369 18.014 18.126 multiply_cannon_sync_h2d 10028 15.6 15.661 15.670 15.661 15.670 hybrid_alltoall_any 5200 16.5 1.302 3.038 11.572 13.640 make_images_data 5014 15.6 0.058 0.063 11.512 13.552 init_scf_run 11 5.9 0.000 0.001 12.773 12.774 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.773 12.773 pw_transfer 1547 11.6 0.095 0.095 12.067 12.075 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 11.829 11.837 fft_wrap_pw1pw2_140 523 13.2 3.076 3.104 10.383 10.394 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.079 10.128 mp_alltoall_d11v 2415 14.1 8.844 9.572 8.844 9.572 cp_fm_cholesky_decompose 22 10.9 9.460 9.565 9.460 9.565 wfi_extrapolate 11 7.9 0.001 0.001 9.328 9.328 dbcsr_mm_accdrv_process 20762 16.1 3.724 5.671 7.234 9.057 grid_integrate_task_list 128 12.3 8.836 9.036 8.836 9.036 density_rs2pw 128 9.7 0.005 0.006 8.519 8.664 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 8.077 8.078 fft3d_ps 1291 14.7 2.855 2.884 6.898 6.940 calculate_dm_sparse 128 9.5 0.001 0.001 6.868 6.938 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.521 6.567 grid_collocate_task_list 128 9.7 6.483 6.540 6.483 6.540 rs_scatter_matrices 139 9.7 3.642 4.599 6.147 6.386 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.258 6.376 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=311.060000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2853.000000, yerr=161.030771 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 1.255690E+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 39803024. 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 0.035 0.044 85.439 85.440 qs_energies 1 2.0 0.000 0.000 84.948 84.961 ls_scf 1 3.0 0.003 0.021 84.028 84.041 dbcsr_multiply_generic 111 6.7 0.014 0.016 72.904 73.085 multiply_cannon 111 7.7 0.018 0.020 55.925 57.183 multiply_cannon_loop 111 8.7 0.205 0.221 52.456 53.928 ls_scf_main 1 4.0 0.001 0.009 52.290 52.291 density_matrix_trs4 2 5.0 0.002 0.003 46.726 46.801 ls_scf_init_scf 1 4.0 0.000 0.001 28.669 28.671 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.536 27.593 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.426 25.448 mp_waitall_1 11031 10.9 22.718 25.229 22.718 25.229 multiply_cannon_multrec 2664 9.7 8.166 8.906 15.461 17.176 multiply_cannon_sync_h2d 2664 9.7 13.536 15.443 13.536 15.443 make_m2s 222 7.7 0.008 0.010 13.303 13.791 make_images 222 8.7 0.099 0.110 13.281 13.769 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.695 12.544 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.570 9.680 make_images_data 222 9.7 0.004 0.005 7.792 8.412 hybrid_alltoall_any 227 10.6 0.216 1.839 6.672 8.128 dbcsr_mm_accdrv_process 4760 10.4 0.506 0.644 6.914 7.880 dbcsr_mm_accdrv_process_sort 4760 11.4 6.207 7.099 6.207 7.099 calculate_norms 4752 9.8 5.509 6.137 5.509 6.137 mp_sum_l 887 5.1 3.142 5.247 3.142 5.247 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.051 5.195 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.326 3.925 make_images_sizes 222 9.7 0.000 0.000 0.845 3.804 mp_alltoall_i44 222 10.7 0.844 3.803 0.844 3.803 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.062 3.746 mp_irecv_dv 6231 10.9 2.046 3.722 2.046 3.722 arnoldi_extremal 4 6.8 0.000 0.000 3.328 3.356 arnoldi_normal_ev 4 7.8 0.001 0.002 3.328 3.356 build_subspace 16 8.4 0.009 0.012 3.220 3.222 ls_scf_post 1 4.0 0.000 0.001 3.066 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.878 2.931 dbcsr_special_finalize 555 9.7 0.005 0.006 2.308 2.855 dbcsr_merge_single_wm 555 10.7 0.466 0.582 2.300 2.845 make_images_pack 222 9.7 2.212 2.614 2.214 2.615 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.319 2.555 dbcsr_sort_data 658 11.4 2.090 2.551 2.090 2.551 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.448 2.065 2.450 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.330 2.388 buffer_matrices_ensure_size 222 8.7 1.775 2.091 1.775 2.091 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.816 1.818 rebuild_ks_matrix 3 7.3 0.000 0.000 1.806 1.809 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.017 1.806 1.809 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.440000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1124.