=== This is the CP2K Performance-Test === Already up to date. Current branch master is up to date. Already up to date. Current branch master is up to date. GIT Revision: fe1068d0649e3e881f064d86a0513eff1752c2c8 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (31.05.2023) # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/01 job id: 47059264 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/02 job id: 47059267 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/03 job id: 47059272 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/04 job id: 47059274 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/05 job id: 47059275 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/06 job id: 47059276 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/07 job id: 47059277 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/08 job id: 47059278 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/09 job id: 47059279 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/10 job id: 47059280 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/11 job id: 47059281 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/12 job id: 47059282 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/13 job id: 47059283 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/14 job id: 47059285 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/15 job id: 47059286 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/16 job id: 47059288 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/17 job id: 47059289 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/18 job id: 47059290 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/19 job id: 47059291 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/20 job id: 47059292 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/21 job id: 47059294 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/22 job id: 47059295 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/23 job id: 47059296 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/24 job id: 47059297 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/25 job id: 47059298 --- 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/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/26 job id: 47059299 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/27 job id: 47059300 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 137.015 137.016 farming_run 1 2.0 135.657 135.659 136.989 136.991 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.465471E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1103589. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.025 115.193 115.194 qs_energies 1 2.0 0.000 0.000 114.927 114.928 mp2_main 1 3.0 0.000 0.000 112.823 112.824 mp2_gpw_main 1 4.0 0.019 0.025 111.748 111.750 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.739 93.153 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.915 55.330 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.166 41.356 46.567 get_2c_integrals 1 6.0 0.008 0.009 37.125 37.651 integrate_v_rspace 273 8.0 0.435 0.450 24.764 29.792 pw_transfer 6555 10.6 0.373 0.403 27.499 28.212 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 26.110 26.777 grid_integrate_task_list 273 9.0 20.581 26.090 20.581 26.090 fft_wrap_pw1pw2_100 2178 12.4 1.237 1.403 23.633 24.296 compute_2c_integrals 1 7.0 0.002 0.003 19.728 19.730 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.027 19.393 mp2_eri_2c_integrate_gpw 1 9.0 2.355 2.430 19.024 19.391 rpa_ri_compute_en 1 5.0 0.019 0.020 18.899 19.142 cp_fm_cholesky_decompose 12 8.2 17.352 17.867 17.352 17.867 cholesky_decomp 1 7.0 0.000 0.000 16.238 16.752 fft3d_s 5443 13.4 16.232 16.643 16.254 16.665 ao_to_mo_and_store_B_mult_1 272 7.0 10.728 15.305 10.728 15.305 calculate_wavefunction 272 8.0 5.405 5.559 12.512 13.122 rpa_num_int 1 6.0 0.000 0.000 10.667 10.667 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.648 10.661 calc_mat_Q 8 8.0 0.000 0.000 9.494 9.579 contract_S_to_Q 8 9.0 0.000 0.000 8.914 8.999 calc_potential_gpw 544 9.5 0.005 0.006 8.250 8.673 parallel_gemm_fm 14 9.1 0.000 0.000 8.502 8.607 parallel_gemm_fm_cosma 14 10.1 8.502 8.607 8.502 8.607 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.291 8.542 potential_pw2rs 545 10.0 0.108 0.111 7.767 8.434 collocate_single_gaussian 272 10.0 0.039 0.042 7.569 7.781 create_integ_mat 1 6.0 0.022 0.027 7.725 7.725 array2fm 1 7.0 0.000 0.000 6.739 7.166 pw_scatter_s 2720 13.7 4.382 4.586 4.382 4.586 pw_gather_s 2722 13.2 3.852 4.104 3.852 4.104 array2fm_buffer_send 1 8.0 2.956 3.166 2.956 3.166 pw_poisson_solve 545 10.5 1.108 1.146 2.207 2.421 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.747189, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.029 0.038 401.313 401.314 farming_run 1 2.0 399.524 399.530 401.273 401.276 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 897827141120 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 249788822 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 1.227293E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 743 386399. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.032 210.278 210.278 qs_energies 1 2.0 0.004 0.015 209.974 209.994 scf_env_do_scf 1 3.0 0.000 0.000 106.966 106.966 qs_ks_update_qs_env 5 5.0 0.000 0.000 106.046 106.054 rebuild_ks_matrix 4 6.0 0.000 0.000 106.045 106.053 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 106.045 106.053 hfx_ks_matrix 4 8.0 0.001 0.001 105.645 105.650 integrate_four_center 4 9.0 0.144 0.452 105.644 105.649 mp2_main 1 3.0 0.005 0.026 102.687 102.700 mp2_gpw_main 1 4.0 0.079 0.269 101.784 101.801 integrate_four_center_main 4 10.0 0.124 0.505 96.719 100.267 integrate_four_center_bin 263 11.0 96.594 99.827 96.594 99.827 init_scf_loop 1 4.0 0.000 0.000 92.734 92.734 mp2_ri_gpw_compute_in 1 5.0 0.083 0.143 74.616 75.726 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.057 55.158 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.165 41.759 46.752 integrate_v_rspace 95 8.0 0.397 0.568 28.158 32.976 pw_transfer 2240 10.6 0.142 0.175 29.967 30.413 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.967 29.446 ao_to_mo_and_store_B_mult_1 91 7.0 10.614 29.367 10.614 29.367 mp2_ri_gpw_compute_en 1 5.0 0.075 0.130 26.806 28.517 grid_integrate_task_list 95 9.0 23.433 28.480 23.433 28.480 fft_wrap_pw1pw2_100 730 12.4 1.301 1.454 26.650 27.131 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.855 1.918 24.948 24.960 get_2c_integrals 1 6.0 0.010 0.054 20.435 20.502 compute_2c_integrals 1 7.0 0.006 0.017 19.394 19.428 compute_2c_integrals_loop_lm 1 8.0 0.004 0.021 18.951 19.224 mp2_eri_2c_integrate_gpw 1 9.0 1.732 1.853 18.948 19.223 fft3d_s 1823 13.4 18.436 18.773 18.449 18.787 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.230 14.230 calculate_wavefunction 91 8.0 2.012 2.045 9.736 9.925 potential_pw2rs 186 10.0 0.034 0.035 8.673 9.257 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.580 8.723 9.153 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.303 8.648 local_gemm 172 8.0 8.167 8.589 8.167 8.589 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.523 7.970 8.393 calc_potential_gpw 182 9.5 0.002 0.002 7.949 8.227 collocate_single_gaussian 91 10.0 0.023 0.078 7.918 8.161 mp_sendrecv_dm3 2068 8.0 6.004 6.440 6.004 6.440 mp2_ri_gpw_compute_en_ener 172 7.0 6.355 6.431 6.355 6.431 mp_sync 37 10.5 3.596 6.381 3.596 6.381 pw_gather_s 912 13.2 4.878 5.467 4.878 5.467 pw_scatter_s 910 13.7 3.948 4.466 3.948 4.466 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.790959, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1515.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 452.153344E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 588782. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.030 54.452 54.453 qs_mol_dyn_low 1 2.0 0.003 0.003 54.241 54.249 qs_forces 11 3.9 0.002 0.002 54.176 54.177 qs_energies 11 4.9 0.002 0.002 52.659 52.683 scf_env_do_scf 11 5.9 0.000 0.001 45.429 45.430 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 43.317 43.317 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.484 33.970 qs_scf_new_mos 108 7.5 0.000 0.001 33.009 33.316 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.008 33.316 ot_scf_mini 108 9.5 0.002 0.002 31.374 31.574 multiply_cannon 2286 13.5 0.181 0.188 26.406 27.995 multiply_cannon_loop 2286 14.5 1.470 1.572 25.742 27.350 velocity_verlet 10 3.0 0.001 0.001 26.313 26.313 ot_mini 108 10.5 0.001 0.001 18.757 19.028 qs_ot_get_derivative 108 11.5 0.001 0.001 15.852 16.041 mp_waitall_1 245248 16.5 8.537 14.807 8.537 14.807 multiply_cannon_metrocomm3 54864 15.5 0.068 0.074 6.035 13.296 multiply_cannon_multrec 54864 15.5 4.279 6.600 8.008 11.642 qs_ot_get_p 119 10.4 0.001 0.001 8.053 8.398 rebuild_ks_matrix 119 8.3 0.000 0.000 8.018 8.172 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.017 8.171 multiply_cannon_sync_h2d 54864 15.5 5.930 7.341 5.930 7.341 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.062 7.200 mp_sum_l 7287 12.8 5.101 6.795 5.101 6.795 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.658 6.108 init_scf_run 11 5.9 0.000 0.001 5.959 5.959 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.959 5.959 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.324 5.380 dbcsr_mm_accdrv_process 76910 16.1 1.126 1.815 3.651 5.157 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.011 5.116 