=== This is the CP2K Performance-Test === Updating d99951c42..52c01cbe9 Fast-forward src/negf_methods.F | 25 +++---------------------- tools/conventions/conventions.supp | 3 +++ 2 files changed, 6 insertions(+), 22 deletions(-) Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: 52c01cbe9bfe1b67bcd624e507926c0777952c09 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/01 job id: 44682315 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/02 job id: 44682316 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/03 job id: 44682317 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/04 job id: 44682318 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/05 job id: 44682319 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/06 job id: 44682320 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/07 job id: 44682321 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/08 job id: 44682322 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/09 job id: 44682323 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/10 job id: 44682324 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/11 job id: 44682325 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/12 job id: 44682326 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/13 job id: 44682327 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/14 job id: 44682328 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/15 job id: 44682330 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/16 job id: 44682331 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/17 job id: 44682332 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/18 job id: 44682333 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/19 job id: 44682334 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/20 job id: 44682335 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/21 job id: 44682336 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/22 job id: 44682337 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/23 job id: 44682338 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/24 job id: 44682339 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/25 job id: 44682340 --- 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/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/26 job id: 44682341 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 133.771 133.772 farming_run 1 2.0 133.233 133.235 133.743 133.747 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.458094E+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 229 1108280. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.036 0.086 115.624 115.625 qs_energies 1 2.0 0.000 0.001 115.293 115.308 mp2_main 1 3.0 0.001 0.011 113.428 113.437 mp2_gpw_main 1 4.0 0.021 0.028 112.584 112.587 mp2_ri_gpw_compute_in 1 5.0 0.175 0.230 93.790 94.187 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.548 55.944 mp2_eri_3c_integrate_gpw 272 7.0 0.157 0.186 41.876 47.239 get_2c_integrals 1 6.0 0.000 0.000 37.329 38.070 integrate_v_rspace 273 8.0 0.441 0.500 25.217 30.345 pw_transfer 6555 10.6 0.374 0.387 27.453 28.137 grid_integrate_task_list 273 9.0 21.011 26.641 21.011 26.641 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.130 26.622 fft_wrap_pw1pw2_100 2178 12.4 1.181 1.403 23.650 24.145 compute_2c_integrals 1 7.0 0.009 0.011 19.664 19.666 compute_2c_integrals_loop_lm 1 8.0 0.002 0.004 18.871 19.334 mp2_eri_2c_integrate_gpw 1 9.0 2.381 2.442 18.869 19.332 rpa_ri_compute_en 1 5.0 0.010 0.016 18.686 18.932 cp_fm_cholesky_decompose 12 8.2 17.587 18.323 17.587 18.323 cholesky_decomp 1 7.0 0.000 0.000 16.509 17.248 fft3d_s 5443 13.4 16.209 16.566 16.231 16.587 ao_to_mo_and_store_B_mult_1 272 7.0 10.841 15.576 10.841 15.576 calculate_wavefunction 272 8.0 5.432 5.603 12.581 13.183 rpa_num_int 1 6.0 0.000 0.004 10.492 10.502 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.460 10.492 calc_mat_Q 8 8.0 0.000 0.000 9.331 9.436 contract_S_to_Q 8 9.0 0.000 0.001 8.753 8.859 calc_potential_gpw 544 9.5 0.005 0.006 8.218 8.642 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.212 8.541 potential_pw2rs 545 10.0 0.107 0.109 7.730 8.415 parallel_gemm_fm 14 9.1 0.000 0.000 8.330 8.410 parallel_gemm_fm_cosma 14 10.1 8.330 8.410 8.330 8.410 collocate_single_gaussian 272 10.0 0.039 0.041 7.463 7.707 create_integ_mat 1 6.0 0.014 0.028 7.689 7.699 array2fm 1 7.0 0.000 0.000 6.748 7.112 pw_scatter_s 2720 13.7 4.419 4.650 4.419 4.650 pw_gather_s 2722 13.2 3.910 4.293 3.910 4.293 array2fm_buffer_send 1 8.0 2.975 3.154 2.975 3.154 pw_poisson_solve 545 10.5 1.114 1.161 2.144 2.356 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.583641, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2731.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.039 396.537 396.537 farming_run 1 2.0 395.767 395.772 396.503 396.505 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.226027E+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 704 407793. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. 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 0.014 0.053 209.797 209.797 qs_energies 1 2.0 0.000 0.000 209.566 209.572 scf_env_do_scf 1 3.0 0.000 0.000 106.350 106.350 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.485 105.492 rebuild_ks_matrix 4 6.0 0.000 0.000 105.484 105.491 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 105.484 105.491 hfx_ks_matrix 4 8.0 0.001 0.001 105.101 105.104 integrate_four_center 4 9.0 0.145 0.463 105.100 105.103 mp2_main 1 3.0 0.000 0.000 102.928 102.934 mp2_gpw_main 1 4.0 0.032 0.046 102.073 102.082 integrate_four_center_main 4 10.0 0.103 0.423 96.915 99.412 integrate_four_center_bin 266 11.0 96.811 99.370 96.811 99.370 init_scf_loop 1 4.0 0.000 0.000 92.203 92.204 mp2_ri_gpw_compute_in 1 5.0 0.064 0.106 74.989 76.047 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.510 55.567 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.164 42.134 47.241 integrate_v_rspace 95 8.0 0.400 0.567 28.518 33.433 pw_transfer 2240 10.6 0.142 0.162 29.969 30.426 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.963 29.439 ao_to_mo_and_store_B_mult_1 91 7.0 10.684 29.159 10.684 29.159 grid_integrate_task_list 95 9.0 23.808 28.928 23.808 28.928 mp2_ri_gpw_compute_en 1 5.0 0.056 0.063 26.940 28.704 fft_wrap_pw1pw2_100 730 12.4 1.307 1.432 26.645 27.077 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.821 1.879 25.225 25.235 get_2c_integrals 1 6.0 0.000 0.000 20.384 20.413 compute_2c_integrals 1 7.0 0.006 0.011 19.361 19.364 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.990 19.219 mp2_eri_2c_integrate_gpw 1 9.0 1.737 1.867 18.989 19.218 fft3d_s 1823 13.4 18.407 18.713 18.421 18.726 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.145 14.145 calculate_wavefunction 91 8.0 2.018 2.048 9.735 9.963 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.582 8.829 9.509 potential_pw2rs 186 10.0 0.034 0.035 8.644 9.194 local_gemm 172 8.0 8.272 8.932 8.272 8.932 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.519 8.190 8.673 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.325 8.621 calc_potential_gpw 182 9.5 0.002 0.002 7.985 8.260 collocate_single_gaussian 91 10.0 0.017 0.024 7.918 8.110 mp_sendrecv_dm3 2068 8.0 6.230 6.691 6.230 6.691 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.411 6.340 6.411 mp_sync 38 10.4 2.867 6.294 2.867 6.294 pw_gather_s 912 13.2 4.880 5.362 4.880 5.362 pw_scatter_s 910 13.7 3.967 4.237 3.967 4.237 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.068124, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.256320E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 394227. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.026 52.266 52.267 qs_mol_dyn_low 1 2.0 0.003 0.006 51.948 51.955 qs_forces 11 3.9 0.002 0.002 51.866 51.868 qs_energies 11 4.9 0.002 0.002 50.406 50.415 scf_env_do_scf 11 5.9 0.000 0.001 44.228 44.228 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.207 42.207 dbcsr_multiply_generic 2286 12.5 0.094 0.097 33.806 34.245 qs_scf_new_mos 108 7.5 0.000 0.001 32.319 32.603 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.318 32.603 ot_scf_mini 108 9.5 0.002 0.002 30.658 30.870 multiply_cannon 2286 13.5 0.178 0.186 26.284 27.827 multiply_cannon_loop 2286 14.5 1.463 1.551 25.640 27.142 velocity_verlet 10 3.0 0.001 0.001 25.717 25.717 ot_mini 108 10.5 0.001 0.001 19.505 19.781 qs_ot_get_derivative 108 11.5 0.001 0.001 16.564 16.782 mp_waitall_1 245248 16.5 8.698 14.901 8.698 14.901 multiply_cannon_metrocomm3 54864 15.5 0.068 0.072 6.257 13.246 multiply_cannon_multrec 54864 15.5 4.233 6.564 7.707 11.184 rebuild_ks_matrix 119 8.3 0.000 0.000 7.782 7.941 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.781 7.941 mp_sum_l 7207 12.9 5.716 7.766 5.716 7.766 multiply_cannon_sync_h2d 54864 15.5 5.872 7.270 5.872 7.270 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.862 6.998 qs_ot_get_p 119 10.4 0.001 0.001 6.483 6.777 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.597 6.045 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.591 5.722 init_scf_run 11 5.9 0.000 0.001 4.962 4.962 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.961 4.962 dbcsr_mm_accdrv_process 76910 16.1 1.131 1.819 3.398 4.766 sum_up_and_integrate 119 10.3 0.012 0.014 4.420 4.428 integrate_v_rspace 119 11.3 0.002 0.003 4.408 4.417 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.028 4.144 calculate_rho_elec 119 8.7 0.011 0.017 4.027 4.144 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.644 3.728 multiply_cannon_metrocomm1 54864 15.5 0.052 0.057 1.872 3.153 calculate_dm_sparse 119 9.5 0.000 0.001 3.017 3.145 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.897 3.095 apply_single 119 13.6 0.000 0.000 2.896 3.094 jit_kernel_multiply 13 15.8 2.207 2.858 2.207 2.858 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.826 2.827 calculate_first_density_matrix 1 7.0 0.000 0.000 2.813 2.818 rs_pw_transfer 974 11.9 0.011 0.012 2.586 2.713 ot_diis_step 108 11.5 0.006 0.006 2.658 2.659 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.640 2.640 cp_fm_redistribute_end 50 14.0 2.404 2.620 2.409 2.621 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.502 2.574 cp_fm_diag_elpa_base 50 14.0 0.210 2.535 0.211 2.544 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.365 2.367 acc_transpose_blocks 54864 15.5 0.222 0.245 1.758 2.225 density_rs2pw 119 9.7 0.004 0.005 2.060 2.205 grid_integrate_task_list 119 12.3 2.042 2.163 2.042 2.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.094 2.144 wfi_extrapolate 11 7.9 0.001 0.001 2.091 2.091 init_scf_loop 11 6.9 0.000 0.001 2.006 2.006 mp_sum_d 4127 12.0 1.335 1.984 1.335 1.984 potential_pw2rs 119 12.3 0.004 0.004 1.775 1.783 pw_transfer 1439 11.6 0.053 0.059 1.649 1.715 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.572 1.642 make_m2s 4572 13.5 0.053 0.056 1.498 1.551 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.515 1.540 make_images 4572 14.5 0.132 0.137 1.416 1.468 mp_alltoall_d11v 2130 13.8 1.191 1.450 1.191 1.450 fft3d_ps 1201 14.6 0.366 0.474 1.343 1.407 grid_collocate_task_list 119 9.7 1.296 1.358 1.296 1.358 mp_waitany 12084 13.8 1.209 1.333 1.209 1.333 fft_wrap_pw1pw2_140 487 13.2 0.083 0.097 1.203 1.275 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.731 1.173 acc_transpose_blocks_kernels 54864 16.5 0.248 0.385 0.800 1.084 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.267000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 489.299968E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1030216. 