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 2.098549E+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 0.042 98.150 98.150 qs_energies 1 2.0 0.000 0.000 97.585 97.591 ls_scf 1 3.0 0.001 0.009 96.271 96.276 dbcsr_multiply_generic 111 6.7 0.016 0.018 82.170 82.591 multiply_cannon 111 7.7 0.029 0.040 60.035 64.161 multiply_cannon_loop 111 8.7 0.116 0.122 56.738 60.404 ls_scf_main 1 4.0 0.002 0.038 54.795 54.803 density_matrix_trs4 2 5.0 0.002 0.003 49.036 49.233 ls_scf_init_scf 1 4.0 0.000 0.002 37.931 37.932 ls_scf_init_matrix_S 1 5.0 0.000 0.000 36.712 36.827 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 34.292 34.305 mp_waitall_1 9105 10.9 21.695 32.154 21.695 32.154 multiply_cannon_multrec 1332 9.7 13.233 16.752 22.468 27.109 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.791 21.124 make_m2s 222 7.7 0.006 0.007 15.492 16.190 make_images 222 8.7 1.576 1.932 15.462 16.161 dbcsr_mm_accdrv_process 4041 10.4 0.264 0.452 8.833 10.364 dbcsr_mm_accdrv_process_sort 4041 11.4 8.424 9.927 8.424 9.927 make_images_data 222 9.7 0.004 0.004 8.894 9.807 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.640 9.372 hybrid_alltoall_any 227 10.6 0.519 2.436 8.241 9.216 mp_sum_l 887 5.1 5.349 8.356 5.349 8.356 compute_matrix_preconditioner 1 6.0 0.000 0.002 7.876 7.882 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.215 7.625 mp_irecv_dv 3311 11.0 3.195 7.560 3.195 7.560 acc_transpose_blocks 1332 9.7 0.007 0.007 6.434 7.339 acc_transpose_blocks_kernels 1332 10.7 0.018 0.019 6.387 7.291 jit_kernel_transpose 1 13.0 6.369 7.274 6.369 7.274 calculate_norms 2376 9.8 6.009 6.690 6.009 6.690 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.133 6.672 multiply_cannon_sync_h2d 1332 9.7 4.816 6.405 4.816 6.405 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.034 5.205 arnoldi_extremal 4 6.8 0.000 0.000 4.656 4.676 arnoldi_normal_ev 4 7.8 0.001 0.005 4.656 4.676 build_subspace 16 8.4 0.014 0.021 4.376 4.383 ls_scf_post 1 4.0 0.000 0.001 3.544 3.550 ls_scf_store_result 1 5.0 0.000 0.000 3.232 3.367 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.125 3.357 dbcsr_matrix_vector_mult_local 304 10.0 2.738 3.221 2.740 3.223 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.648 2.739 mp_allgather_i34 111 8.7 0.989 2.530 0.989 2.530 make_images_pack 222 9.7 2.023 2.431 2.025 2.433 dbcsr_sort_data 436 11.2 1.837 2.109 1.837 2.109 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=98.150000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1761.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 2.798834E+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 0.060 0.094 101.578 101.579 qs_energies 1 2.0 0.000 0.000 100.878 100.899 ls_scf 1 3.0 0.000 0.001 99.458 99.479 dbcsr_multiply_generic 111 6.7 0.016 0.019 84.017 84.363 multiply_cannon 111 7.7 0.051 0.111 59.797 63.820 multiply_cannon_loop 111 8.7 0.100 0.104 56.269 60.611 ls_scf_main 1 4.0 0.001 0.025 58.120 58.125 density_matrix_trs4 2 5.0 0.002 0.005 52.089 52.273 ls_scf_init_scf 1 4.0 0.001 0.005 37.788 37.791 ls_scf_init_matrix_S 1 5.0 0.000 0.002 36.550 36.629 mp_waitall_1 7281 11.0 24.864 34.383 24.864 34.383 