sum_up_and_integrate 119 10.3 0.012 0.014 4.604 4.613 integrate_v_rspace 119 11.3 0.002 0.002 4.593 4.602 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.551 4.552 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.318 4.420 calculate_rho_elec 119 8.7 0.012 0.017 4.318 4.420 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.342 4.342 cp_fm_redistribute_end 50 14.0 2.215 4.315 2.221 4.318 cp_fm_diag_elpa_base 50 14.0 2.091 4.201 2.095 4.209 calculate_first_density_matrix 1 7.0 0.000 0.000 3.735 3.738 jit_kernel_multiply 13 15.8 2.466 3.281 2.466 3.281 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.266 3.268 calculate_dm_sparse 119 9.5 0.000 0.001 3.081 3.208 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.904 3.112 apply_single 119 13.6 0.000 0.000 2.903 3.112 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.779 3.090 rs_pw_transfer 974 11.9 0.011 0.012 2.833 2.960 ot_diis_step 108 11.5 0.006 0.006 2.692 2.693 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.525 2.575 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.411 2.486 density_rs2pw 119 9.7 0.004 0.005 2.269 2.392 acc_transpose_blocks 54864 15.5 0.223 0.248 1.774 2.289 wfi_extrapolate 11 7.9 0.001 0.001 2.154 2.154 grid_integrate_task_list 119 12.3 2.018 2.106 2.018 2.106 init_scf_loop 11 6.9 0.000 0.000 2.096 2.097 potential_pw2rs 119 12.3 0.004 0.004 1.933 1.945 mp_sum_d 4135 12.0 1.257 1.889 1.257 1.889 pw_transfer 1439 11.6 0.052 0.058 1.787 1.857 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 1.711 1.784 make_m2s 4572 13.5 0.053 0.055 1.659 1.710 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.612 1.652 make_images 4572 14.5 0.133 0.138 1.578 1.628 mp_alltoall_d11v 2130 13.8 1.323 1.507 1.323 1.507 mp_waitany 12084 13.8 1.273 1.482 1.273 1.482 grid_collocate_task_list 119 9.7 1.348 1.432 1.348 1.432 fft3d_ps 1201 14.6 0.375 0.481 1.353 1.422 fft_wrap_pw1pw2_140 487 13.2 0.185 0.203 1.327 1.403 acc_transpose_blocks_kernels 54864 16.5 0.241 0.377 0.799 1.118 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=54.453000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 486.825984E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 998897. 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 0.035 41.775 41.777 qs_mol_dyn_low 1 2.0 0.003 0.003 41.416 41.483 qs_forces 11 3.9 0.003 0.003 40.892 40.893 qs_energies 11 4.9 0.002 0.003 39.178 39.182 scf_env_do_scf 11 5.9 0.001 0.002 33.154 33.154 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 30.604 30.604 dbcsr_multiply_generic 2286 12.5 0.101 0.104 23.068 23.487 qs_scf_new_mos 108 7.5 0.001 0.001 21.232 21.480 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.232 21.480 ot_scf_mini 108 9.5 0.002 0.003 20.332 20.506 multiply_cannon 2286 13.5 0.207 0.219 17.943 19.442 velocity_verlet 10 3.0 0.001 0.002 19.215 19.226 multiply_cannon_loop 2286 14.5 0.904 0.977 16.783 18.389 ot_mini 108 10.5 0.001 0.001 12.730 12.968 mp_waitall_1 200699 16.5 6.334 12.399 6.334 12.399 multiply_cannon_metrocomm3 27432 15.5 0.067 0.070 4.831 10.894 qs_ot_get_derivative 108 11.5 0.001 0.001 10.241 10.418 multiply_cannon_multrec 27432 15.5 2.066 5.028 6.500 9.640 rebuild_ks_matrix 119 8.3 0.000 0.000 7.283 7.421 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.282 7.420 dbcsr_mm_accdrv_process 47894 16.0 3.084 5.849 4.364 6.750 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.421 6.546 qs_ot_get_p 119 10.4 0.001 0.001 4.669 4.898 init_scf_run 11 5.9 0.000 0.001 4.781 4.782 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.781 4.781 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.909 4.745 sum_up_and_integrate 119 10.3 0.025 0.028 4.271 4.276 integrate_v_rspace 119 11.3 0.002 0.003 4.246 4.252 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.086 4.147 apply_single 119 13.6 0.000 0.000 3.086 4.147 mp_sum_l 7287 12.8 2.117 4.131 2.117 4.131 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.947 4.023 calculate_rho_elec 119 8.7 0.021 0.024 3.947 4.023 jit_kernel_multiply 12 16.1 1.228 3.431 1.228 3.431 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.173 3.194 calculate_first_density_matrix 1 7.0 0.000 0.000 3.178 3.183 rs_pw_transfer 974 11.9 0.010 0.011 2.673 3.113 multiply_cannon_sync_h2d 27432 15.5 2.167 2.814 2.167 2.814 make_m2s 4572 13.5 0.052 0.054 2.556 2.786 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.729 2.729 make_images 4572 14.5 0.200 0.236 2.466 2.695 density_rs2pw 119 9.7 0.004 0.004 2.224 2.688 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.577 2.580 init_scf_loop 11 6.9 0.000 0.000 2.527 2.528 ot_diis_step 108 11.5 0.011 0.011 2.433 2.434 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.362 2.363 calculate_dm_sparse 119 9.5 0.000 0.001 2.265 2.343 cp_fm_redistribute_end 50 14.0 1.193 2.324 1.196 2.327 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.214 2.308 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.250 2.293 cp_fm_diag_elpa_base 50 14.0 1.098 2.227 1.126 2.271 pw_transfer 1439 11.6 0.065 0.070 2.017 2.057 acc_transpose_blocks 27432 15.5 0.110 0.115 1.537 2.025 potential_pw2rs 119 12.3 0.006 0.006 1.980 1.991 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.925 1.967 grid_integrate_task_list 119 12.3 1.832 1.920 1.832 1.920 prepare_preconditioner 11 7.9 0.000 0.000 1.621 1.650 make_preconditioner 11 8.9 0.000 0.000 1.621 1.