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 0.017 0.030 38.640 38.642 qs_mol_dyn_low 1 2.0 0.003 0.004 38.363 38.370 qs_forces 11 3.9 0.002 0.004 38.223 38.223 qs_energies 11 4.9 0.002 0.007 36.554 36.557 scf_env_do_scf 11 5.9 0.001 0.001 30.919 30.920 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.525 28.526 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.629 21.983 qs_scf_new_mos 108 7.5 0.001 0.001 19.585 19.832 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.584 19.832 ot_scf_mini 108 9.5 0.002 0.003 18.717 18.890 multiply_cannon 2286 13.5 0.206 0.216 16.771 18.411 velocity_verlet 10 3.0 0.001 0.001 17.909 17.912 multiply_cannon_loop 2286 14.5 0.896 0.970 15.707 17.080 ot_mini 108 10.5 0.001 0.001 11.617 11.854 mp_waitall_1 200699 16.5 5.482 10.763 5.482 10.763 multiply_cannon_metrocomm3 27432 15.5 0.067 0.071 4.156 9.522 qs_ot_get_derivative 108 11.5 0.001 0.001 9.185 9.363 multiply_cannon_multrec 27432 15.5 1.976 4.438 6.428 9.279 rebuild_ks_matrix 119 8.3 0.000 0.000 7.066 7.205 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.015 7.066 7.205 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.247 6.372 dbcsr_mm_accdrv_process 47894 16.0 3.398 5.533 4.383 6.283 qs_ot_get_p 119 10.4 0.001 0.001 4.244 4.477 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.588 4.460 init_scf_run 11 5.9 0.000 0.001 4.408 4.409 scf_env_initial_rho_setup 11 6.9 0.003 0.004 4.408 4.409 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.046 4.117 apply_single 119 13.6 0.000 0.000 3.046 4.117 sum_up_and_integrate 119 10.3 0.024 0.027 4.083 4.089 integrate_v_rspace 119 11.3 0.002 0.003 4.059 4.068 mp_sum_l 7207 12.9 2.018 3.970 2.018 3.970 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.643 3.676 calculate_rho_elec 119 8.7 0.021 0.024 3.642 3.676 calculate_first_density_matrix 1 7.0 0.005 0.039 2.937 2.940 rs_pw_transfer 974 11.9 0.010 0.011 2.439 2.896 qs_ot_p2m_diag 50 11.0 0.008 0.012 2.865 2.883 multiply_cannon_sync_h2d 27432 15.5 2.195 2.844 2.195 2.844 make_m2s 4572 13.5 0.052 0.054 2.393 2.620 calculate_dm_sparse 119 9.5 0.000 0.000 2.491 2.569 make_images 4572 14.5 0.198 0.235 2.305 2.528 density_rs2pw 119 9.7 0.004 0.004 1.990 2.456 cp_dbcsr_syevd 50 12.0 0.005 0.006 2.441 2.441 ot_diis_step 108 11.5 0.010 0.011 2.383 2.384 init_scf_loop 11 6.9 0.001 0.003 2.375 2.375 jit_kernel_multiply 9 16.1 0.933 2.306 0.933 2.306 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.066 2.156 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.129 2.131 cp_fm_redistribute_end 50 14.0 1.764 2.107 1.767 2.108 cp_fm_diag_elpa_base 50 14.0 0.326 2.001 0.339 2.053 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.966 1.968 grid_integrate_task_list 119 12.3 1.837 1.938 1.837 1.938 pw_transfer 1439 11.6 0.065 0.071 1.880 1.912 potential_pw2rs 119 12.3 0.006 0.006 1.837 1.844 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.785 1.825 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.788 1.823 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.728 1.735 acc_transpose_blocks 27432 15.5 0.109 0.113 1.220 1.548 fft3d_ps 1201 14.6 0.509 0.561 1.490 1.520 prepare_preconditioner 11 7.9 0.000 0.001 1.486 1.514 make_preconditioner 11 8.9 0.000 0.001 1.486 1.514 make_images_data 4572 15.5 0.044 0.050 1.104 1.500 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.390 1.446 fft_wrap_pw1pw2_140 487 13.2 0.079 0.086 1.396 1.430 wfi_extrapolate 11 7.9 0.001 0.001 1.421 1.421 hybrid_alltoall_any 4725 16.4 0.050 0.110 0.970 1.419 grid_collocate_task_list 119 9.7 1.228 1.357 1.228 1.357 mp_allgather_i34 2286 14.5 0.512 1.247 0.512 1.247 mp_alltoall_d11v 2130 13.8 1.096 1.218 1.096 1.218 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.148 1.194 mp_sum_d 4127 12.0 0.607 1.074 0.607 1.074 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.549 1.006 mp_waitany 5720 13.7 0.522 0.999 0.522 0.999 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.965 0.966 acc_transpose_blocks_kernels 27432 16.5 0.182 0.271 0.689 0.923 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.877 0.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=38.642000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.454545, yerr=2.189381 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 524.029952E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1587286. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.105 32.734 32.765 qs_mol_dyn_low 1 2.0 0.003 0.003 32.407 32.414 qs_forces 11 3.9 0.002 0.005 32.346 32.346 qs_energies 11 4.9 0.002 0.006 30.793 30.795 scf_env_do_scf 11 5.9 0.001 0.002 25.994 25.995 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.483 23.483 dbcsr_multiply_generic 2286 12.5 0.093 0.095 17.224 17.308 qs_scf_new_mos 108 7.5 0.001 0.001 15.444 15.469 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.444 15.468 velocity_verlet 10 3.0 0.001 0.001 14.976 14.978 multiply_cannon 2286 13.5 0.195 0.201 14.100 14.891 ot_scf_mini 108 9.5 0.002 0.003 14.730 14.742 multiply_cannon_loop 2286 14.5 0.631 0.665 13.334 14.140 ot_mini 108 10.5 0.001 0.001 9.481 9.492 multiply_cannon_multrec 18288 15.5 2.165 3.699 7.164 8.005 qs_ot_get_derivative 108 11.5 0.001 0.001 7.988 8.002 rebuild_ks_matrix 119 8.3 0.000 0.000 6.298 6.313 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.297 6.313 dbcsr_mm_accdrv_process 38222 16.0 4.182 5.634 4.915 5.902 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.566 5.580 mp_waitall_1 158411 16.6 3.137 4.489 3.137 4.489 sum_up_and_integrate 119 10.3 0.031 0.032 3.891 3.896 integrate_v_rspace 119 11.3 0.002 0.003 3.859 3.870 init_scf_run 11 5.9 0.000 0.001 3.621 3.621 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.621 3.621 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.362 3.374 calculate_rho_elec 119 8.7 0.030 0.031 3.362 3.374 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.679 3.309 qs_ot_get_p 119 10.4 0.001 0.001 3.102 3.126 init_scf_loop 11 6.9 0.001 0.004 2.496 2.497 rs_pw_transfer 974 11.9 0.009 0.010 2.196 2.482 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.444 2.436 calculate_first_density_matrix 1 7.0 0.000 0.002 2.417 2.417 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.047 2.365 apply_single 119 13.6 0.000 0.000 2.047 2.365 density_rs2pw 119 9.7 0.004 0.004 1.898 2.176 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.098 2.103 jit_kernel_multiply 10 16.2 0.682 2.003 0.682 2.003 calculate_dm_sparse 119 9.5 0.000 0.000 1.910 1.924 grid_integrate_task_list 119 12.3 1.799 1.874 1.799 1.874 pw_transfer 1439 11.6 0.065 0.071 1.854 1.861 make_m2s 4572 13.5 0.044 0.045 1.685 1.845 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.842 1.842 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.760 1.772 make_images 4572 14.5 0.188 0.199 1.601 1.760 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.754 1.755 prepare_preconditioner 11 7.9 0.000 0.000 1.714 1.717 make_preconditioner 11 8.9 0.000 0.001 1.714 1.717 potential_pw2rs 119 12.3 0.007 0.008 1.671 1.677 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.561 1.647 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.626 1.627 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.618 1.623 multiply_cannon_sync_h2d 18288 15.5 1.419 1.618 1.419 1.618 cp_fm_redistribute_end 50 14.0 1.213 1.607 1.214 1.607 cp_fm_diag_elpa_base 50 14.0 0.378 1.537 0.392 1.574 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.509 1.515 ot_diis_step 108 11.5 0.011 0.011 1.471 1.471 fft3d_ps 1201 14.6 0.520 0.538 1.440 1.450 fft_wrap_pw1pw2_140 487 13.2 0.090 0.094 1.418 1.430 mp_sum_l 7207 12.9 1.092 1.429 1.092 1.429 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.978 1.396 grid_collocate_task_list 119 9.7 1.211 1.355 1.211 1.355 acc_transpose_blocks 18288 15.5 0.075 0.077 1.302 1.333 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.286 1.293 wfi_extrapolate 11 7.9 0.001 0.001 1.159 1.159 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.957 0.959 make_images_data 4572 15.5 0.044 0.048 0.764 0.927 acc_transpose_blocks_kernels 18288 16.5 0.209 0.218 0.865 0.884 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.859 0.880 hybrid_alltoall_any 4725 16.4 0.055 0.114 0.650 0.854 mp_waitany 9880 13.7 0.532 0.835 0.532 0.835 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.802 0.804 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.121 0.517 0.801 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.662 0.720 mp_alltoall_d11v 2130 13.8 0.605 0.714 0.605 0.714 cp_fm_cholesky_invert 11 10.9 0.701 0.705 0.701 0.705 jit_kernel_transpose 5 15.6 0.656 0.672 0.656 0.672 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.765000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.363636, yerr=2.568081 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 557.637632E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.059 0.144 35.825 35.863 qs_mol_dyn_low 1 2.0 0.003 0.004 35.393 35.401 qs_forces 11 3.9 0.003 0.005 34.926 34.927 qs_energies 11 4.9 0.004 0.009 33.212 33.218 scf_env_do_scf 11 5.9 0.001 0.002 27.718 27.720 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 24.292 24.294 dbcsr_multiply_generic 2286 12.5 0.099 0.102 18.421 18.542 velocity_verlet 10 3.0 0.001 0.001 17.898 17.917 qs_scf_new_mos 108 7.5 0.001 0.001 15.920 15.975 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.919 15.974 multiply_cannon 2286 13.5 0.229 0.265 14.936 15.382 ot_scf_mini 108 9.5 0.003 0.004 14.990 15.046 multiply_cannon_loop 2286 14.5 0.935 0.968 14.018 14.344 multiply_cannon_multrec 27432 15.5 2.353 3.030 8.858 9.277 ot_mini 108 10.5 0.001 0.005 9.164 9.230 dbcsr_mm_accdrv_process 47916 15.9 5.416 6.942 6.410 7.776 qs_ot_get_derivative 108 11.5 0.001 0.001 7.366 7.421 rebuild_ks_matrix 119 8.3 0.000 0.000 6.538 6.588 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.016 6.538 6.587 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.806 5.850 init_scf_run 11 5.9 0.000 0.001 4.116 4.116 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.116 4.116 sum_up_and_integrate 119 10.3 0.036 0.040 3.770 3.774 integrate_v_rspace 119 11.3 0.003 0.003 3.734 3.739 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.485 3.513 calculate_rho_elec 119 8.7 0.040 0.046 3.484 3.513 init_scf_loop 11 6.9 0.013 0.030 3.406 3.410 qs_ot_get_p 119 10.4 0.004 0.013 3.269 3.340 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.781 3.270 calculate_first_density_matrix 1 7.0 0.001 0.003 2.775 2.778 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 2.118 2.587 apply_single 119 13.6 0.000 0.000 2.118 2.586 prepare_preconditioner 11 7.9 0.000 0.000 2.497 2.507 make_preconditioner 11 8.9 0.001 0.001 2.497 2.507 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.113 2.432 mp_waitall_1 137007 16.6 1.696 2.295 1.696 2.295 calculate_dm_sparse 119 9.5 0.000 0.000 2.155 2.213 jit_kernel_multiply 10 15.6 0.934 2.205 0.934 2.205 make_m2s 4572 13.5 0.054 0.055 2.069 2.195 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.192 2.192 rs_pw_transfer 974 11.9 0.009 0.009 1.952 2.155 density_rs2pw 119 9.7 0.004 0.004 1.919 2.102 make_images 4572 14.5 0.269 0.328 1.963 2.089 pw_transfer 1439 11.6 0.066 0.071 2.008 2.049 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.018 2.047 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.035 2.044 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.915 1.961 grid_integrate_task_list 119 12.3 1.816 1.893 1.816 1.893 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.848 1.877 acc_transpose_blocks 27432 15.5 0.111 0.114 1.780 1.820 ot_diis_step 108 11.5 0.012 0.017 1.759 1.759 cp_dbcsr_syevd 50 12.0 0.006 0.007 1.730 1.730 mp_sum_l 7207 12.9 1.023 1.641 1.023 1.641 fft_wrap_pw1pw2_140 487 13.2 0.089 0.096 1.589 1.639 fft3d_ps 1201 14.6 0.557 0.606 1.589 1.627 potential_pw2rs 119 12.3 0.008 0.010 1.537 1.540 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.503 1.507 cp_fm_redistribute_end 50 14.0 0.995 1.481 0.996 1.481 cp_fm_diag_elpa_base 50 14.0 0.463 1.418 0.482 1.457 grid_collocate_task_list 119 9.7 1.222 1.358 1.222 1.358 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.345 1.355 multiply_cannon_metrocomm3 27432 15.5 0.037 0.038 0.777 1.293 wfi_extrapolate 11 7.9 0.001 0.001 1.292 1.292 acc_transpose_blocks_kernels 27432 16.5 0.269 0.276 1.157 1.190 cp_fm_upper_to_full 72 13.5 0.795 1.101 0.795 1.101 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.078 1.098 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.093 1.094 multiply_cannon_sync_h2d 27432 15.5 1.001 1.078 1.001 1.078 dbcsr_complete_redistribute 329 12.2 0.120 0.148 0.763 1.051 make_images_data 4572 15.5 0.045 0.048 0.810 0.947 jit_kernel_transpose 5 15.6 0.888 0.915 0.888 0.915 hybrid_alltoall_any 4725 16.4 0.061 0.150 0.703 0.880 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.794 0.874 mp_alltoall_d11v 2130 13.8 0.673 0.867 0.673 0.867 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.562 0.848 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.822 0.827 mp_alltoall_i22 627 13.8 0.444 0.748 0.444 0.748 cp_fm_cholesky_invert 11 10.9 0.727 0.730 0.727 0.730 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.863000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.545455, yerr=3.201239 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 604.852224E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 28.341 28.343 qs_mol_dyn_low 1 2.0 0.003 0.004 28.114 28.121 qs_forces 11 3.9 0.001 0.002 28.056 28.057 qs_energies 11 4.9 0.002 0.002 26.329 26.333 scf_env_do_scf 11 5.9 0.000 0.001 21.519 21.519 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.089 19.091 velocity_verlet 10 3.0 0.001 0.001 14.104 14.106 dbcsr_multiply_generic 2286 12.5 0.091 0.094 12.654 12.728 qs_scf_new_mos 108 7.5 0.001 0.001 11.195 11.216 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.195 11.215 multiply_cannon 2286 13.5 0.229 0.232 10.271 10.714 ot_scf_mini 108 9.5 0.002 0.002 10.535 10.563 multiply_cannon_loop 2286 14.5 0.328 0.340 9.382 9.563 multiply_cannon_multrec 9144 15.5 1.921 2.570 6.215 6.806 ot_mini 108 10.5 0.001 0.001 6.164 6.197 rebuild_ks_matrix 119 8.3 0.000 0.000 5.962 5.982 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.962 5.982 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.307 5.325 qs_ot_get_derivative 108 11.5 0.001 0.001 4.897 4.922 dbcsr_mm_accdrv_process 12550 15.8 3.000 4.226 4.193 4.274 sum_up_and_integrate 119 10.3 0.038 0.042 3.617 3.622 integrate_v_rspace 119 11.3 0.003 0.003 3.579 3.