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 34.129 34.143 multiply_cannon_multrec 888 9.7 12.600 15.336 21.136 24.348 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.821 23.610 make_m2s 222 7.7 0.006 0.007 17.239 18.465 make_images 222 8.7 1.984 2.308 17.201 18.426 hybrid_alltoall_any 227 10.6 0.622 2.875 9.557 10.897 make_images_data 222 9.7 0.003 0.004 9.859 10.860 mp_sum_l 887 5.1 5.452 9.381 5.452 9.381 dbcsr_mm_accdrv_process 3754 10.4 0.248 0.419 8.063 9.298 dbcsr_mm_accdrv_process_sort 3754 11.4 7.689 8.882 7.689 8.882 compute_matrix_preconditioner 1 6.0 0.000 0.001 7.498 7.503 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.234 7.496 multiply_cannon_sync_h2d 888 9.7 6.002 7.344 6.002 7.344 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.492 7.057 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.810 7.041 mp_irecv_dv 2335 11.1 2.477 7.014 2.477 7.014 acc_transpose_blocks 888 9.7 0.004 0.005 6.277 6.920 acc_transpose_blocks_kernels 888 10.7 0.012 0.013 6.237 6.879 jit_kernel_transpose 1 13.0 6.225 6.868 6.225 6.868 arnoldi_extremal 4 6.8 0.000 0.000 5.120 5.140 arnoldi_normal_ev 4 7.8 0.001 0.004 5.120 5.140 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.934 5.127 build_subspace 16 8.4 0.014 0.020 4.808 4.820 calculate_norms 1584 9.8 4.304 4.651 4.304 4.651 mp_allgather_i34 111 8.7 1.366 3.788 1.366 3.788 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.432 3.767 dbcsr_matrix_vector_mult_local 304 10.0 3.015 3.593 3.017 3.595 ls_scf_post 1 4.0 0.003 0.021 3.549 3.571 ls_scf_store_result 1 5.0 0.000 0.000 3.296 3.368 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.897 3.010 make_images_sizes 222 9.7 0.000 0.000 1.024 2.255 mp_alltoall_i44 222 10.7 1.023 2.255 1.023 2.255 dbcsr_sort_data 325 11.1 1.896 2.140 1.896 2.140 make_images_pack 222 9.7 1.823 2.120 1.826 2.122 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=101.579000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2191.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 3.354051E+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 0.058 0.082 104.670 104.671 qs_energies 1 2.0 0.000 0.000 103.926 103.931 ls_scf 1 3.0 0.000 0.000 102.286 102.293 dbcsr_multiply_generic 111 6.7 0.017 0.017 85.953 86.274 multiply_cannon 111 7.7 0.053 0.100 58.864 63.031 ls_scf_main 1 4.0 0.000 0.000 59.190 59.191 multiply_cannon_loop 111 8.7 0.114 0.125 53.875 57.144 density_matrix_trs4 2 5.0 0.002 0.003 53.040 53.176 ls_scf_init_scf 1 4.0 0.001 0.003 39.858 39.859 ls_scf_init_matrix_S 1 5.0 0.000 0.000 38.605 38.694 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 36.150 36.161 mp_waitall_1 6369 11.0 24.020 35.127 24.020 35.127 multiply_cannon_multrec 1332 9.7 14.162 17.071 22.019 24.583 make_m2s 222 7.7 0.006 0.008 21.159 22.546 make_images 222 8.7 3.145 3.611 21.109 22.497 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 10.452 21.166 make_images_data 222 9.7 0.004 0.004 11.834 13.386 hybrid_alltoall_any 227 10.6 0.796 3.765 11.252 12.912 dbcsr_mm_accdrv_process 3641 10.4 0.212 0.413 7.492 9.024 dbcsr_mm_accdrv_process_sort 3641 11.4 7.116 8.610 7.116 8.610 mp_sum_l 887 5.1 4.347 8.459 4.347 8.459 compute_matrix_preconditioner 1 6.0 0.001 0.001 8.027 8.033 acc_transpose_blocks 1332 9.7 0.007 0.007 6.286 7.521 acc_transpose_blocks_kernels 1332 10.7 0.016 0.018 6.226 7.463 jit_kernel_transpose 1 13.0 6.210 7.446 6.210 7.446 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.394 6.778 multiply_cannon_sync_h2d 1332 9.7 5.547 6.196 5.547 6.196 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.082 5.948 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.562 5.925 mp_irecv_dv 3229 