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.517 1.576 make_images_data 4572 15.5 0.044 0.051 1.161 1.572 wfi_extrapolate 11 7.9 0.001 0.001 1.532 1.532 mp_alltoall_d11v 2130 13.8 1.272 1.529 1.272 1.529 fft_wrap_pw1pw2_140 487 13.2 0.203 0.213 1.474 1.514 fft3d_ps 1201 14.6 0.524 0.582 1.474 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.478 1.490 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.013 1.489 grid_collocate_task_list 119 9.7 1.267 1.417 1.267 1.417 acc_transpose_blocks_kernels 27432 16.5 0.183 0.271 0.985 1.381 mp_allgather_i34 2286 14.5 0.597 1.374 0.597 1.374 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.202 1.247 jit_kernel_transpose 5 15.5 0.803 1.111 0.803 1.111 mp_sum_d 4135 12.0 0.587 1.023 0.587 1.023 rs_pw_transfer_RS2PW_140 130 11.5 0.137 0.145 0.544 0.993 mp_waitany 5720 13.7 0.522 0.991 0.522 0.991 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.936 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.921 0.935 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.777000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.545455, yerr=2.016461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 522.137600E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607820. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.045 34.962 34.964 qs_mol_dyn_low 1 2.0 0.003 0.007 34.232 34.240 qs_forces 11 3.9 0.002 0.003 34.061 34.063 qs_energies 11 4.9 0.002 0.005 32.449 32.454 scf_env_do_scf 11 5.9 0.047 0.374 26.885 26.885 scf_env_do_scf_inner_loop 108 6.5 0.100 0.390 24.188 24.236 dbcsr_multiply_generic 2286 12.5 0.096 0.098 17.638 17.915 velocity_verlet 10 3.0 0.001 0.002 15.928 15.932 qs_scf_new_mos 108 7.5 0.001 0.001 15.706 15.869 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.705 15.868 ot_scf_mini 108 9.5 0.004 0.014 14.937 15.095 multiply_cannon 2286 13.5 0.194 0.200 13.875 14.828 multiply_cannon_loop 2286 14.5 0.637 0.668 13.050 14.047 ot_mini 108 10.5 0.001 0.001 9.241 9.395 qs_ot_get_derivative 108 11.5 0.001 0.001 7.688 7.846 multiply_cannon_multrec 18288 15.5 1.943 2.863 7.306 7.701 rebuild_ks_matrix 119 8.3 0.000 0.000 6.501 6.530 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.500 6.530 dbcsr_mm_accdrv_process 38222 16.0 4.587 5.951 5.278 6.098 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.730 5.758 init_scf_run 11 5.9 0.000 0.001 4.290 4.290 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.289 4.289 mp_waitall_1 158411 16.6 2.921 4.111 2.921 4.111 sum_up_and_integrate 119 10.3 0.030 0.031 4.060 4.064 integrate_v_rspace 119 11.3 0.002 0.003 4.030 4.036 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.072 3.844 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.589 3.598 calculate_rho_elec 119 8.7 0.031 0.031 3.589 3.597 qs_ot_get_p 119 10.4 0.001 0.001 3.481 3.503 calculate_first_density_matrix 1 7.0 0.000 0.002 3.012 3.013 rs_pw_transfer 974 11.9 0.009 0.010 2.455 2.725 init_scf_loop 11 6.9 0.001 0.004 2.630 2.632 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.513 2.484 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.060 2.373 apply_single 119 13.6 0.000 0.000 2.060 2.373 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.339 2.346 density_rs2pw 119 9.7 0.004 0.004 2.069 2.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.238 2.240 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.074 2.083 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.065 2.065 calculate_dm_sparse 119 9.5 0.000 0.000 2.019 2.030 make_m2s 4572 13.5 0.044 0.046 1.901 2.028 pw_transfer 1439 11.6 0.065 0.069 2.002 2.012 mp_sum_l 7287 12.8 1.484 1.950 1.484 1.950 make_images 4572 14.5 0.191 0.203 1.814 1.939 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.909 1.923 grid_integrate_task_list 119 12.3 1.798 1.874 1.798 1.874 prepare_preconditioner 11 7.9 0.000 0.000 1.819 1.821 make_preconditioner 11 8.9 0.000 0.003 1.819 1.821 potential_pw2rs 119 12.3 0.007 0.008 1.807 1.813 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.799 1.810 cp_fm_diag_elpa_base 50 