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.463 3.472 calculate_rho_elec 119 8.7 0.060 0.061 3.462 3.471 init_scf_run 11 5.9 0.000 0.001 3.394 3.394 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.394 3.394 qs_ot_get_p 119 10.4 0.001 0.001 2.633 2.677 init_scf_loop 11 6.9 0.000 0.000 2.411 2.411 calculate_first_density_matrix 1 7.0 0.000 0.000 2.285 2.287 mp_waitall_1 115863 16.7 1.654 2.268 1.654 2.268 pw_transfer 1439 11.6 0.066 0.069 2.075 2.086 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.981 1.995 grid_integrate_task_list 119 12.3 1.848 1.977 1.848 1.977 density_rs2pw 119 9.7 0.004 0.004 1.817 1.961 jit_kernel_multiply 10 15.5 1.155 1.889 1.155 1.889 make_m2s 4572 13.5 0.034 0.035 1.647 1.822 calculate_dm_sparse 119 9.5 0.000 0.000 1.766 1.783 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.771 1.774 rs_pw_transfer 974 11.9 0.008 0.008 1.605 1.746 make_images 4572 14.5 0.267 0.301 1.559 1.731 prepare_preconditioner 11 7.9 0.000 0.000 1.663 1.668 make_preconditioner 11 8.9 0.000 0.000 1.663 1.668 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.645 1.657 fft3d_ps 1201 14.6 0.560 0.571 1.640 1.653 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.632 1.633 cp_dbcsr_syevd 50 12.0 0.006 0.006 1.582 1.583 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.557 1.582 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.470 1.490 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.454 1.462 grid_collocate_task_list 119 9.7 1.271 1.392 1.271 1.392 potential_pw2rs 119 12.3 0.010 0.011 1.367 1.370 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.325 1.328 cp_fm_redistribute_end 50 14.0 0.662 1.306 0.662 1.307 cp_fm_diag_elpa_base 50 14.0 0.601 1.240 0.644 1.295 ot_diis_step 108 11.5 0.012 0.012 1.256 1.256 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.212 1.213 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.198 1.202 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.140 1.161 apply_single 119 13.6 0.000 0.000 1.140 1.161 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.086 1.098 wfi_extrapolate 11 7.9 0.001 0.001 1.060 1.060 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.735 0.993 make_images_data 4572 15.5 0.038 0.042 0.781 0.972 acc_transpose_blocks 9144 15.5 0.037 0.038 0.932 0.938 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.870 0.921 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.537 0.875 cp_fm_cholesky_invert 11 10.9 0.833 0.836 0.833 0.836 mp_alltoall_d11v 2130 13.8 0.682 0.813 0.682 0.813 multiply_cannon_sync_h2d 9144 15.5 0.711 0.800 0.711 0.800 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.777 0.783 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.665 0.718 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.700 0.709 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.693 0.697 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.375 0.650 mp_allgather_i34 2286 14.5 0.226 0.638 0.226 0.638 mp_alltoall_z22v 1201 16.6 0.529 0.583 0.529 0.583 jit_kernel_transpose 5 15.6 0.576 0.579 0.576 0.579 mp_waitany 5200 13.7 0.442 0.572 0.442 0.572 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.343000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=570.454545, yerr=6.638231 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 760.745984E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.039 41.342 41.343 qs_mol_dyn_low 1 2.0 0.003 0.003 41.099 41.105 qs_forces 11 3.9 0.002 0.002 41.039 41.040 qs_energies 11 4.9 0.001 0.002 39.067 39.070 scf_env_do_scf 11 5.9 0.001 0.001 33.172 33.172 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.306 25.308 velocity_verlet 10 3.0 0.001 0.001 23.219 23.224 dbcsr_multiply_generic 2286 12.5 0.100 0.101 17.421 17.525 qs_scf_new_mos 108 7.5 0.001 0.001 15.522 15.618 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.521 15.617 ot_scf_mini 108 9.5 0.002 0.002 14.460 14.558 multiply_cannon 2286 13.5 0.300 0.306 13.698 14.509 multiply_cannon_loop 2286 14.5 0.342 0.347 12.479 13.323 multiply_cannon_multrec 9144 15.5 3.528 5.153 8.679 8.803 ot_mini 108 10.5 0.001 0.001 8.671 8.785 init_scf_loop 11 6.9 0.000 0.000 7.839 7.843 rebuild_ks_matrix 119 8.3 0.000 0.000 7.378 7.517 qs_ks_build_kohn_sham_matrix 119 9.3 0.014 0.014 7.377 7.516 prepare_preconditioner 11 7.9 0.000 0.000 6.788 6.801 make_preconditioner 11 8.9 0.000 0.000 6.788 6.801 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.674 6.799 qs_ot_get_derivative 108 11.5 0.001 0.001 6.681 6.780 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.349 6.681 dbcsr_mm_accdrv_process 12550 15.8 3.940 5.321 5.028 6.425 cp_fm_upper_to_full 72 14.2 3.203 4.585 3.203 4.585 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.326 4.389 calculate_rho_elec 119 8.7 0.118 0.121 4.325 4.389 sum_up_and_integrate 119 10.3 0.065 0.067 4.013 4.041 integrate_v_rspace 119 11.3 0.003 0.003 3.948 3.974 init_scf_run 11 5.9 0.000 0.001 3.875 3.875 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.875 3.875 qs_ot_get_p 119 10.4 0.001 0.001 3.103 3.236 mp_waitall_1 94719 16.7 2.234 3.168 2.234 3.168 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.453 2.861 pw_transfer 1439 11.6 0.069 0.069 2.806 2.810 dbcsr_complete_redistribute 329 12.2 0.288 0.293 1.952 2.788 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.707 2.711 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.656 2.472 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.455 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.138 2.423 apply_single 119 13.6 0.000 0.000 2.138 2.423 make_m2s 4572 13.5 0.038 0.038 2.192 2.343 fft3d_ps 1201 14.6 0.594 0.603 2.331 2.334 mp_alltoall_i22 627 13.8 1.475 2.324 1.475 2.324 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.292 2.300 calculate_dm_sparse 119 9.5 0.000 0.000 2.240 2.300 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.434 2.243 make_images 4572 14.5 0.352 0.382 2.070 2.220 density_rs2pw 119 9.7 0.004 0.004 2.195 2.215 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.308 2.163 grid_integrate_task_list 119 12.3 2.048 2.067 2.048 2.067 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.968 1.969 ot_diis_step 108 11.5 0.014 0.015 1.966 1.966 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.878 1.879 jit_kernel_multiply 10 15.8 1.061 1.830 1.061 1.830 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.762 1.763 mp_sum_l 7207 12.9 0.992 1.712 0.992 1.712 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.654 1.703 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.662 1.690 cp_dbcsr_syevd 50 12.0 0.008 0.008 1.609 1.610 grid_collocate_task_list 119 9.7 1.482 1.498 1.482 1.498 rs_pw_transfer 974 11.9 0.009 0.010 1.426 1.464 potential_pw2rs 119 12.3 0.014 0.015 1.417 1.420 cp_fm_cholesky_invert 11 10.9 1.394 1.397 1.394 1.397 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.323 1.383 wfi_extrapolate 11 7.9 0.001 0.001 1.355 1.355 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.325 1.326 cp_fm_diag_elpa_base 50 14.0 1.182 1.235 1.324 1.324 hybrid_alltoall_any 4725 16.4 0.086 0.147 1.038 1.244 mp_alltoall_d11v 2130 13.8 1.102 1.160 1.102 1.160 make_images_data 4572 15.5 0.043 0.045 0.972 1.149 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.117 1.135 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.103 1.117 multiply_cannon_sync_h2d 9144 15.5 1.043 1.045 1.043 1.045 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.985 1.015 acc_transpose_blocks 9144 15.5 0.038 0.039 0.954 0.968 qs_create_task_list 11 7.9 0.001 0.001 0.936 0.948 generate_qs_task_list 11 8.9 0.370 0.389 0.935 0.948 yz_to_x 606 15.1 0.458 0.469 0.931 0.943 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.914 0.927 mp_alltoall_z22v 1201 16.6 0.788 0.836 0.788 0.836 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.343000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=711.000000, yerr=16.558436 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 503.701504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.035 82.128 82.135 qs_mol_dyn_low 1 2.0 0.003 0.003 81.822 81.835 qs_forces 11 3.9 0.002 0.003 81.755 81.757 qs_energies 11 4.9 0.001 0.002 78.945 78.964 scf_env_do_scf 11 5.9 0.001 0.001 70.148 70.150 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.687 64.687 dbcsr_multiply_generic 2055 12.4 0.107 0.110 51.755 52.054 qs_scf_new_mos 99 7.5 0.000 0.001 47.685 47.814 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.684 47.813 ot_scf_mini 99 9.5 0.002 0.002 45.323 45.422 multiply_cannon 2055 13.4 0.173 0.181 42.582 43.374 multiply_cannon_loop 2055 14.4 1.523 1.560 41.703 42.394 velocity_verlet 10 3.0 0.001 0.001 41.556 41.557 ot_mini 99 10.5 0.001 0.001 27.355 27.433 qs_ot_get_derivative 99 11.5 0.001 0.002 20.493 20.597 multiply_cannon_multrec 49320 15.4 12.239 12.891 17.464 18.086 rebuild_ks_matrix 110 8.3 0.000 0.001 14.353 14.470 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.353 14.469 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.592 12.695 mp_waitall_1 220248 16.4 11.159 12.264 11.159 12.264 multiply_cannon_sync_h2d 49320 15.4 9.992 10.726 9.992 10.726 qs_ot_get_p 110 10.4 0.001 0.001 9.480 9.609 multiply_cannon_metrocomm3 49320 15.4 0.078 0.082 6.939 7.907 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.191 7.637 apply_single 110 13.6 0.000 0.000 7.191 7.637 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.801 7.357 sum_up_and_integrate 110 10.3 0.037 0.043 6.918 6.932 integrate_v_rspace 110 11.3 0.003 0.003 6.881 6.905 init_scf_run 11 5.9 0.000 0.001 6.800 6.800 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.800 6.800 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.620 6.680 ot_diis_step 99 11.5 0.006 0.006 6.567 6.567 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.376 6.489 calculate_rho_elec 110 8.6 0.021 0.026 6.375 6.489 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.235 6.278 mp_sum_l 6514 12.8 5.030 5.731 5.030 5.731 init_scf_loop 11 6.9 0.000 0.001 5.435 5.435 dbcsr_mm_accdrv_process 87628 16.1 2.021 2.144 5.104 5.421 cp_dbcsr_syevd 48 12.0 0.004 0.007 5.328 5.329 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.873 4.876 cp_fm_redistribute_end 48 14.0 4.252 4.847 4.256 4.848 cp_fm_diag_elpa_base 48 14.0 0.586 4.703 0.590 4.735 rs_pw_transfer 902 11.9 0.012 0.013 3.513 4.109 wfi_extrapolate 11 7.9 0.001 0.001 3.946 3.946 multiply_cannon_metrocomm1 49320 15.4 0.059 0.063 2.569 3.932 calculate_dm_sparse 110 9.5 0.000 0.001 3.815 3.907 density_rs2pw 110 9.6 0.004 0.005 3.270 3.819 make_m2s 4110 13.4 0.060 0.065 3.680 3.773 make_images 4110 14.4 0.176 0.189 3.586 3.682 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.594 3.598 grid_integrate_task_list 110 12.3 3.239 3.383 3.239 3.383 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.269 3.328 prepare_preconditioner 11 7.9 0.000 0.000 3.299 3.317 make_preconditioner 11 8.9 0.000 0.000 3.299 3.317 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.259 3.290 pw_transfer 1331 11.6 0.054 0.067 3.067 3.124 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.056 3.107 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.979 3.039 jit_kernel_multiply 13 15.9 2.804 2.820 2.804 2.820 calculate_first_density_matrix 1 7.0 0.000 0.000 2.772 2.775 fft_wrap_pw1pw2_140 451 13.1 0.172 0.193 2.549 2.616 fft3d_ps 1111 14.6 0.775 0.882 2.505 2.560 potential_pw2rs 110 12.3 0.006 0.007 2.509 2.529 mp_waitany 14300 13.8 1.757 2.368 1.757 2.368 grid_collocate_task_list 110 9.6 2.084 2.312 2.084 2.312 mp_alltoall_d11v 2046 13.8 1.884 2.297 1.884 2.297 acc_transpose_blocks 49320 15.4 0.218 0.229 2.144 2.242 mp_sum_d 3881 11.9 1.470 1.925 1.470 1.925 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.901 1.916 cp_fm_cholesky_invert 11 10.9 1.795 1.799 1.795 1.799 make_images_data 4110 15.4 0.041 0.044 1.652 1.774 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.727 1.751 hybrid_alltoall_any 4261 16.3 0.081 0.481 1.439 1.679 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.135000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.363636, yerr=3.310389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 590.213120E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2463018. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.031 68.948 68.949 qs_mol_dyn_low 1 2.0 0.005 0.020 68.661 68.670 qs_forces 11 3.9 0.007 0.043 68.477 68.478 qs_energies 11 4.9 0.004 0.026 65.097 65.101 scf_env_do_scf 11 5.9 0.000 0.001 56.336 56.340 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.788 48.788 dbcsr_multiply_generic 2055 12.4 0.114 0.119 37.663 37.836 velocity_verlet 10 3.0 0.004 0.028 36.182 36.185 qs_scf_new_mos 99 7.5 0.001 0.001 32.382 32.504 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.381 32.504 multiply_cannon 2055 13.4 0.221 0.247 31.164 32.371 ot_scf_mini 99 9.5 0.003 0.003 30.735 30.850 multiply_cannon_loop 2055 14.4 0.922 0.956 29.887 30.758 ot_mini 99 10.5 0.001 0.001 18.098 18.226 multiply_cannon_multrec 24660 15.4 7.621 9.379 14.019 15.923 rebuild_ks_matrix 110 8.3 0.000 0.000 13.904 13.973 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 13.903 13.973 qs_ot_get_derivative 99 11.5 0.001 0.002 12.331 12.449 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.248 12.303 mp_waitall_1 176588 16.5 7.549 10.299 7.549 10.299 multiply_cannon_sync_h2d 24660 15.4 7.036 8.036 7.036 8.036 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.119 7.994 init_scf_loop 11 6.9 0.000 0.000 7.514 7.515 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 6.465 7.136 apply_single 110 13.6 0.000 0.001 6.465 7.136 sum_up_and_integrate 110 10.3 0.052 0.061 6.692 6.706 integrate_v_rspace 110 11.3 0.002 0.003 6.639 6.655 dbcsr_mm_accdrv_process 52282 16.1 4.706 5.828 6.238 6.544 init_scf_run 11 5.9 0.000 0.001 6.303 6.304 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.303 6.303 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.185 6.196 calculate_rho_elec 110 8.6 0.040 0.048 6.185 6.196 qs_ot_get_p 110 10.4 0.001 0.001 6.008 6.166 ot_diis_step 99 11.5 0.010 0.012 5.721 5.721 prepare_preconditioner 11 7.9 0.000 0.000 5.451 5.467 make_preconditioner 11 8.9 0.000 0.000 5.451 5.467 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.653 5.388 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.047 5.202 make_m2s 4110 13.4 0.057 0.059 4.137 4.636 make_images 4110 14.4 0.397 0.442 4.029 4.523 qs_ot_p2m_diag 48 11.0 0.028 0.044 4.178 4.198 pw_transfer 1331 11.6 0.066 0.074 3.785 3.925 density_rs2pw 110 9.6 0.004 0.004 3.311 3.898 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.678 3.820 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.752 3.752 rs_pw_transfer 902 11.9 0.012 0.014 2.959 3.572 wfi_extrapolate 11 7.9 0.001 0.002 3.515 3.515 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.341 3.344 grid_integrate_task_list 110 12.3 3.146 3.314 3.146 3.314 fft_wrap_pw1pw2_140 451 13.1 0.203 0.220 3.134 3.283 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.227 3.229 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.145 3.206 cp_fm_redistribute_end 48 14.0 2.415 3.204 2.416 3.205 fft3d_ps 1111 14.6 1.101 1.321 3.029 3.153 cp_fm_diag_elpa_base 48 14.0 0.753 3.062 0.784 3.149 calculate_dm_sparse 110 9.5 0.001 0.001 3.049 3.081 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.995 3.017 make_images_data 4110 15.4 0.046 0.050 2.314 2.821 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.035 2.765 calculate_first_density_matrix 1 7.0 0.000 0.001 2.698 2.699 cp_fm_cholesky_invert 11 10.9 2.527 2.533 2.527 2.533 potential_pw2rs 110 12.3 0.008 0.009 2.475 2.486 grid_collocate_task_list 110 9.6 2.055 2.483 2.055 2.483 mp_sum_l 6514 12.8 1.739 2.358 1.739 2.358 jit_kernel_multiply 11 16.3 1.185 2.319 1.185 2.319 mp_alltoall_d11v 2046 13.8 1.739 2.007 1.739 2.007 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.941 1.960 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.886 1.887 mp_waitany 10164 13.8 1.228 1.869 1.228 1.869 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.819 1.831 multiply_cannon_metrocomm4 22605 15.4 0.073 0.078 0.788 1.708 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.634 1.646 mp_allgather_i34 2055 14.4 0.611 1.623 0.611 1.623 acc_transpose_blocks 24660 15.4 0.111 0.114 1.575 1.602 mp_irecv_dv 57340 16.2 0.664 1.591 0.664 1.591 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.216 0.972 1.576 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.372 1.482 dbcsr_complete_redistribute 325 12.2 0.274 0.328 1.130 1.395 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.949000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.909091, yerr=7.633116 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 659.968000E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.111 59.606 59.607 qs_mol_dyn_low 1 2.0 0.003 0.004 59.167 59.176 qs_forces 11 3.9 0.003 0.007 59.088 59.091 qs_energies 11 4.9 0.002 0.009 55.887 55.893 scf_env_do_scf 11 5.9 0.001 0.002 47.874 47.874 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.446 39.447 velocity_verlet 10 3.0 0.001 0.001 32.261 32.262 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.330 28.615 qs_scf_new_mos 99 7.5 0.001 0.001 24.480 24.596 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.480 24.596 multiply_cannon 2055 13.4 0.211 0.220 22.320 23.687 ot_scf_mini 99 9.5 0.002 0.003 23.271 23.403 multiply_cannon_loop 2055 14.4 0.614 0.627 21.118 22.209 ot_mini 99 10.5 0.001 0.001 13.371 13.509 rebuild_ks_matrix 110 8.3 0.000 0.000 12.345 12.494 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.019 12.345 12.494 multiply_cannon_multrec 16440 15.4 3.944 5.358 9.988 11.327 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.871 11.005 mp_waitall_1 139946 16.5 6.676 9.145 6.676 9.145 qs_ot_get_derivative 99 11.5 0.001 0.001 8.945 9.077 init_scf_loop 11 6.9 0.001 0.004 8.395 8.396 multiply_cannon_metrocomm3 16440 15.4 0.043 0.044 4.184 7.078 prepare_preconditioner 11 7.9 0.000 0.000 6.615 6.630 make_preconditioner 11 8.9 0.000 0.001 6.615 6.630 sum_up_and_integrate 110 10.3 0.059 0.060 6.511 6.524 integrate_v_rspace 110 11.3 0.003 0.003 6.451 6.464 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.940 6.296 dbcsr_mm_accdrv_process 34862 16.1 4.972 5.503 5.898 6.145 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.928 5.939 calculate_rho_elec 110 8.6 0.058 0.059 5.928 5.939 init_scf_run 11 5.9 0.000 0.001 5.618 5.619 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.618 5.618 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.979 5.414 apply_single 110 13.6 0.000 0.000 4.979 5.414 qs_ot_get_p 110 10.4 0.001 0.001 5.127 5.284 ot_diis_step 99 11.5 0.011 0.011 4.400 4.400 make_m2s 4110 13.4 0.049 0.051 3.969 4.398 multiply_cannon_sync_h2d 16440 15.4 3.735 4.381 3.735 4.381 density_rs2pw 110 9.6 0.004 0.005 3.056 4.287 make_images 4110 14.4 0.390 0.509 3.855 4.286 rs_pw_transfer 902 11.9 0.010 0.011 2.616 3.837 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.049 3.696 pw_transfer 1331 11.6 0.065 0.075 3.639 3.649 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.602 3.606 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.533 3.544 grid_integrate_task_list 110 12.3 3.179 3.436 3.179 3.436 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.255 3.256 fft_wrap_pw1pw2_140 451 13.1 0.213 0.218 3.072 3.085 wfi_extrapolate 11 7.9 0.001 0.001 2.919 2.919 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.861 2.862 fft3d_ps 1111 14.6 1.088 1.104 2.832 2.846 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.100 2.824 make_images_data 4110 15.4 0.043 0.048 2.301 2.823 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.761 2.762 cp_fm_redistribute_end 48 14.0 1.725 2.739 1.727 2.739 cp_fm_diag_elpa_base 48 14.0 0.950 2.593 1.009 2.700 calculate_dm_sparse 110 9.5 0.001 0.001 2.590 2.623 calculate_first_density_matrix 1 7.0 0.000 0.003 2.616 2.617 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.534 2.582 cp_fm_cholesky_invert 11 10.9 2.546 2.552 2.546 2.552 grid_collocate_task_list 110 9.6 2.081 2.490 2.081 2.490 mp_waitany 17072 13.8 1.189 2.466 1.189 2.466 multiply_cannon_metrocomm4 14385 15.4 0.044 0.048 0.866 2.463 mp_irecv_dv 48980 15.7 0.797 2.337 0.797 2.337 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.248 2.318 mp_alltoall_d11v 2046 13.8 1.740 2.300 1.740 2.300 potential_pw2rs 110 12.3 0.010 0.010 2.260 2.266 mp_sum_l 6514 12.8 1.471 2.170 1.471 2.170 rs_pw_transfer_RS2PW_140 121 11.5 0.176 0.182 0.894 2.107 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.968 1.971 cp_fm_upper_to_full 70 13.6 1.399 1.863 1.399 1.863 dbcsr_complete_redistribute 325 12.2 0.348 0.378 1.382 1.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.672 1.688 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.558 1.570 cp_fm_cholesky_decompose 22 10.9 1.526 1.547 1.526 1.547 mp_allgather_i34 2055 14.4 0.530 1.540 0.530 1.540 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.378 1.474 rs_gather_matrices 110 12.3 0.232 0.260 0.935 1.451 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.975 1.428 jit_kernel_multiply 6 16.6 0.541 1.291 0.541 1.291 acc_transpose_blocks 16440 15.4 0.072 0.075 1.237 1.253 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.246 1.252 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=59.607000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=624.181818, yerr=9.163348 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 729.931776E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.074 0.168 65.616 65.645 qs_mol_dyn_low 1 2.0 0.005 0.024 65.114 65.128 qs_forces 11 3.9 0.019 0.075 64.963 64.964 qs_energies 11 4.9 0.017 0.082 61.480 61.493 scf_env_do_scf 11 5.9 0.001 0.002 52.865 52.868 scf_env_do_scf_inner_loop 99 6.5 0.002 0.008 41.380 41.381 velocity_verlet 10 3.0 0.003 0.015 36.860 36.866 dbcsr_multiply_generic 2055 12.4 0.115 0.117 29.616 29.915 qs_scf_new_mos 99 7.5 0.001 0.001 26.108 26.199 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.108 26.199 ot_scf_mini 99 9.5 0.003 0.004 24.493 24.579 multiply_cannon 2055 13.4 0.244 0.268 22.831 24.008 multiply_cannon_loop 2055 14.4 0.880 0.900 21.314 22.021 ot_mini 99 10.5 0.001 0.001 14.155 14.263 multiply_cannon_multrec 24660 15.4 4.246 6.697 12.833 14.043 rebuild_ks_matrix 110 8.3 0.000 0.000 12.335 12.435 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.020 12.335 12.435 init_scf_loop 11 6.9 0.005 0.018 11.441 11.441 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.903 10.985 qs_ot_get_derivative 99 11.5 0.001 0.002 10.018 10.108 prepare_preconditioner 11 7.9 0.000 0.000 9.628 9.643 make_preconditioner 11 8.9 0.002 0.010 9.628 9.643 dbcsr_mm_accdrv_process 52304 16.0 6.898 8.281 8.440 9.366 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.868 9.306 sum_up_and_integrate 110 10.3 0.068 0.072 6.559 6.570 integrate_v_rspace 110 11.3 0.003 0.003 6.490 6.504 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.235 6.242 calculate_rho_elec 110 8.6 0.077 0.081 6.234 6.242 mp_waitall_1 121746 16.5 4.248 6.103 4.248 6.103 make_m2s 4110 13.4 0.059 0.061 5.239 5.730 init_scf_run 11 5.9 0.000 0.001 5.657 5.657 scf_env_initial_rho_setup 11 6.9 0.001 0.003 5.657 5.657 make_images 4110 14.4 0.574 0.701 5.102 5.589 qs_ot_get_p 110 10.4 0.001 0.001 5.344 5.491 cp_fm_upper_to_full 70 13.8 3.218 4.576 3.218 4.576 apply_preconditioner_dbcsr 110 12.6 0.000 0.001 4.000 4.112 apply_single 110 13.6 0.000 0.000 4.000 4.112 ot_diis_step 99 11.5 0.011 0.014 4.101 4.102 pw_transfer 1331 11.6 0.065 0.075 3.911 3.940 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.805 3.838 dbcsr_complete_redistribute 325 12.2 0.435 0.492 2.674 3.792 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.610 3.624 density_rs2pw 110 9.6 0.004 0.004 3.110 3.617 grid_integrate_task_list 110 12.3 3.260 3.483 3.260 3.483 multiply_cannon_sync_h2d 24660 15.4 3.203 3.329 3.203 3.329 fft_wrap_pw1pw2_140 451 13.1 0.203 0.213 3.288 3.324 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.262 3.304 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.158 3.254 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.476 3.221 hybrid_alltoall_any 4261 16.3 0.119 0.455 2.319 3.181 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.142 3.142 fft3d_ps 1111 14.6 1.092 1.128 3.097 3.122 make_images_data 4110 15.4 0.046 0.051 2.630 3.101 calculate_dm_sparse 110 9.5 0.001 0.001 2.988 3.019 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.968 3.017 wfi_extrapolate 11 7.9 0.001 0.002 2.982 2.982 rs_pw_transfer 902 11.9 0.010 0.011 2.414 2.950 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.897 2.898 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.749 2.830 mp_alltoall_i22 605 13.7 1.646 2.807 1.646 2.807 cp_fm_cholesky_invert 11 10.9 2.591 2.600 2.591 2.600 calculate_first_density_matrix 1 7.0 0.000 0.002 2.580 2.582 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.566 2.568 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.522 2.555 cp_fm_redistribute_end 48 14.0 1.285 2.544 1.286 2.545 cp_fm_diag_elpa_base 48 14.0 1.180 2.410 1.256 2.521 grid_collocate_task_list 110 9.6 2.183 2.464 2.183 2.464 jit_kernel_multiply 12 15.6 1.212 2.373 1.212 2.373 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.312 2.323 mp_alltoall_d11v 2046 13.8 1.868 2.236 1.868 2.236 potential_pw2rs 110 12.3 0.012 0.013 2.194 2.203 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.601 1.710 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.669 1.704 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.681 1.693 cp_fm_cholesky_decompose 22 10.9 1.604 1.664 1.604 1.664 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.608 1.624 multiply_cannon_metrocomm4 20550 15.4 0.057 0.060 0.844 1.608 mp_allgather_i34 2055 14.4 0.627 1.602 0.627 1.602 mp_waitany 13376 13.8 1.087 1.600 1.087 1.600 acc_transpose_blocks 24660 15.4 0.105 0.106 1.561 1.585 mp_irecv_dv 62702 16.1 0.745 1.530 0.745 1.530 mp_sum_l 6514 12.8 0.883 1.471 0.883 1.471 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.176 0.855 1.397 qs_env_update_s_mstruct 11 6.9 0.000 0.001 1.267 1.373 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=65.645000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=691.636364, yerr=9.772622 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 825.225216E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 55.015 55.016 qs_mol_dyn_low 1 2.0 0.003 0.003 54.729 54.738 qs_forces 11 3.9 0.002 0.002 54.642 54.644 qs_energies 11 4.9 0.001 0.001 50.962 50.968 scf_env_do_scf 