10.9 2.059 5.885 2.059 5.885 arnoldi_extremal 4 6.8 0.000 0.000 5.230 5.245 arnoldi_normal_ev 4 7.8 0.001 0.005 5.230 5.245 build_subspace 16 8.4 0.014 0.021 4.860 4.866 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.534 4.749 calculate_norms 2376 9.8 4.195 4.517 4.195 4.517 mp_allgather_i34 111 8.7 2.111 4.327 2.111 4.327 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.577 3.875 dbcsr_matrix_vector_mult_local 304 10.0 3.179 3.682 3.182 3.683 dbcsr_sort_data 658 11.4 3.099 3.366 3.099 3.366 ls_scf_post 1 4.0 0.000 0.000 3.238 3.243 dbcsr_special_finalize 555 9.7 0.006 0.007 2.841 3.151 dbcsr_merge_single_wm 555 10.7 0.536 0.656 2.832 3.143 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.050 3.109 ls_scf_store_result 1 5.0 0.000 0.000 2.985 3.039 dbcsr_data_release 10477 10.7 1.586 2.453 1.586 2.453 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=104.671000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2705.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 4.672840E+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 0.044 0.056 106.921 106.921 qs_energies 1 2.0 0.000 0.000 106.100 106.103 ls_scf 1 3.0 0.000 0.000 104.172 104.178 dbcsr_multiply_generic 111 6.7 0.017 0.018 85.325 85.570 multiply_cannon 111 7.7 0.090 0.199 62.542 67.708 ls_scf_main 1 4.0 0.000 0.000 61.983 61.983 multiply_cannon_loop 111 8.7 0.069 0.079 57.973 59.331 density_matrix_trs4 2 5.0 0.002 0.003 54.789 54.895 ls_scf_init_scf 1 4.0 0.000 0.001 38.570 38.574 ls_scf_init_matrix_S 1 5.0 0.000 0.000 37.271 37.316 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 35.089 35.098 mp_waitall_1 5436 11.0 26.571 32.322 26.571 32.322 multiply_cannon_multrec 444 9.7 14.006 16.575 21.039 24.553 make_m2s 222 7.7 0.004 0.005 17.909 20.442 make_images 222 8.7 3.719 4.419 17.847 20.382 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.123 15.746 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.072 14.261 make_images_data 222 9.7 0.003 0.004 10.173 12.594 hybrid_alltoall_any 227 10.6 0.786 3.764 9.839 12.382 multiply_cannon_sync_h2d 444 9.7 6.512 9.063 6.512 9.063 compute_matrix_preconditioner 1 6.0 0.001 0.002 8.087 8.093 dbcsr_mm_accdrv_process 3003 10.4 0.179 0.353 6.737 7.838 acc_transpose_blocks 444 9.7 0.002 0.002 7.594 7.604 acc_transpose_blocks_kernels 444 10.7 0.006 0.007 7.556 7.565 jit_kernel_transpose 1 13.0 7.550 7.559 7.550 7.559 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.487 6.417 7.487 mp_allgather_i34 111 8.7 2.798 7.038 2.798 7.038 arnoldi_extremal 4 6.8 0.000 0.000 6.044 6.052 arnoldi_normal_ev 4 7.8 0.001 0.004 6.044 6.052 build_subspace 16 8.4 0.015 0.019 5.638 5.652 mp_sum_l 887 5.1 2.905 5.547 2.905 5.547 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.604 4.736 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.193 4.418 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.983 4.239 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.620 4.187 mp_irecv_dv 1241 11.2 1.603 4.164 1.603 4.164 dbcsr_matrix_vector_mult_local 304 10.0 3.679 4.154 3.681 4.157 make_images_sizes 222 9.7 0.000 0.000 1.095 3.729 mp_alltoall_i44 222 10.7 1.095 3.728 1.095 3.728 calculate_norms 792 9.8 3.545 3.710 3.545 3.710 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.598 3.687 ls_scf_post 1 4.0 0.000 0.000 3.620 3.623 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.436 dbcsr_finalize 304 7.8 0.062 0.078 2.201 2.282 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=106.921000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3683.