14.0 1.774 1.790 1.797 1.808 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.663 1.749 jit_kernel_multiply 8 16.5 0.640 1.650 0.640 1.650 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.625 1.635 multiply_cannon_sync_h2d 18288 15.5 1.380 1.585 1.380 1.585 ot_diis_step 108 11.5 0.011 0.016 1.528 1.528 fft_wrap_pw1pw2_140 487 13.2 0.258 0.287 1.513 1.525 fft3d_ps 1201 14.6 0.532 0.547 1.391 1.410 grid_collocate_task_list 119 9.7 1.243 1.393 1.243 1.393 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.348 1.354 acc_transpose_blocks 18288 15.5 0.076 0.078 1.293 1.319 wfi_extrapolate 11 7.9 0.001 0.001 1.213 1.213 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.043 1.044 make_images_data 4572 15.5 0.044 0.048 0.855 1.005 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.913 0.940 hybrid_alltoall_any 4725 16.4 0.055 0.112 0.736 0.905 mp_waitany 9880 13.7 0.591 0.866 0.591 0.866 acc_transpose_blocks_kernels 18288 16.5 0.211 0.220 0.838 0.860 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.124 0.582 0.851 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.843 0.845 mp_alltoall_d11v 2130 13.8 0.730 0.821 0.730 0.821 mp_alltoall_z22v 1201 16.6 0.704 0.786 0.704 0.786 multiply_cannon_metrocomm1 18288 15.5 0.028 0.030 0.508 0.771 cp_fm_cholesky_invert 11 10.9 0.753 0.757 0.753 0.757 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.650 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.964000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.363636, yerr=2.143605 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 548.950016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.091 39.194 39.196 qs_mol_dyn_low 1 2.0 0.003 0.003 38.086 38.093 qs_forces 11 3.9 0.002 0.004 37.830 37.832 qs_energies 11 4.9 0.032 0.066 36.090 36.094 scf_env_do_scf 11 5.9 0.001 0.002 29.586 29.588 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 25.779 25.779 dbcsr_multiply_generic 2286 12.5 0.110 0.117 19.689 19.911 velocity_verlet 10 3.0 0.001 0.002 18.318 18.328 qs_scf_new_mos 108 7.5 0.001 0.001 17.182 17.242 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.181 17.242 multiply_cannon 2286 13.5 0.236 0.303 15.750 16.730 ot_scf_mini 108 9.5 0.005 0.023 16.229 16.283 multiply_cannon_loop 2286 14.5 0.946 0.978 14.777 15.568 ot_mini 108 10.5 0.001 0.001 10.097 10.166 multiply_cannon_multrec 27432 15.5 2.370 3.266 8.901 9.332 qs_ot_get_derivative 108 11.5 0.001 0.001 8.265 8.321 dbcsr_mm_accdrv_process 47916 15.9 5.591 7.458 6.441 7.628 rebuild_ks_matrix 119 8.3 0.000 0.000 6.667 6.721 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.667 6.720 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.912 5.959 init_scf_run 11 5.9 0.000 0.001 4.589 4.590 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.589 4.590 sum_up_and_integrate 119 10.3 0.035 0.039 3.863 3.872 integrate_v_rspace 119 11.3 0.003 0.003 3.828 3.838 init_scf_loop 11 6.9 0.042 0.328 3.782 3.782 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.115 3.715 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.631 3.665 calculate_rho_elec 119 8.7 0.040 0.046 3.630 3.665 qs_ot_get_p 119 10.4 0.001 0.001 3.485 3.563 mp_waitall_1 137007 16.6 2.457 3.360 2.457 3.360 calculate_first_density_matrix 1 7.0 0.002 0.011 3.158 3.160 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.270 2.707 apply_single 119 13.6 0.000 0.000 2.269 2.707 acc_transpose_blocks 27432 15.5 0.111 0.113 1.836 2.672 prepare_preconditioner 11 7.9 0.000 0.000 2.593 2.601 make_preconditioner 11 8.9 0.000 0.001 2.593 2.601 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.205 2.526 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.384 2.386 make_m2s 4572 13.5 0.054 0.056 2.261 2.360 calculate_dm_sparse 119 9.5 0.000 0.001 2.267 2.326 rs_pw_transfer 974 11.9 0.009 0.010 2.080 2.277 make_images 4572 14.5 0.270 0.331 2.154 2.253 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.189 2.202 density_rs2pw 119 9.7 0.004 0.004 2.013 2.194 qs_ot_p2m_diag 50 11.0 0.016 0.023 2.163 2.173 pw_transfer 1439 11.6 0.065 0.070 2.068 2.102 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 1.189 2.030 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.976 2.014 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.978 2.005 jit_kernel_multiply 10 16.1 0.789 1.965 0.789 1.965 multiply_cannon_metrocomm3 27432 15.5 0.038 0.040 1.193 1.941 grid_integrate_task_list 119 12.3 1.816 1.919 1.816 1.919 mp_sum_l 7287 12.8 1.276 1.843 1.276 1.843 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.842 1.843 ot_diis_step 108 11.5 0.012 0.013 1.791 1.792 jit_kernel_transpose 5 15.6 0.919 1.766 0.919 1.766 fft_wrap_pw1pw2_140 487 13.2 0.290 0.302 1.623 1.661 