11 5.9 0.001 0.001 42.662 42.662 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.103 35.104 velocity_verlet 10 3.0 0.001 0.001 30.772 30.776 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.698 22.894 qs_scf_new_mos 99 7.5 0.001 0.001 19.948 19.997 qs_scf_loop_do_ot 99 8.5 0.001 0.001 19.948 19.996 ot_scf_mini 99 9.5 0.002 0.002 18.722 18.744 multiply_cannon 2055 13.4 0.246 0.258 17.436 18.709 multiply_cannon_loop 2055 14.4 0.319 0.330 16.144 16.545 rebuild_ks_matrix 110 8.3 0.000 0.000 11.990 12.027 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.990 12.027 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.652 10.684 ot_mini 99 10.5 0.001 0.001 10.307 10.330 multiply_cannon_multrec 8220 15.4 3.315 4.586 7.697 8.666 init_scf_loop 11 6.9 0.000 0.000 7.513 7.514 mp_waitall_1 103326 16.6 5.799 7.388 5.799 7.388 qs_ot_get_derivative 99 11.5 0.001 0.001 6.561 6.583 sum_up_and_integrate 110 10.3 0.079 0.081 6.549 6.559 integrate_v_rspace 110 11.3 0.003 0.003 6.470 6.480 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.333 6.349 calculate_rho_elec 110 8.6 0.115 0.116 6.333 6.348 prepare_preconditioner 11 7.9 0.000 0.000 5.845 5.848 make_preconditioner 11 8.9 0.000 0.000 5.845 5.848 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.420 5.497 dbcsr_mm_accdrv_process 17442 15.9 2.869 4.196 4.251 5.294 init_scf_run 11 5.9 0.000 0.001 5.178 5.178 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.177 5.177 qs_ot_get_p 110 10.4 0.001 0.001 4.544 4.574 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.998 4.408 make_m2s 4110 13.4 0.038 0.040 4.051 4.298 make_images 4110 14.4 0.636 0.694 3.922 4.173 pw_transfer 1331 11.6 0.066 0.071 4.113 4.123 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.006 4.020 ot_diis_step 99 11.5 0.012 0.012 3.727 3.727 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.645 3.668 apply_single 110 13.6 0.000 0.000 3.645 3.668 grid_integrate_task_list 110 12.3 3.356 3.543 3.356 3.543 fft_wrap_pw1pw2_140 451 13.1 0.216 0.219 3.485 3.502 density_rs2pw 110 9.6 0.004 0.004 3.085 3.458 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.293 3.296 fft3d_ps 1111 14.6 1.143 1.162 3.251 3.265 cp_dbcsr_syevd 48 12.0 0.004 0.004 2.996 2.997 multiply_cannon_sync_h2d 8220 15.4 2.927 2.976 2.927 2.976 cp_fm_cholesky_invert 11 10.9 2.776 2.779 2.776 2.779 wfi_extrapolate 11 7.9 0.001 0.001 2.694 2.694 make_images_data 4110 15.4 0.038 0.043 2.316 2.691 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.681 2.683 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.244 2.683 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.564 2.565 calculate_dm_sparse 110 9.5 0.001 0.001 2.504 2.544 grid_collocate_task_list 110 9.6 2.279 2.539 2.279 2.539 rs_pw_transfer 902 11.9 0.010 0.010 2.132 2.534 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.499 2.500 cp_fm_redistribute_end 48 14.0 0.635 2.477 0.639 2.477 cp_fm_diag_elpa_base 48 14.0 1.663 2.299 1.833 2.453 calculate_first_density_matrix 1 7.0 0.000 0.000 2.388 2.389 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.236 2.244 potential_pw2rs 110 12.3 0.015 0.016 2.094 2.099 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.985 mp_alltoall_d11v 2046 13.8 1.572 1.871 1.572 1.871 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.858 1.871 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.705 1.717 jit_kernel_multiply 9 15.8 1.076 1.716 1.076 1.716 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.670 1.675 cp_fm_cholesky_decompose 22 10.9 1.645 1.657 1.645 1.657 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.520 1.647 mp_allgather_i34 2055 14.4 0.483 1.544 0.483 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.478 1.487 dbcsr_complete_redistribute 325 12.2 0.567 0.593 1.395 1.486 mp_waitany 9240 13.8 1.060 1.465 1.060 1.465 qs_create_task_list 11 7.9 0.000 0.001 1.225 1.328 generate_qs_task_list 11 8.9 0.379 0.448 1.224 1.328 rs_gather_matrices 110 12.3 0.325 0.364 0.931 1.210 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.792 1.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=55.016000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.272727, yerr=8.486255 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 1.386721E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.075 87.305 87.314 qs_mol_dyn_low 1 2.0 0.003 0.003 86.950 86.959 qs_forces 11 3.9 0.002 0.003 86.642 86.643 qs_energies 11 4.9 0.002 0.002 82.471 82.473 scf_env_do_scf 11 5.9 0.001 0.001 72.203 72.204 velocity_verlet 10 3.0 0.001 0.001 55.882 55.918 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 43.852 43.853 dbcsr_multiply_generic 2055 12.4 0.118 0.119 28.274 28.354 init_scf_loop 11 6.9 0.000 0.000 28.281 28.283 prepare_preconditioner 11 7.9 0.000 0.000 26.265 26.271 make_preconditioner 11 8.9 0.000 0.000 26.264 26.271 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.296 25.728 qs_scf_new_mos 99 7.5 0.001 0.001 25.488 25.530 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.488 25.529 ot_scf_mini 99 9.5 0.002 0.002 23.763 23.788 multiply_cannon 2055 13.4 0.338 0.373 21.447 22.268 multiply_cannon_loop 2055 14.4 0.339 0.345 19.700 19.999 cp_fm_upper_to_full 70 14.2 12.716 18.398 12.716 18.398 rebuild_ks_matrix 110 8.3 0.001 0.001 14.161 14.189 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 14.160 14.188 ot_mini 99 10.5 0.001 0.001 13.240 13.264 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.835 12.860 dbcsr_complete_redistribute 325 12.2 1.023 1.042 7.545 10.755 multiply_cannon_multrec 8220 15.4 4.371 4.572 9.707 9.812 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.517 9.717 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.953 9.120 qs_ot_get_derivative 99 11.5 0.001 0.001 8.799 8.823 mp_alltoall_i22 605 13.7 5.558 8.770 5.558 8.770 mp_waitall_1 84994 16.7 7.232 8.062 7.232 8.062 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.788 7.824 calculate_rho_elec 110 8.6 0.227 0.227 7.788 7.823 sum_up_and_integrate 110 10.3 0.150 0.151 7.354 7.366 integrate_v_rspace 110 11.3 0.004 0.004 7.203 7.217 init_scf_run 11 5.9 0.000 0.001 5.994 5.995 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.993 5.995 make_m2s 4110 13.4 0.043 0.043 5.170 5.666 make_images 4110 14.4 0.876 0.933 4.983 5.477 qs_ot_get_p 110 10.4 0.001 0.001 5.411 5.438 dbcsr_mm_accdrv_process 11614 15.7 3.346 3.637 5.194 5.406 pw_transfer 1331 11.6 0.075 0.076 5.306 5.311 fft_wrap_pw1pw2 1111 12.6 0.009 0.010 5.189 5.193 cp_fm_cholesky_invert 11 10.9 5.084 5.088 5.084 5.088 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.574 5.044 apply_single 110 13.6 0.000 0.000 4.574 5.044 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.565 4.928 fft_wrap_pw1pw2_140 451 13.1 0.221 0.222 4.528 4.534 ot_diis_step 99 11.5 0.015 0.015 4.425 4.425 fft3d_ps 1111 14.6 1.294 1.311 4.383 4.394 multiply_cannon_sync_h2d 8220 15.4 3.950 3.956 3.950 3.956 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.928 3.937 density_rs2pw 110 9.6 0.004 0.004 3.827 3.867 grid_integrate_task_list 110 12.3 3.665 3.721 3.665 3.721 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.680 3.681 hybrid_alltoall_any 4261 16.3 0.257 0.554 2.815 3.532 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.477 3.483 make_images_data 4110 15.4 0.041 0.044 2.798 3.393 wfi_extrapolate 11 7.9 0.001 0.001 3.299 3.300 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.825 3.274 calculate_dm_sparse 110 9.5 0.001 0.001 3.250 3.269 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.915 2.918 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.881 2.881 cp_fm_diag_elpa_base 48 14.0 2.348 2.547 2.879 2.879 grid_collocate_task_list 110 9.6 2.635 2.668 2.635 2.668 calculate_first_density_matrix 1 7.0 0.000 0.000 2.573 2.574 potential_pw2rs 110 12.3 0.021 0.022 2.484 2.487 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.431 2.441 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.318 2.331 rs_pw_transfer 902 11.9 0.011 0.011 2.257 2.279 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.208 2.268 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.093 2.184 cp_fm_cholesky_decompose 22 10.9 1.969 2.000 1.969 2.000 mp_alltoall_d11v 2046 13.8 1.904 1.999 1.904 1.999 qs_create_task_list 11 7.9 0.001 0.001 1.893 1.941 generate_qs_task_list 11 8.9 0.735 0.788 1.892 1.940 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.926 1.929 jit_kernel_multiply 10 15.2 1.650 1.899 1.650 1.899 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.819 1.832 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.747 1.782 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.314000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1230.909091, yerr=70.341317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 629.755904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56460. MP_Allreduce 11308 783. MP_Sync 170 MP_Alltoall 2242 2521907. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_comm_split 83 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.064 204.939 204.945 qs_mol_dyn_low 1 2.0 0.004 0.007 204.502 204.516 qs_forces 11 3.9 0.004 0.009 204.419 204.420 qs_energies 11 4.9 0.001 0.002 198.822 198.837 scf_env_do_scf 11 5.9 0.001 0.001 182.214 182.218 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 161.829 161.831 dbcsr_multiply_generic 2527 12.6 0.179 0.183 125.087 125.871 velocity_verlet 10 3.0 0.001 0.001 124.195 124.196 qs_scf_new_mos 118 7.6 0.001 0.001 121.370 121.683 qs_scf_loop_do_ot 118 8.6 0.001 0.001 121.369 121.683 ot_scf_mini 118 9.6 0.003 0.003 114.749 115.024 multiply_cannon 2527 13.6 0.238 0.247 101.907 104.395 multiply_cannon_loop 2527 14.6 2.101 2.189 99.731 102.079 ot_mini 118 10.6 0.001 0.001 66.103 66.419 multiply_cannon_multrec 60648 15.6 33.635 35.904 42.213 44.531 qs_ot_get_derivative 118 11.6 0.001 0.001 40.960 41.250 rebuild_ks_matrix 129 8.3 0.001 0.001 34.078 34.748 qs_ks_build_kohn_sham_matrix 129 9.3 0.016 0.018 34.077 34.747 mp_waitall_1 269254 16.5 28.294 32.126 28.294 32.126 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.658 31.282 multiply_cannon_sync_h2d 60648 15.6 27.567 30.061 27.567 30.061 qs_ot_get_p 129 10.4 0.001 0.001 26.494 26.766 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.484 25.115 apply_single 129 13.6 0.001 0.001 24.484 25.114 ot_diis_step 118 11.6 0.008 0.008 24.781 24.782 init_scf_loop 11 6.9 0.000 0.001 20.310 20.311 qs_ot_p2m_diag 83 11.4 0.077 0.091 19.873 19.959 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.873 19.062 multiply_cannon_metrocomm3 60648 15.6 0.115 0.121 15.784 18.140 cp_dbcsr_syevd 83 12.4 0.005 0.006 17.315 17.316 prepare_preconditioner 11 7.9 0.000 0.000 15.671 15.717 make_preconditioner 11 8.9 0.000 0.000 15.671 15.717 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.903 15.062 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.552 14.559 cp_fm_redistribute_end 83 14.4 11.511 14.490 11.521 14.493 sum_up_and_integrate 129 10.3 0.090 0.109 14.339 14.352 integrate_v_rspace 129 11.3 0.003 0.004 14.248 14.266 cp_fm_diag_elpa_base 83 14.4 2.928 14.131 2.959 14.245 make_m2s 5054 13.6 0.105 0.115 13.636 14.153 make_images 5054 14.6 0.400 0.423 13.453 13.980 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.787 13.907 calculate_rho_elec 129 8.7 0.046 0.065 13.786 13.907 init_scf_run 11 5.9 0.000 0.001 12.435 12.436 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.435 12.436 density_rs2pw 129 9.7 0.007 0.007 7.089 10.341 mp_sum_l 7930 13.1 8.287 9.818 8.287 9.818 wfi_extrapolate 11 7.9 0.001 0.001 9.041 9.041 cp_fm_cholesky_invert 11 10.9 8.920 8.928 8.920 8.928 rs_pw_transfer 1054 12.0 0.016 0.018 5.579 8.853 dbcsr_mm_accdrv_process 125468 16.2 3.238 3.467 8.131 8.746 calculate_dm_sparse 129 9.5 0.001 0.001 8.530 8.621 multiply_cannon_metrocomm1 60648 15.6 0.087 0.093 6.152 8.504 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.114 8.302 pw_transfer 1559 11.6 0.076 0.104 8.032 8.280 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 7.826 8.079 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.890 7.981 grid_integrate_task_list 129 12.3 7.083 7.715 7.083 7.715 make_images_data 5054 15.6 0.065 0.073 6.702 7.667 hybrid_alltoall_any 5240 16.5 0.291 2.263 5.867 7.282 fft_wrap_pw1pw2_140 527 13.2 0.451 0.521 6.870 7.187 fft3d_ps 1301 14.7 2.122 2.920 6.601 6.776 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.733 6.742 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.849 5.956 mp_waitany 16140 13.9 2.648 5.880 2.648 5.880 grid_collocate_task_list 129 9.7 4.601 5.854 4.601 5.854 mp_alltoall_d11v 2423 14.1 4.265 5.768 4.265 5.768 rs_pw_transfer_RS2PW_140 140 11.5 0.281 0.297 2.131 5.419 potential_pw2rs 129 12.3 0.009 0.011 4.628 4.644 cp_fm_cholesky_decompose 22 10.9 4.570 4.586 4.570 4.586 mp_sum_d 4491 12.2 3.691 4.326 3.691 4.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=204.945000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.545455, yerr=6.760361 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 830.599168E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57200. MP_Allreduce 11252 947. MP_Sync 168 MP_Alltoall 1955 5620331. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_comm_split 82 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 189.018 189.018 qs_mol_dyn_low 1 2.0 0.003 0.005 188.601 188.614 qs_forces 11 3.9 0.003 0.003 188.518 188.519 qs_energies 11 4.9 0.002 0.002 181.697 181.708 scf_env_do_scf 11 5.9 0.001 0.001 165.222 165.232 scf_env_do_scf_inner_loop 116 6.6 0.002 0.008 132.725 132.728 velocity_verlet 10 3.0 0.001 0.001 118.334 118.335 dbcsr_multiply_generic 2485 12.5 0.184 0.189 96.705 97.853 qs_scf_new_mos 116 7.6 0.001 0.001 92.824 93.334 qs_scf_loop_do_ot 116 8.6 0.001 0.001 92.823 93.333 ot_scf_mini 116 9.6 0.004 0.004 88.082 88.668 multiply_cannon 2485 13.5 0.474 0.531 77.083 81.622 multiply_cannon_loop 2485 14.5 1.237 1.286 73.879 76.248 ot_mini 116 10.6 0.001 0.001 49.192 49.715 mp_waitall_1 212858 16.6 24.096 38.218 24.096 38.218 multiply_cannon_multrec 29820 15.5 21.888 26.206 31.688 36.243 rebuild_ks_matrix 127 8.3 0.001 0.001 33.241 33.714 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.241 33.714 init_scf_loop 11 6.9 0.000 0.000 32.400 32.401 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.881 30.320 qs_ot_get_derivative 116 11.6 0.001 0.002 27.511 28.067 prepare_preconditioner 11 7.9 0.000 0.000 27.985 28.058 make_preconditioner 11 8.9 0.000 0.000 27.985 28.058 multiply_cannon_metrocomm3 29820 15.5 0.094 0.099 15.550 27.777 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.705 27.258 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.770 22.830 apply_single 127 13.6 0.001 0.001 21.770 22.830 multiply_cannon_sync_h2d 29820 15.5 19.222 21.743 19.222 21.743 qs_ot_get_p 127 10.4 0.001 0.001 20.801 21.559 ot_diis_step 116 11.6 0.014 0.015 21.508 21.511 cp_fm_cholesky_invert 11 10.9 16.378 16.390 16.378 16.390 qs_ot_p2m_diag 82 11.4 0.185 0.213 16.074 16.108 make_m2s 4970 13.5 0.089 0.094 14.000 15.581 make_images 4970 14.5 1.139 1.329 13.791 15.370 cp_dbcsr_syevd 82 12.4 0.006 0.006 14.961 14.962 sum_up_and_integrate 127 10.3 0.115 0.132 14.681 14.709 integrate_v_rspace 127 11.3 0.003 0.004 14.566 14.596 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.017 14.052 calculate_rho_elec 127 8.7 0.087 0.105 14.016 14.051 cp_fm_diag_elpa 82 13.4 0.000 0.001 11.954 11.960 cp_fm_redistribute_end 82 14.4 6.980 11.891 6.994 11.895 cp_fm_diag_elpa_base 82 14.4 4.673 11.401 4.886 11.757 init_scf_run 11 5.9 0.000 0.001 11.713 11.714 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.713 11.714 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.736 11.146 make_images_data 4970 15.5 0.062 0.072 8.547 10.491 multiply_cannon_metrocomm4 27335 15.5 0.094 0.108 3.729 10.373 mp_irecv_dv 68888 16.3 3.537 10.001 3.537 10.001 hybrid_alltoall_any 5155 16.4 0.338 1.459 7.272 9.885 density_rs2pw 127 9.7 0.006 0.007 7.445 9.848 dbcsr_mm_accdrv_process 61726 16.2 4.568 5.382 9.262 9.846 pw_transfer 1535 11.6 0.086 0.110 8.987 9.075 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.764 8.852 wfi_extrapolate 11 7.9 0.001 0.001 8.367 8.367 rs_pw_transfer 1038 11.9 0.014 0.016 5.547 7.961 fft_wrap_pw1pw2_140 519 13.2 0.470 0.526 7.756 7.856 grid_integrate_task_list 127 12.3 7.120 7.481 7.120 7.481 fft3d_ps 1281 14.7 2.729 2.907 7.141 7.197 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.192 6.985 cp_fm_cholesky_decompose 22 10.9 6.797 6.879 6.797 6.879 calculate_dm_sparse 127 9.5 0.001 0.001 6.484 6.645 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.248 6.259 grid_collocate_task_list 127 9.7 4.666 5.817 4.666 5.817 mp_sum_l 7804 13.0 3.846 5.698 3.846 5.698 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.406 5.552 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.256 5.323 potential_pw2rs 127 12.3 0.014 0.016 4.990 5.005 mp_alltoall_d11v 2401 14.1 4.044 4.940 4.044 4.940 mp_allgather_i34 2485 14.5 1.799 4.864 1.799 4.864 mp_waitany 11660 13.9 2.440 4.832 2.440 4.832 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.374 2.051 4.436 mp_sum_d 4444 12.1 2.619 3.927 2.619 3.927 dbcsr_complete_redistribute 393 12.7 0.789 0.869 3.056 3.900 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.762 3.805 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=189.018000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=790.000000, yerr=4.670994 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/17/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 2.928533E+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 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 931.241984E+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 931530938576 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57335. MP_Allreduce 11226 986. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 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 0.041 0.128 176.144 176.150 qs_mol_dyn_low 1 2.0 0.003 0.004 175.583 175.596 qs_forces 11 3.9 0.003 0.006 175.467 175.476 qs_energies 11 4.9 0.002 0.006 168.854 168.870 scf_env_do_scf 11 5.9 0.002 0.012 153.233 153.234 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.563 118.565 velocity_verlet 10 3.0 0.001 0.001 112.767 112.769 dbcsr_multiply_generic 2507 12.6 0.182 0.187 81.314 82.502 qs_scf_new_mos 117 7.6 0.001 0.001 80.215 80.506 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.214 80.505 ot_scf_mini 117 9.6 0.004 0.005 76.089 76.408 multiply_cannon 2507 13.6 0.502 0.526 62.007 66.638 multiply_cannon_loop 2507 14.6 0.852 0.876 58.796 61.294 ot_mini 117 10.6 0.001 0.001 42.441 42.771 init_scf_loop 11 6.9 0.001 0.004 34.570 34.571 mp_waitall_1 170520 16.6 24.445 32.725 24.445 32.725 rebuild_ks_matrix 128 8.3 0.001 0.001 30.760 31.187 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.020 30.759 31.186 prepare_preconditioner 11 7.9 0.000 0.000 30.484 30.526 make_preconditioner 11 8.9 0.000 0.001 30.484 30.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.140 29.540 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.722 28.103 multiply_cannon_multrec 20056 15.6 13.578 16.251 22.485 25.108 multiply_cannon_metrocomm3 20056 15.6 0.059 0.064 15.083 24.380 qs_ot_get_derivative 117 11.6 0.002 0.002 22.797 23.127 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.794 20.785 apply_single 128 13.6 0.001 0.001 19.794 20.785 qs_ot_get_p 128 10.4 0.001 0.001 19.246 19.655 ot_diis_step 117 11.6 0.018 0.018 19.546 19.547 multiply_cannon_sync_h2d 20056 15.6 14.244 15.803 14.244 15.803 make_m2s 5014 13.6 0.080 0.086 14.303 15.489 make_images 5014 14.6 1.178 1.262 14.070 15.258 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.949 14.956 sum_up_and_integrate 128 10.3 0.134 0.145 14.700 14.724 integrate_v_rspace 128 11.3 0.004 0.005 14.565 14.592 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.409 14.432 calculate_rho_elec 128 8.7 0.132 0.147 14.409 14.432 cp_fm_cholesky_invert 11 10.9 14.071 14.079 14.071 14.079 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.964 13.966 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.863 10.863 cp_fm_redistribute_end 83 14.4 4.106 10.807 4.119 10.809 cp_fm_diag_elpa_base 83 14.4 6.267 10.231 6.670 10.699 init_scf_run 11 5.9 0.000 0.001 10.554 10.556 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.554 10.555 make_images_data 5014 15.6 0.060 0.068 8.711 10.389 hybrid_alltoall_any 5200 16.5 0.434 1.975 7.634 9.557 pw_transfer 1547 11.6 0.086 0.108 9.158 9.288 density_rs2pw 128 9.7 0.006 0.006 7.273 9.221 multiply_cannon_metrocomm4 17549 15.6 0.060 0.070 3.420 9.152 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.861 9.096 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.933 9.074 mp_irecv_dv 50230 16.2 3.300 8.907 3.300 8.907 dbcsr_mm_accdrv_process 41502 16.2 4.611 5.434 8.353 8.494 fft_wrap_pw1pw2_140 523 13.2 0.479 0.528 7.927 8.077 grid_integrate_task_list 128 12.3 7.249 7.763 7.249 7.763 cp_fm_upper_to_full 105 14.5 5.790 7.501 5.790 7.501 wfi_extrapolate 11 7.9 0.001 0.001 7.359 7.360 fft3d_ps 1291 14.7 2.666 2.915 7.196 7.297 cp_fm_cholesky_decompose 22 10.9 7.246 7.275 7.246 7.275 rs_pw_transfer 1046 11.9 0.014 0.015 5.088 7.105 dbcsr_complete_redistribute 395 12.7 1.168 1.194 4.525 6.269 calculate_dm_sparse 128 9.5 0.001 0.001 5.838 5.967 grid_collocate_task_list 128 9.7 4.900 5.763 4.900 5.763 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.468 5.473 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.614 5.277 copy_fm_to_dbcsr 209 11.7 0.002 0.003 3.370 5.110 mp_alltoall_d11v 2415 14.1 4.331 5.026 4.331 5.026 potential_pw2rs 128 12.3 0.020 0.022 4.778 4.790 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.595 4.716 mp_sum_l 7870 13.0 3.228 4.599 3.228 4.599 mp_waitany 11748 13.9 2.349 4.407 2.349 4.407 mp_allgather_i34 2507 14.6 1.745 4.219 1.745 4.219 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.324 4.042 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.974 4.006 rs_pw_transfer_RS2PW_140 139 11.5 0.328 0.349 1.951 3.937 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 3.818 3.825 mp_alltoall_i22 716 14.1 1.949 3.809 1.949 3.809 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.548 3.593 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.150000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=882.090909, yerr=10.281975 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 4.387043E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 6024576 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.8 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.131254E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1152312 MPI messages size (bytes): total size 2.039354E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.769794E+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 321600 36333158400 131072 < size <= 4194304 721560 792041881600 4194304 < size <= 16777216 70800 669921260160 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4107 56895. MP_Allreduce 11306 1065. MP_Sync 170 MP_Alltoall 1724 12509439. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_comm_split 83 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.047 0.108 190.295 190.308 qs_mol_dyn_low 1 2.0 0.003 0.004 189.643 189.656 qs_forces 11 3.9 0.009 0.028 189.520 189.525 qs_energies 11 4.9 0.044 0.089 182.259 182.273 scf_env_do_scf 11 5.9 0.004 0.007 165.098 165.106 velocity_verlet 10 3.0 0.001 0.001 125.952 125.955 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 118.670 118.673 dbcsr_multiply_generic 2527 12.6 0.192 0.198 80.038 80.703 qs_scf_new_mos 118 7.6 0.001 0.001 80.417 80.695 qs_scf_loop_do_ot 118 8.6 0.001 0.001 80.417 80.694 ot_scf_mini 118 9.6 0.004 0.004 75.748 76.039 multiply_cannon 2527 13.6 0.563 0.590 55.292 57.558 multiply_cannon_loop 2527 14.6 1.184 1.210 51.282 52.824 init_scf_loop 11 6.9 0.010 0.029 46.301 46.305 ot_mini 118 10.6 0.001 0.001 42.588 42.870 prepare_preconditioner 11 7.9 0.000 0.001 42.129 42.152 make_preconditioner 11 8.9 0.005 0.016 42.129 42.152 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.736 40.752 multiply_cannon_multrec 30324 15.6 14.247 20.350 26.412 32.008 rebuild_ks_matrix 129 8.3 0.001 0.001 30.249 30.541 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.022 30.249 30.540 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.234 27.503 mp_waitall_1 149060 16.7 17.175 27.258 17.175 27.258 qs_ot_get_derivative 118 11.6 0.002 0.002 22.852 23.150 make_m2s 5054 13.6 0.096 0.100 20.398 21.411 make_images 5054 14.6 1.957 2.252 20.100 21.111 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.175 19.677 apply_single 129 13.6 0.001 0.001 19.175 19.677 qs_ot_get_p 129 10.4 0.001 0.002 19.331 19.646 ot_diis_step 118 11.6 0.017 0.019 19.605 19.606 cp_fm_upper_to_full 105 14.7 11.161 16.483 11.161 16.483 cp_fm_cholesky_invert 11 10.9 16.006 16.014 16.006 16.014 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.159 15.211 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.079 15.119 calculate_rho_elec 129 8.7 0.176 0.192 15.078 15.118 multiply_cannon_metrocomm3 30324 15.6 0.046 0.048 6.377 15.075 sum_up_and_integrate 129 10.3 0.142 0.156 14.937 14.981 integrate_v_rspace 129 11.3 0.004 0.004 14.794 14.843 cp_dbcsr_syevd 83 12.4 0.006 0.006 13.896 13.898 make_images_data 5054 15.6 0.064 0.069 11.003 12.985 dbcsr_complete_redistribute 395 12.7 1.511 1.618 9.028 12.820 multiply_cannon_sync_h2d 30324 15.6 11.688 12.398 11.688 12.398 dbcsr_mm_accdrv_process 62756 16.2 7.406 8.617 11.728 12.198 hybrid_alltoall_any 5240 16.5 0.529 2.213 9.864 11.700 copy_fm_to_dbcsr 209 11.7 0.002 0.003 7.686 11.484 init_scf_run 11 5.9 0.000 0.001 10.939 10.940 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.938 10.940 cp_fm_diag_elpa 83 13.4 0.000 0.001 10.693 10.696 cp_fm_redistribute_end 83 14.4 1.844 10.624 1.856 10.629 cp_fm_diag_elpa_base 83 14.4 8.168 10.015 8.742 10.510 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.365 10.064 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.414 9.614 pw_transfer 1559 11.6 0.088 0.108 9.525 9.603 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.295 9.383 mp_alltoall_i22 718 14.1 5.626 9.359 5.626 9.359 density_rs2pw 129 9.7 0.006 0.006 7.120 8.560 fft_wrap_pw1pw2_140 527 13.2 0.487 0.497 8.255 8.363 grid_integrate_task_list 129 12.3 7.530 7.948 7.530 7.948 wfi_extrapolate 11 7.9 0.001 0.001 7.686 7.686 fft3d_ps 1301 14.7 2.798 2.911 7.509 7.578 cp_fm_cholesky_decompose 22 10.9 7.471 7.566 7.471 7.566 multiply_cannon_metrocomm4 25270 15.6 0.076 0.086 2.865 7.074 mp_irecv_dv 76685 16.2 2.717 6.811 2.717 6.811 calculate_dm_sparse 129 9.5 0.001 0.001 6.417 6.516 mp_alltoall_d11v 2423 14.1 5.349 6.211 5.349 6.211 rs_pw_transfer 1054 12.0 0.013 0.015 4.753 6.170 grid_collocate_task_list 129 9.7 5.117 5.829 5.117 5.829 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.550 5.599 potential_pw2rs 129 12.3 0.022 0.023 4.702 4.729 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.615 4.617 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.505 4.597 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.477 4.554 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.248 4.317 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=190.308000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1073.272727, yerr=14.967182 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 5.909858E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1976192 0.0% 0.0% 100.0% average stack size 0.0 0.0 3444.2 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 1.503351E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242592 MPI messages size (bytes): total size 1.341705E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.530704E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 114816 60196651008 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726591885696 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9003 51. MP_Alltoall 9708 795382. MP_ISend 40388 2101966. MP_IRecv 40388 2101070. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57813. MP_Allreduce 11072 1172. MP_Sync 87 MP_Alltoall 1724 18848449. MP_SendRecv 3870 122880. MP_ISendRecv 3870 122880. MP_Wait 16244 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 173.082 173.084 qs_mol_dyn_low 1 2.0 0.003 0.004 172.641 172.654 qs_forces 11 3.9 0.003 0.003 172.537 172.540 qs_energies 11 4.9 0.003 0.004 164.944 164.954 scf_env_do_scf 11 5.9 0.001 0.001 147.728 147.732 velocity_verlet 10 3.0 0.001 0.001 113.802 113.806 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 112.403 112.404 dbcsr_multiply_generic 2527 12.6 0.183 0.189 74.190 74.824 qs_scf_new_mos 118 7.6 0.001 0.001 73.310 73.418 qs_scf_loop_do_ot 118 8.6 0.001 0.001 73.309 73.417 ot_scf_mini 118 9.6 0.003 0.004 68.865 68.970 multiply_cannon 2527 13.6 0.591 0.619 54.613 58.683 multiply_cannon_loop 2527 14.6 0.446 0.462 50.026 51.156 ot_mini 118 10.6 0.001 0.001 39.618 39.732 init_scf_loop 11 6.9 0.000 0.000 35.180 35.183 mp_waitall_1 126780 16.7 25.424 33.211 25.424 33.211 prepare_preconditioner 11 7.9 0.000 0.000 31.266 31.289 make_preconditioner 11 8.9 0.000 0.000 31.266 31.289 rebuild_ks_matrix 129 8.3 0.001 0.001 30.207 30.336 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 30.207 30.335 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.135 29.392 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.445 27.565 multiply_cannon_multrec 10108 15.6 10.491 14.423 18.122 21.475 ot_diis_step 118 11.6 0.020 0.021 19.944 19.945 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.434 19.905 apply_single 129 13.6 0.001 0.001 19.434 19.904 qs_ot_get_derivative 118 11.6 0.002 0.002 19.606 19.712 multiply_cannon_metrocomm3 10108 15.6 0.023 0.025 12.189 19.579 make_m2s 5054 13.6 0.066 0.070 15.978 18.056 cp_fm_cholesky_invert 11 10.9 17.813 17.819 17.813 17.819 make_images 5054 14.6 2.285 2.811 15.669 17.741 qs_ot_get_p 129 10.4 0.001 0.001 16.373 16.514 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.547 15.589 calculate_rho_elec 129 8.7 0.261 0.270 15.546 15.588 sum_up_and_integrate 129 10.3 0.181 0.191 15.219 15.266 integrate_v_rspace 129 11.3 0.004 0.004 15.037 15.093 qs_ot_p2m_diag 83 11.4 0.495 0.501 12.704 12.722 multiply_cannon_sync_h2d 10108 15.6 11.732 12.522 11.732 12.522 make_images_data 5054 15.6 0.052 0.060 9.733 12.101 hybrid_alltoall_any 5240 16.5 0.828 3.648 9.554 12.016 cp_dbcsr_syevd 83 12.4 0.006 0.007 11.597 11.599 init_scf_run 11 5.9 0.000 0.001 10.442 10.442 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.442 10.442 pw_transfer 1559 11.6 0.086 0.094 9.937 9.975 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 9.712 9.758 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.635 8.647 fft_wrap_pw1pw2_140 527 13.2 0.500 0.520 8.596 8.642 cp_fm_diag_elpa_base 83 14.4 8.396 8.470 8.624 8.637 grid_integrate_task_list 129 12.3 7.786 8.207 7.786 8.207 density_rs2pw 129 9.7 0.005 0.006 7.052 7.936 dbcsr_mm_accdrv_process 20926 16.1 2.792 3.583 7.256 7.922 cp_fm_cholesky_decompose 22 10.9 7.776 7.883 7.776 7.883 fft3d_ps 1301 14.7 2.725 2.800 7.859 7.879 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.762 7.843 multiply_cannon_metrocomm1 10108 15.6 0.029 0.029 4.288 7.525 wfi_extrapolate 11 7.9 0.001 0.001 7.315 7.315 mp_allgather_i34 2527 14.6 2.796 6.961 2.796 6.961 calculate_dm_sparse 129 9.5 0.001 0.001 6.158 6.254 grid_collocate_task_list 129 9.7 5.438 6.118 5.438 6.118 mp_alltoall_d11v 2423 14.1 5.076 6.094 5.076 6.094 dbcsr_complete_redistribute 395 12.7 2.124 2.177 5.095 5.463 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.281 5.282 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.211 5.223 rs_pw_transfer 1054 12.0 0.013 0.013 4.065 4.985 potential_pw2rs 129 12.3 0.027 0.028 4.558 4.572 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.158 4.196 multiply_cannon_metrocomm4 7581 15.6 0.024 0.027 1.859 4.004 mp_irecv_dv 29086 15.9 1.821 3.928 1.821 3.928 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.567 3.881 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.697 3.747 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.673 3.687 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 3.628 3.665 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.349 3.637 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.498 3.579 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.084000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1400.090909, yerr=46.504065 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 2.982064E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 59070. MP_Allreduce 11015 1514. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.052 0.119 294.458 294.485 qs_mol_dyn_low 1 2.0 0.003 0.005 293.801 293.813 qs_forces 11 3.9 0.003 0.003 293.708 293.713 qs_energies 11 4.9 0.002 0.002 284.267 284.285 scf_env_do_scf 11 5.9 0.001 0.001 261.723 261.730 velocity_verlet 10 3.0 0.001 0.001 212.014 212.021 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 138.348 138.350 init_scf_loop 11 6.9 0.000 0.000 123.119 123.124 prepare_preconditioner 11 7.9 0.000 0.000 117.951 117.977 make_preconditioner 11 8.9 0.000 0.000 117.951 117.977 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.355 115.138 qs_scf_new_mos 116 7.6 0.001 0.001 88.323 88.434 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.322 88.433 ot_scf_mini 116 9.6 0.004 0.004 83.644 83.677 dbcsr_multiply_generic 2485 12.5 0.211 0.218 82.240 82.788 cp_fm_upper_to_full 104 14.8 52.285 74.880 52.285 74.880 multiply_cannon 2485 13.5 0.702 0.739 58.277 58.857 multiply_cannon_loop 2485 14.5 0.463 0.469 54.858 56.203 ot_mini 116 10.6 0.001 0.001 43.868 43.899 dbcsr_complete_redistribute 393 12.7 3.971 4.041 29.217 41.927 copy_fm_to_dbcsr 208 11.6 0.002 0.002 25.848 38.569 rebuild_ks_matrix 127 8.3 0.001 0.001 37.814 37.842 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 37.813 37.841 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.551 36.157 qs_ks_update_qs_env 127 7.6 0.001 0.001 34.725 34.764 mp_alltoall_i22 712 14.1 21.362 33.964 21.362 33.964 cp_fm_cholesky_invert 11 10.9 33.044 33.050 33.044 33.050 mp_waitall_1 102768 16.8 27.832 31.529 27.832 31.529 qs_ot_get_p 127 10.4 0.001 0.001 24.596 24.613 qs_ot_get_derivative 116 11.6 0.002 0.002 24.017 24.048 qs_ot_p2m_diag 82 11.4 0.868 0.873 20.675 20.702 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.513 20.533 calculate_rho_elec 127 8.7 0.479 0.480 20.512 20.533 make_m2s 4970 13.5 0.075 0.077 19.465 20.242 ot_diis_step 116 11.6 0.022 0.022 19.821 19.821 make_images 4970 14.5 3.747 3.900 18.989 19.768 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 18.730 19.754 sum_up_and_integrate 127 10.3 0.318 0.319 19.272 19.351 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.081 19.272 apply_single 127 13.6 0.001 0.001 19.081 19.272 integrate_v_rspace 127 11.3 0.004 0.004 18.953 19.034 cp_dbcsr_syevd 82 12.4 0.007 0.007 18.988 18.990 multiply_cannon_multrec 9940 15.5 10.440 12.278 17.959 18.036 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.874 15.875 cp_fm_diag_elpa_base 82 14.4 11.560 13.136 15.869 15.869 multiply_cannon_sync_h2d 9940 15.5 15.540 15.551 15.540 15.551 init_scf_run 11 5.9 0.000 0.001 12.498 12.499 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.498 12.499 pw_transfer 1535 11.6 0.092 0.092 12.227 12.232 hybrid_alltoall_any 5155 16.4 1.288 3.009 10.525 12.093 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.994 11.999 make_images_data 4970 15.5 0.059 0.063 10.361 11.889 fft_wrap_pw1pw2_140 519 13.2 0.539 0.541 10.671 10.689 fft3d_ps 1281 14.7 2.719 2.732 10.047 10.055 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.252 9.282 wfi_extrapolate 11 7.9 0.001 0.001 9.160 9.160 dbcsr_mm_accdrv_process 20590 16.0 3.810 5.769 7.282 9.119 mp_alltoall_d11v 2401 14.1 7.788 8.826 7.788 8.826 cp_fm_cholesky_decompose 22 10.9 8.720 8.736 8.720 8.736 grid_integrate_task_list 127 12.3 8.510 8.685 8.510 8.685 density_rs2pw 127 9.7 0.005 0.005 8.219 8.362 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.077 8.084 calculate_dm_sparse 127 9.5 0.001 0.001 6.583 6.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.483 6.547 grid_collocate_task_list 127 9.7 6.309 6.339 6.309 6.339 rs_scatter_matrices 138 9.7 3.626 4.559 5.961 6.186 copy_dbcsr_to_fm 185 11.7 0.005 0.005 6.020 6.081 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=294.485000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2633.181818, yerr=170.025035 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 1.261609E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.022 0.071 85.122 85.123 qs_energies 1 2.0 0.000 0.001 84.531 84.537 ls_scf 1 3.0 0.000 0.000 83.607 83.613 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.650 72.858 multiply_cannon 111 7.7 0.018 0.021 56.014 57.167 multiply_cannon_loop 111 8.7 0.200 0.214 52.481 53.974 ls_scf_main 1 4.0 0.000 0.000 52.256 52.257 density_matrix_trs4 2 5.0 0.002 0.003 46.703 46.791 ls_scf_init_scf 1 4.0 0.000 0.000 28.345 28.347 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.216 27.268 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.061 25.079 mp_waitall_1 11031 10.9 22.547 24.989 22.547 24.989 multiply_cannon_multrec 2664 9.7 8.154 8.958 15.478 17.281 multiply_cannon_sync_h2d 2664 9.7 13.534 15.417 13.534 15.417 make_m2s 222 7.7 0.008 0.011 12.972 13.567 make_images 222 8.7 0.099 0.108 12.950 13.547 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.773 11.841 make_images_data 222 9.7 0.004 0.005 7.535 8.121 dbcsr_mm_accdrv_process 4760 10.4 0.505 0.624 6.942 7.895 hybrid_alltoall_any 227 10.6 0.215 1.829 6.472 7.862 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.570 7.776 dbcsr_mm_accdrv_process_sort 4760 11.4 6.238 7.115 6.238 7.115 calculate_norms 4752 9.8 5.489 6.106 5.489 6.106 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.073 5.190 mp_sum_l 807 5.4 3.140 4.701 3.140 4.701 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.346 3.619 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.077 3.412 make_images_sizes 222 9.7 0.000 0.000 0.674 3.396 mp_alltoall_i44 222 10.7 0.673 3.395 0.673 3.395 mp_irecv_dv 6231 10.9 2.059 3.380 2.059 3.380 arnoldi_extremal 4 6.8 0.000 0.000 3.176 3.205 arnoldi_normal_ev 4 7.8 0.001 0.003 3.176 3.205 build_subspace 16 8.4 0.009 0.012 3.084 3.086 ls_scf_post 1 4.0 0.000 0.000 3.005 3.011 dbcsr_special_finalize 555 9.7 0.005 0.006 2.412 2.888 dbcsr_merge_single_wm 555 10.7 0.484 0.634 2.404 2.880 ls_scf_store_result 1 5.0 0.000 0.000 2.825 2.868 make_images_pack 222 9.7 2.206 2.625 2.208 2.626 dbcsr_sort_data 658 11.4 2.177 2.586 2.177 2.586 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.306 2.548 dbcsr_matrix_vector_mult_local 304 10.0 2.066 2.455 2.068 2.457 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.289 2.346 buffer_matrices_ensure_size 222 8.7 1.813 2.193 1.813 2.193 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.808 1.809 rebuild_ks_matrix 3 7.3 0.000 0.000 1.799 1.800 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.799 1.800 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.123000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1111.