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_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 8.789324E+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 0.076 0.093 108.600 108.601 qs_energies 1 2.0 0.000 0.000 107.166 107.179 ls_scf 1 3.0 0.000 0.000 104.226 104.238 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.893 78.016 ls_scf_main 1 4.0 0.000 0.000 65.971 65.972 density_matrix_trs4 2 5.0 0.002 0.003 56.765 56.819 multiply_cannon 111 7.7 0.143 0.209 50.088 52.070 multiply_cannon_loop 111 8.7 0.067 0.069 46.574 47.601 ls_scf_init_scf 1 4.0 0.001 0.001 34.605 34.605 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.018 33.044 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.171 30.182 mp_waitall_1 4527 11.1 22.516 26.455 22.516 26.455 make_m2s 222 7.7 0.005 0.005 24.165 25.180 make_images 222 8.7 4.584 4.967 24.059 25.073 multiply_cannon_multrec 444 9.7 17.866 18.628 22.464 23.135 hybrid_alltoall_any 227 10.6 1.661 3.639 13.043 15.849 make_images_data 222 9.7 0.003 0.003 13.285 15.827 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.659 11.056 multiply_cannon_sync_h2d 444 9.7 8.846 8.889 8.846 8.889 arnoldi_extremal 4 6.8 0.000 0.000 7.322 7.336 arnoldi_normal_ev 4 7.8 0.002 0.008 7.322 7.336 build_subspace 16 8.4 0.026 0.036 6.771 6.787 dbcsr_matrix_vector_mult 304 9.0 0.017 0.032 5.433 5.587 dbcsr_matrix_vector_mult_local 304 10.0 5.010 5.329 5.012 5.331 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.192 5.284 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.026 5.275 dbcsr_mm_accdrv_process 1814 10.4 0.231 0.324 4.427 4.558 dbcsr_mm_accdrv_process_sort 1814 11.4 4.125 4.259 4.125 4.259 make_images_sizes 222 9.7 0.000 0.000 1.559 3.757 mp_alltoall_i44 222 10.7 1.559 3.756 1.559 3.756 ls_scf_post 1 4.0 0.000 0.000 3.650 3.662 mp_allgather_i34 111 8.7 1.143 3.586 1.143 3.586 ls_scf_store_result 1 5.0 0.000 0.000 3.402 3.414 calculate_norms 792 9.8 3.242 3.283 3.242 3.283 dbcsr_finalize 304 7.8 0.082 0.089 3.074 3.131 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.909 2.910 dbcsr_merge_all 275 8.9 0.890 0.916 2.862 2.910 dbcsr_complete_redistribute 5 7.6 1.435 1.470 2.757 2.876 dbcsr_data_release 12724 10.6 2.325 2.838 2.325 2.838 matrix_ls_to_qs 2 6.0 0.000 0.000 2.397 2.532 dbcsr_sort_data 325 11.1 2.439 2.500 2.439 2.500 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.482 2.484 rebuild_ks_matrix 3 7.3 0.000 0.000 2.414 2.416 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.414 2.416 dbcsr_new_transposed 4 7.5 0.243 0.254 2.252 2.264 dbcsr_frobenius_norm 74 6.6 2.057 2.138 2.206 2.244 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.203 dbcsr_add_anytype 103 7.2 0.860 0.892 2.131 2.203 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.601000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6850.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b314dfac77ca549718eae9848a54a5a4c852e9f1_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 586.436608E+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 656765760. 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 0.070 0.189 280.847 280.849 qs_mol_dyn_low 1 2.0 0.005 0.033 279.469 279.491 qs_forces 5 3.8 0.006 0.038 279.324 279.325 qs_energies 5 4.8 0.003 0.023 275.684 275.726 scf_env_do_scf 5 5.8 0.001 0.005 255.052 255.055 scf_env_do_scf_inner_loop 105 6.6 0.003 0.023 214.126 214.144 qs_scf_new_mos 105 7.6 0.001 0.001 160.190 160.403 qs_scf_loop_do_ot 105 8.6 0.001 0.001 160.190 160.403 ot_scf_mini 105 9.6 0.003 0.004 150.155 150.350 dbcsr_multiply_generic 1445 12.2 0.129 0.138 141.016 141.556 velocity_verlet 4 3.0 0.001 0.011 127.904 127.906 multiply_cannon 1445 13.2 0.280 0.298 118.522 121.249 multiply_cannon_loop 1445 14.2 2.353 2.483 115.863 117.402 ot_mini 105 10.6 0.001 0.001 62.884 