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.595 1.604 cp_fm_diag_elpa_base 50 14.0 1.561 1.579 1.593 1.602 potential_pw2rs 119 12.3 0.008 0.010 1.591 1.594 qs_energies_init_hamiltonians 11 5.9 0.217 0.453 1.571 1.573 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.433 1.444 fft3d_ps 1201 14.6 0.557 0.609 1.412 1.438 dbcsr_complete_redistribute 329 12.2 0.126 0.152 1.074 1.376 wfi_extrapolate 11 7.9 0.001 0.001 1.375 1.375 grid_collocate_task_list 119 9.7 1.248 1.370 1.248 1.370 mp_sum_d 4133 12.0 0.798 1.163 0.798 1.163 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.862 1.162 cp_fm_upper_to_full 72 14.2 0.816 1.158 0.816 1.158 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.125 1.143 multiply_cannon_sync_h2d 27432 15.5 0.981 1.051 0.981 1.051 mp_alltoall_i22 627 13.8 0.722 1.044 0.722 1.044 make_images_data 4572 15.5 0.044 0.048 0.885 0.988 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.762 0.911 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.868 multiply_cannon_metrocomm1 27432 15.5 0.034 0.034 0.383 0.864 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.847 0.852 mp_alltoall_d11v 2130 13.8 0.767 0.849 0.767 0.849 create_qs_kind_set 1 2.0 0.000 0.001 0.847 0.849 read_qs_kind 2 3.0 0.054 0.816 0.847 0.849 parser_read_line 2807 4.0 0.001 0.001 0.793 0.837 parser_read_line_low 5 5.0 0.001 0.009 0.793 0.836 broadcast_input_information 5 6.0 0.002 0.002 0.792 0.836 mp_bcast_i_src 20 7.0 0.776 0.828 0.776 0.828 cp_fm_cholesky_invert 11 10.9 0.799 0.804 0.799 0.804 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=39.196000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=519.454545, yerr=3.916483 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 597.495808E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.027 35.026 35.026 qs_mol_dyn_low 1 2.0 0.003 0.003 34.849 34.855 qs_forces 11 3.9 0.002 0.002 34.780 34.781 qs_energies 11 4.9 0.007 0.009 32.884 32.888 scf_env_do_scf 11 5.9 0.001 0.002 26.780 26.780 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.572 23.573 velocity_verlet 10 3.0 0.001 0.001 17.652 17.655 dbcsr_multiply_generic 2286 12.5 0.094 0.098 15.319 15.503 qs_scf_new_mos 108 7.5 0.001 0.001 13.867 13.915 qs_scf_loop_do_ot 108 8.5 0.001 0.001 13.867 13.914 ot_scf_mini 108 9.5 0.002 0.002 13.066 13.093 multiply_cannon 2286 13.5 0.233 0.243 11.717 12.405 multiply_cannon_loop 2286 14.5 0.333 0.345 10.678 11.118 ot_mini 108 10.5 0.001 0.001 7.373 7.413 multiply_cannon_multrec 9144 15.5 1.597 1.829 7.122 7.375 rebuild_ks_matrix 119 8.3 0.000 0.000 7.081 7.133 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.081 7.133 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.276 6.324 qs_ot_get_derivative 108 11.5 0.001 0.001 5.920 5.952 dbcsr_mm_accdrv_process 12550 15.8 3.639 5.324 5.425 5.541 sum_up_and_integrate 119 10.3 0.037 0.041 4.559 4.565 init_scf_run 11 5.9 0.000 0.001 4.564 4.564 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.564 4.564 integrate_v_rspace 119 11.3 0.003 0.003 4.522 4.528 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.469 4.493 calculate_rho_elec 119 8.7 0.060 0.061 4.469 4.492 qs_ot_get_p 119 10.4 0.001 0.001 3.631 3.672 mp_waitall_1 115863 16.7 2.926 3.615 2.926 3.615 init_scf_loop 11 6.9 0.000 0.000 3.171 3.173 jit_kernel_multiply 10 15.9 1.748 3.150 1.748 3.150 calculate_first_density_matrix 1 7.0 0.000 0.000 3.061 3.062 rs_pw_transfer 974 11.9 0.008 0.008 2.753 2.906 density_rs2pw 119 9.7 0.004 0.004 2.604 2.755 make_m2s 4572 13.5 0.034 0.036 2.494 2.678 pw_transfer 1439 11.6 0.066 0.069 2.596 2.616 make_images 4572 14.5 0.267 0.300 2.403 2.587 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.502 2.523 calculate_dm_sparse 119 9.5 0.000 0.000 2.363 2.394 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.376 2.393 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.371 2.376 prepare_preconditioner 11 7.9 0.000 0.000 2.312 2.319 make_preconditioner 11 8.9 0.000 0.000 2.312 2.319 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.158 2.208 potential_pw2rs 119 12.3 0.010 0.011 2.152 2.166 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.136 2.140 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.055 2.058 fft_wrap_pw1pw2_140 487 13.2 0.366 0.376 1.942 1.964 grid_integrate_task_list 119 12.3 1.887 1.943 1.887 1.943 fft3d_ps 1201 14.6 0.566 0.579 1.834 1.852 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.821 1.837 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.748 1.753 cp_fm_diag_elpa_base 50 14.0 1.717 1.736 1.746 1.752 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.671 1.679 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.573 1.590 make_images_data 4572 15.5 