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 2.104832E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.053 91.270 91.273 qs_energies 1 2.0 0.000 0.001 90.687 90.691 ls_scf 1 3.0 0.003 0.018 89.365 89.371 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.257 75.704 multiply_cannon 111 7.7 0.030 0.041 53.423 56.976 ls_scf_main 1 4.0 0.000 0.008 55.039 55.050 multiply_cannon_loop 111 8.7 0.115 0.121 50.133 53.432 density_matrix_trs4 2 5.0 0.002 0.004 49.300 49.460 ls_scf_init_scf 1 4.0 0.000 0.002 30.799 30.804 mp_waitall_1 9105 10.9 21.311 29.975 21.311 29.975 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.506 29.625 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.106 27.124 multiply_cannon_multrec 1332 9.7 13.092 16.603 22.337 27.022 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.975 20.686 make_m2s 222 7.7 0.006 0.008 15.226 15.896 make_images 222 8.7 1.569 1.963 15.195 15.866 dbcsr_mm_accdrv_process 4041 10.4 0.296 0.478 8.838 10.396 dbcsr_mm_accdrv_process_sort 4041 11.4 8.420 9.933 8.420 9.933 make_images_data 222 9.7 0.004 0.004 8.733 9.685 hybrid_alltoall_any 227 10.6 0.519 2.460 8.153 9.172 mp_sum_l 807 5.4 5.301 8.170 5.301 8.170 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.227 7.872 mp_irecv_dv 3311 11.0 3.207 7.819 3.207 7.819 calculate_norms 2376 9.8 5.979 6.820 5.979 6.820 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.084 6.411 multiply_cannon_sync_h2d 1332 9.7 4.816 5.935 4.816 5.935 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.048 5.223 arnoldi_extremal 4 6.8 0.000 0.000 4.648 4.675 arnoldi_normal_ev 4 7.8 0.001 0.004 4.648 4.675 build_subspace 16 8.4 0.014 0.021 4.386 4.389 ls_scf_post 1 4.0 0.000 0.001 3.525 3.529 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.163 3.404 ls_scf_store_result 1 5.0 0.000 0.000 3.216 3.349 dbcsr_matrix_vector_mult_local 304 10.0 2.774 3.275 2.776 3.277 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.610 2.702 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.233 2.573 make_images_pack 222 9.7 2.016 2.378 2.018 2.380 mp_allgather_i34 111 8.7 1.001 2.352 1.001 2.352 dbcsr_sort_data 436 11.2 1.809 2.026 1.809 2.026 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.941 1.947 rebuild_ks_matrix 3 7.3 0.000 0.000 1.928 1.934 qs_ks_build_kohn_sham_matrix 3 8.3 0.010 0.035 1.928 1.934 dbcsr_data_new 4174 10.1 1.600 1.843 1.600 1.843 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.273000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1750.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 2.698318E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.064 94.991 94.998 qs_energies 1 2.0 0.000 0.000 94.373 94.379 ls_scf 1 3.0 0.000 0.000 92.974 92.980 dbcsr_multiply_generic 111 6.7 0.015 0.017 77.799 78.108 ls_scf_main 1 4.0 0.000 0.000 58.458 58.465 multiply_cannon 111 7.7 0.039 0.067 53.613 57.567 multiply_cannon_loop 111 8.7 0.099 0.106 50.078 54.433 density_matrix_trs4 2 5.0 0.002 0.003 52.434 52.660 mp_waitall_1 7281 11.0 24.835 35.233 24.835 35.233 ls_scf_init_scf 1 4.0 0.000 0.000 30.915 30.918 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.723 29.792 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.251 27.259 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.611 24.510 multiply_cannon_multrec 888 9.7 12.496 15.268 21.048 24.394 make_m2s 222 7.7 0.006 0.007 17.272 18.523 make_images 222 8.7 1.962 2.294 17.234 18.481 hybrid_alltoall_any 227 10.6 0.617 2.862 9.583 11.028 make_images_data 222 9.7 0.003 0.004 9.902 10.951 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.429 8.062 9.285 mp_sum_l 807 5.4 5.364 9.253 5.364 9.253 dbcsr_mm_accdrv_process_sort 3754 11.4 7.689 8.857 7.689 8.857 multiply_cannon_sync_h2d 888 9.7 6.065 8.098 6.065 8.098 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.101 7.266 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.941 7.234 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.477 6.996 mp_irecv_dv 2335 11.1 2.462 6.940 2.462 6.940 apply_matrix_preconditioner 6 5.3 0.000 0.001 5.017 5.207 arnoldi_extremal 4 6.8 0.000 0.000 5.016 5.026 arnoldi_normal_ev 4 7.8 0.001 0.007 5.016 5.026 calculate_norms 1584 9.8 4.327 4.759 4.327 4.759 build_subspace 16 8.4 0.014 0.020 4.718 4.724 mp_allgather_i34 111 8.7 1.380 3.889 1.380 3.889 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.424 3.751 ls_scf_post 1 4.0 0.000 0.000 3.601 3.608 dbcsr_matrix_vector_mult_local 304 10.0 3.027 3.597 3.029 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.345 3.429 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.861 2.962 make_images_sizes 222 9.7 0.000 0.000 1.073 2.344 mp_alltoall_i44 222 10.7 1.073 2.344 1.073 2.344 dbcsr_sort_data 325 11.1 1.858 2.163 1.858 2.163 make_images_pack 222 9.7 1.830 2.118 1.832 2.121 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.919 1.920 dbcsr_data_release 9322 10.9 1.322 1.903 1.322 1.903 rebuild_ks_matrix 3 7.3 0.000 0.000 1.900 1.902 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.900 1.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.998000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2156.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 3.332456E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.118 0.318 97.138 97.189 qs_energies 1 2.0 0.000 0.000 96.211 96.276 ls_scf 1 3.0 0.003 0.023 94.425 94.480 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.106 78.383 ls_scf_main 1 4.0 0.000 0.003 58.690 58.691 multiply_cannon 111 7.7 0.060 0.152 51.421 55.687 density_matrix_trs4 2 5.0 0.002 0.003 52.569 52.690 multiply_cannon_loop 111 8.7 0.114 0.124 46.350 49.451 ls_scf_init_scf 1 4.0 0.000 0.004 32.442 32.445 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.191 31.267 mp_waitall_1 6369 11.0 22.483 29.763 22.483 29.763 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.725 28.742 multiply_cannon_multrec 1332 9.7 14.171 17.529 22.033 24.741 make_m2s 222 7.7 0.007 0.008 21.089 22.500 make_images 222 8.7 3.123 3.588 21.038 22.451 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.076 16.545 make_images_data 222 9.7 0.004 0.005 11.747 13.387 hybrid_alltoall_any 227 10.6 0.797 3.799 11.002 12.810 dbcsr_mm_accdrv_process 3641 10.4 0.218 0.404 7.504 9.058 dbcsr_mm_accdrv_process_sort 3641 11.4 7.132 8.637 7.132 8.637 mp_sum_l 807 5.4 4.092 7.081 4.092 7.081 multiply_cannon_sync_h2d 1332 9.7 5.535 6.149 5.535 6.149 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.090 5.926 mp_irecv_dv 3229 10.9 2.067 5.851 2.067 5.851 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.029 5.466 arnoldi_extremal 4 6.8 0.000 0.000 5.201 5.218 arnoldi_normal_ev 4 7.8 0.002 0.005 5.201 5.218 build_subspace 16 8.4 0.014 0.021 4.858 4.866 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.536 4.738 mp_allgather_i34 111 8.7 2.209 4.601 2.209 4.601 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.488 4.600 calculate_norms 2376 9.8 4.196 4.568 4.196 4.568 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.565 3.865 dbcsr_matrix_vector_mult_local 304 10.0 3.182 3.689 3.184 3.691 dbcsr_sort_data 658 11.4 3.062 3.516 3.062 3.516 ls_scf_post 1 4.0 0.005 0.019 3.290 3.346 dbcsr_special_finalize 555 9.7 0.006 0.007 2.809 3.207 dbcsr_merge_single_wm 555 10.7 0.533 0.663 2.801 3.199 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.037 3.087 ls_scf_store_result 1 5.0 0.000 0.000 2.999 3.051 dbcsr_data_release 10477 10.7 1.570 2.446 1.570 2.446 dbcsr_finalize 304 7.8 0.050 0.061 1.794 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.982 1.984 rebuild_ks_matrix 3 7.3 0.000 0.000 1.960 1.961 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.005 1.960 1.961 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.189000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2747.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 4.630016E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.061 99.453 99.454 qs_energies 1 2.0 0.000 0.000 98.609 98.612 ls_scf 1 3.0 0.000 0.000 96.657 96.660 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.006 78.252 ls_scf_main 1 4.0 0.000 0.000 62.136 62.144 multiply_cannon 111 7.7 0.104 0.175 55.389 60.786 density_matrix_trs4 2 5.0 0.002 0.003 54.871 54.953 multiply_cannon_loop 111 8.7 0.069 0.077 50.959 52.625 mp_waitall_1 5436 11.0 26.676 32.254 26.676 32.254 ls_scf_init_scf 1 4.0 0.000 0.000 30.924 30.926 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.525 29.558 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.362 27.373 multiply_cannon_multrec 444 9.7 14.041 16.527 21.090 23.524 make_m2s 222 7.7 0.004 0.005 17.760 20.189 make_images 222 8.7 3.709 4.410 17.699 20.129 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.250 16.084 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.236 14.541 make_images_data 222 9.7 0.003 0.004 9.993 12.391 hybrid_alltoall_any 227 10.6 0.789 3.767 9.805 12.177 multiply_cannon_sync_h2d 444 9.7 6.552 8.265 6.552 8.265 dbcsr_mm_accdrv_process 3003 10.4 0.178 0.351 6.750 7.854 dbcsr_mm_accdrv_process_sort 3003 11.4 6.434 7.513 6.434 7.513 mp_allgather_i34 111 8.7 2.673 6.989 2.673 6.989 arnoldi_extremal 4 6.8 0.000 0.000 5.724 5.733 arnoldi_normal_ev 4 7.8 0.002 0.005 5.724 5.732 mp_sum_l 807 5.4 2.903 5.539 2.903 5.539 build_subspace 16 8.4 0.015 0.020 5.339 5.352 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.531 4.683 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.658 4.385 mp_irecv_dv 1241 11.2 1.638 4.357 1.638 4.357 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.066 4.345 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.107 4.310 dbcsr_matrix_vector_mult_local 304 10.0 3.656 4.119 3.658 4.121 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.709 3.798 calculate_norms 792 9.8 3.540 3.656 3.540 3.656 ls_scf_post 1 4.0 0.000 0.000 3.597 3.601 make_images_sizes 222 9.7 0.000 0.000 1.144 3.465 mp_alltoall_i44 222 10.7 1.143 3.465 1.143 3.465 ls_scf_store_result 1 5.0 0.000 0.000 3.381 3.424 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.298 2.300 rebuild_ks_matrix 3 7.3 0.000 0.000 2.266 2.268 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.009 2.266 2.268 dbcsr_finalize 304 7.8 0.062 0.078 2.194 2.260 dbcsr_merge_all 275 8.9 0.473 0.527 2.045 2.099 dbcsr_data_release 10123 10.8 1.329 1.997 1.329 1.997 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.454000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3600.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/52c01cbe9bfe1b67bcd624e507926c0777952c09_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 8.769085E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.077 0.098 108.081 108.083 qs_energies 1 2.0 0.000 0.000 106.674 106.688 ls_scf 1 3.0 0.000 0.000 103.700 103.714 dbcsr_multiply_generic 111 6.7 0.023 0.025 77.181 77.298 ls_scf_main 1 4.0 0.000 0.000 65.793 65.808 density_matrix_trs4 2 5.0 0.002 0.003 56.488 56.533 multiply_cannon 111 7.7 0.176 0.275 49.880 51.993 multiply_cannon_loop 111 8.7 0.068 0.076 46.436 47.172 ls_scf_init_scf 1 4.0 0.000 0.000 34.252 34.255 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.488 32.496 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.696 29.711 mp_waitall_1 4527 11.1 22.088 25.540 22.088 25.540 make_m2s 222 7.7 0.005 0.005 23.762 24.736 make_images 222 8.7 4.570 4.954 23.655 24.626 multiply_cannon_multrec 444 9.7 17.900 18.612 22.500 23.097 hybrid_alltoall_any 227 10.6 1.661 3.640 12.845 15.642 make_images_data 222 9.7 0.003 0.003 13.020 15.405 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.460 11.040 multiply_cannon_sync_h2d 444 9.7 8.846 8.900 8.846 8.900 arnoldi_extremal 4 6.8 0.000 0.000 7.244 7.262 arnoldi_normal_ev 4 7.8 0.003 0.009 7.243 7.262 build_subspace 16 8.4 0.026 0.037 6.687 6.695 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.356 5.507 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.259 5.346 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.008 5.253 dbcsr_matrix_vector_mult_local 304 10.0 4.959 5.250 4.962 5.253 dbcsr_mm_accdrv_process 1814 10.4 0.239 0.338 4.441 4.559 dbcsr_mm_accdrv_process_sort 1814 11.4 4.133 4.262 4.133 4.262 ls_scf_post 1 4.0 0.000 0.000 3.655 3.669 make_images_sizes 222 9.7 0.000 0.000 1.438 3.542 mp_alltoall_i44 222 10.7 1.437 3.542 1.437 3.542 ls_scf_store_result 1 5.0 0.000 0.000 3.408 3.415 mp_allgather_i34 111 8.7 1.099 3.399 1.099 3.399 calculate_norms 792 9.8 3.243 3.280 3.243 3.280 dbcsr_finalize 304 7.8 0.082 0.089 3.087 3.175 dbcsr_merge_all 275 8.9 0.893 0.923 2.873 2.954 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.943 2.943 dbcsr_complete_redistribute 5 7.6 1.435 1.467 2.772 2.879 dbcsr_data_release 12724 10.6 2.329 2.869 2.329 2.869 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.693 2.697 rebuild_ks_matrix 3 7.3 0.000 0.000 2.628 2.631 qs_ks_build_kohn_sham_matrix 3 8.3 0.016 0.018 2.627 2.631 matrix_ls_to_qs 2 6.0 0.000 0.000 2.430 2.542 dbcsr_sort_data 325 11.1 2.440 2.500 2.440 2.500 dbcsr_new_transposed 4 7.5 0.243 0.253 2.289 2.305 dbcsr_frobenius_norm 74 6.6 2.056 2.130 2.187 2.229 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.207 dbcsr_add_anytype 103 7.2 0.860 0.891 2.132 2.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.083000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6833.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 52c01cbe9bfe1b67bcd624e507926c0777952c09 Summary: empty Status: OK