63.050 qs_ot_get_p 112 10.4 0.001 0.002 56.675 57.015 mp_waitall_1 488190 16.1 38.961 47.350 38.961 47.350 multiply_cannon_multrec 69360 15.2 31.811 37.521 39.322 44.959 qs_ot_p2m_diag 40 11.0 0.020 0.031 44.032 44.161 qs_ot_get_derivative 55 11.6 0.001 0.003 41.156 41.307 init_scf_loop 7 6.6 0.000 0.005 40.847 40.852 cp_dbcsr_syevd 40 12.0 0.002 0.003 40.827 40.830 rebuild_ks_matrix 110 8.4 0.000 0.000 35.068 35.266 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.014 35.067 35.266 prepare_preconditioner 7 7.6 0.000 0.001 35.031 35.069 make_preconditioner 7 8.6 0.000 0.006 35.031 35.069 cp_fm_syevd 40 13.0 0.000 0.003 34.835 34.987 multiply_cannon_sync_h2d 69360 15.2 30.757 34.850 30.757 34.850 multiply_cannon_metrocomm3 69360 15.2 0.194 0.206 25.321 34.048 qs_ks_update_qs_env 112 7.6 0.001 0.001 32.160 32.326 cp_fm_redistribute_end 40 14.0 14.775 29.423 14.784 29.430 cp_fm_syevd_base 40 14.0 14.625 29.269 14.625 29.269 qs_rho_update_rho_low 110 7.6 0.001 0.004 26.696 27.038 calculate_rho_elec 110 8.6 0.030 0.035 26.696 27.038 make_full_inverse_cholesky 7 9.6 0.000 0.000 25.936 26.014 rs_pw_transfer 690 11.5 0.011 0.013 22.799 24.247 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.020 23.230 apply_single 62 13.6 0.000 0.000 23.020 23.230 density_rs2pw 110 9.6 0.006 0.008 19.778 21.083 ot_new_cg_direction 55 11.6 0.001 0.001 20.983 20.988 init_scf_run 5 5.8 0.000 0.001 17.512 17.514 scf_env_initial_rho_setup 5 6.8 0.000 0.002 17.512 17.513 cp_fm_cholesky_invert 7 10.6 16.871 16.887 16.871 16.887 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.357 15.485 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.192 15.410 pw_transfer 1645 12.4 0.082 0.107 14.432 14.741 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 14.291 14.601 mp_sum_l 4764 12.2 12.311 13.438 12.311 13.438 sum_up_and_integrate 60 10.3 0.028 0.034 13.294 13.306 integrate_v_rspace 60 11.3 0.002 0.002 13.266 13.279 fft_wrap_pw1pw2_240 915 15.0 1.183 1.257 11.900 12.231 calculate_dm_sparse 110 9.5 0.000 0.001 11.669 11.838 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 11.728 11.748 qs_vxc_create 110 10.4 0.002 0.005 11.422 11.456 calculate_first_density_matrix 1 7.0 0.000 0.007 11.336 11.354 acc_transpose_blocks 69360 15.2 0.344 0.377 10.412 11.250 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 11.152 11.221 check_diag 80 13.5 8.587 8.859 10.654 10.804 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.465 10.519 make_m2s 2890 13.2 0.078 0.084 9.772 10.357 fft3d_pb 915 16.0 2.344 2.568 10.006 10.313 make_images 2890 14.2 0.241 0.261 9.666 10.250 rs_pw_transfer_RS2PW_30 110 11.6 1.585 1.663 8.548 9.611 multiply_cannon_metrocomm1 69360 15.2 0.088 0.095 4.877 9.516 acc_transpose_blocks_kernels 69360 16.2 0.824 0.894 8.610 9.295 mp_sendrecv_dv 168740 12.6 8.846 9.079 8.846 9.079 potential_pw2rs 60 12.3 0.003 0.003 8.964 9.046 make_full_single_inverse 7 9.6 0.001 0.008 8.654 8.691 jit_kernel_transpose 5 15.0 7.786 8.438 7.786 8.438 cp_fm_cholesky_decompose 14 10.2 8.331 8.341 8.331 8.341 dbcsr_mm_accdrv_process 154766 15.8 3.911 4.214 7.380 8.246 xc_rho_set_and_dset_create 110 12.4 0.077 0.109 7.923 8.170 xc_vxc_pw_create 60 11.3 0.039 0.051 7.704 7.739 mp_alltoall_z22v 2340 17.7 7.173 7.633 7.173 7.633 xc_pw_derive 510 13.4 0.005 0.006 7.105 7.168 mp_alltoall_d11v 1300 13.8 5.527 6.173 5.527 6.173 mp_waitany 7680 13.5 4.787 6.118 4.787 6.118 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=280.849000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=555.600000, yerr=4.029888 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b314dfac77ca549718eae9848a54a5a4c852e9f1 Summary: empty Status: OK