0.038 0.042 1.154 1.545 hybrid_alltoall_any 4725 16.4 0.063 0.176 1.077 1.465 grid_collocate_task_list 119 9.7 1.323 1.459 1.323 1.459 wfi_extrapolate 11 7.9 0.001 0.001 1.444 1.444 ot_diis_step 108 11.5 0.012 0.013 1.416 1.417 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.256 1.311 apply_single 119 13.6 0.000 0.000 1.256 1.311 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.988 1.305 mp_alltoall_d11v 2130 13.8 1.218 1.292 1.218 1.292 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.238 1.239 cp_fm_cholesky_invert 11 10.9 1.186 1.189 1.186 1.189 mp_alltoall_z22v 1201 16.6 1.140 1.167 1.140 1.167 make_images_sizes 4572 15.5 0.005 0.005 0.674 1.025 mp_alltoall_i44 4572 16.5 0.669 1.020 0.669 1.020 acc_transpose_blocks 9144 15.5 0.038 0.039 0.934 0.944 mp_sum_l 7287 12.8 0.674 0.936 0.674 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.929 0.934 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.920 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.890 0.902 mp_allgather_i34 2286 14.5 0.369 0.879 0.369 0.879 multiply_cannon_sync_h2d 9144 15.5 0.700 0.773 0.700 0.773 dbcsr_complete_redistribute 329 12.2 0.161 0.170 0.713 0.760 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.692 0.744 mp_waitany 5200 13.7 0.573 0.707 0.573 0.707 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=35.026000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.090909, yerr=4.144099 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 745.156608E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.034 42.550 42.550 qs_mol_dyn_low 1 2.0 0.003 0.003 42.332 42.343 qs_forces 11 3.9 0.002 0.003 42.055 42.057 qs_energies 11 4.9 0.001 0.002 40.049 40.053 scf_env_do_scf 11 5.9 0.002 0.014 34.215 34.215 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.209 26.211 velocity_verlet 10 3.0 0.001 0.001 23.775 23.807 dbcsr_multiply_generic 2286 12.5 0.110 0.119 18.012 18.175 qs_scf_new_mos 108 7.5 0.001 0.001 16.403 16.499 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.402 16.498 ot_scf_mini 108 9.5 0.002 0.002 15.317 15.415 multiply_cannon 2286 13.5 0.299 0.305 13.762 14.965 multiply_cannon_loop 2286 14.5 0.344 0.350 12.542 13.779 ot_mini 108 10.5 0.001 0.001 9.360 9.473 multiply_cannon_multrec 9144 15.5 3.458 5.103 8.714 9.156 init_scf_loop 11 6.9 0.000 0.000 7.961 7.964 qs_ot_get_derivative 108 11.5 0.001 0.001 7.363 7.460 rebuild_ks_matrix 119 8.3 0.000 0.000 7.300 7.439 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.300 7.438 prepare_preconditioner 11 7.9 0.000 0.000 6.995 7.009 make_preconditioner 11 8.9 0.000 0.000 6.995 7.009 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.461 6.889 dbcsr_mm_accdrv_process 12550 15.8 4.206 5.677 5.134 6.871 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.584 6.710 cp_fm_upper_to_full 72 14.2 3.272 4.754 3.272 4.754 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.310 4.330 calculate_rho_elec 119 8.7 0.118 0.121 4.309 4.330 sum_up_and_integrate 119 10.3 0.064 0.066 4.038 4.045 integrate_v_rspace 119 11.3 0.003 0.003 3.973 3.979 init_scf_run 11 5.9 0.000 0.001 3.797 3.797 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.796 3.797 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.637 3.408 qs_ot_get_p 119 10.4 0.001 0.001 3.243 3.374 mp_waitall_1 94719 16.7 2.368 3.275 2.368 3.275 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.513 2.956 apply_single 119 13.6 0.000 0.000 2.513 2.956 dbcsr_complete_redistribute 329 12.2 0.286 0.293 2.085 2.955 pw_transfer 1439 11.6 0.068 0.069 2.715 2.720 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.763 2.637 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.616 2.621 mp_alltoall_i22 627 13.8 1.594 2.471 1.594 2.471 make_m2s 4572 13.5 0.037 0.038 2.301 2.458 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.529 2.396 make_images 4572 14.5 0.353 0.385 2.180 2.336 calculate_first_density_matrix 1 7.0 0.000 0.000 2.312 2.315 mp_sum_l 7287 12.8 1.394 2.236 1.394 2.236 density_rs2pw 119 9.7 0.004 0.004 2.209 2.228 fft_wrap_pw1pw2_140 487 13.2 0.619 0.622 2.207 2.213 calculate_dm_sparse 119 9.5 0.000 0.000 2.184 2.203 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.324 2.149 grid_integrate_task_list 119 12.3 2.064 2.087 2.064 2.087 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.006 2.006 ot_diis_step 108 11.5 0.014 0.014 1.975 1.975 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.959 1.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.765 1.767 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.682 1.731 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.681 1.716 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.682 1.683 fft3d_ps 1201 14.6 0.602 0.612 1.628 1.634 grid_collocate_task_list 119 9.7 1.511 1.527 1.511 1.527 rs_pw_transfer 974 11.9 0.009 0.009 1.492 1.515 jit_kernel_multiply 6 15.6 0.900 1.486 0.900 1.486 cp_fm_cholesky_invert 11 10.9 1.425 1.428 1.425 1.428 potential_pw2rs 119 12.3 0.014 0.015 1.419 1.421 wfi_extrapolate 11 7.9 0.001 0.001 1.412 1.412 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.400 1.401 cp_fm_diag_elpa_base 50 14.0 1.251 1.308 1.399 1.399 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.091 1.301 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.245 1.255 make_images_data 4572 15.5 0.042 0.045 1.030 1.219 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.132 1.150 mp_alltoall_d11v 2130 13.8 1.109 1.145 1.109 1.145 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.099 1.114 multiply_cannon_sync_h2d 9144 15.5 1.040 1.044 1.040 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.982 1.038 acc_transpose_blocks 9144 15.5 0.038 0.039 0.931 0.946 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.930 0.944 qs_create_task_list 11 7.9 0.000 0.000 0.930 0.941 generate_qs_task_list 11 8.9 0.368 0.387 0.929 0.940 mp_alltoall_z22v 1201 16.6 0.890 0.908 0.890 0.908 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.550000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=694.727273, yerr=13.652112 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 499.875840E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 2297799. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.029 81.774 81.775 qs_mol_dyn_low 1 2.0 0.003 0.003 81.394 81.522 qs_forces 11 3.9 0.003 0.003 81.258 81.258 qs_energies 11 4.9 0.002 0.003 78.371 78.384 scf_env_do_scf 11 5.9 0.001 0.002 68.932 68.934 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 63.395 63.395 dbcsr_multiply_generic 2055 12.4 0.108 0.111 50.457 50.797 qs_scf_new_mos 99 7.5 0.000 0.001 46.011 46.130 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.011 46.130 ot_scf_mini 99 9.5 0.002 0.002 43.692 43.813 multiply_cannon 2055 13.4 0.179 0.183 42.500 43.461 multiply_cannon_loop 2055 14.4 1.518 1.552 41.532 42.535 velocity_verlet 10 3.0 0.001 0.002 42.385 42.386 ot_mini 99 10.5 0.001 0.001 25.840 25.962 qs_ot_get_derivative 99 11.5 0.001 0.001 19.101 19.209 multiply_cannon_multrec 49320 15.4 12.542 13.314 17.501 18.378 rebuild_ks_matrix 110 8.3 0.000 0.001 14.639 14.815 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.639 14.815 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.811 12.972 mp_waitall_1 220248 16.4 10.283 11.378 10.283 11.378 multiply_cannon_sync_h2d 49320 15.4 10.474 11.177 10.474 11.177 qs_ot_get_p 110 10.4 0.001 0.001 9.567 9.708 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.364 7.908 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.205 7.640 apply_single 110 13.6 0.000 0.001 7.205 7.640 multiply_cannon_metrocomm3 49320 15.4 0.079 0.085 6.144 7.541 init_scf_run 11 5.9 0.000 0.001 7.389 7.389 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.389 7.389 sum_up_and_integrate 110 10.3 0.036 0.042 7.141 7.157 integrate_v_rspace 110 11.3 0.003 0.004 7.105 7.129 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.615 6.722 calculate_rho_elec 110 8.6 0.021 0.026 6.614 6.721 ot_diis_step 99 11.5 0.005 0.006 6.563 6.563 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.516 6.532 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.745 5.746 init_scf_loop 11 6.9 0.000 0.000 5.508 5.509 dbcsr_mm_accdrv_process 87628 16.1 1.908 2.019 4.833 5.263 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.168 5.230 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.181 5.204 cp_fm_diag_elpa_base 48 14.0 5.169 5.192 5.180 5.202 mp_sum_l 6594 12.7 3.575 4.364 3.575 4.364 rs_pw_transfer 902 11.9 0.011 0.013 3.703 4.302 wfi_extrapolate 11 7.9 0.001 0.001 4.080 4.080 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.041 4.045 make_m2s 4110 13.4 0.060 0.064 3.923 4.034 density_rs2pw 110 9.6 0.004 0.005 3.395 4.025 make_images 4110 14.4 0.177 0.191 3.827 3.942 calculate_dm_sparse 110 9.5 0.001 0.001 3.815 3.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.666 3.728 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.258 3.430 grid_integrate_task_list 110 12.3 3.248 3.393 3.248 3.393 prepare_preconditioner 11 7.9 0.000 0.000 3.346 3.369 make_preconditioner 11 8.9 0.000 0.000 3.346 3.369 pw_transfer 1331 11.6 0.055 0.065 3.158 3.236 calculate_first_density_matrix 1 7.0 0.000 0.000 3.205 3.209 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.127 3.174 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.069 3.151 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.989 3.030 jit_kernel_multiply 13 15.9 2.651 2.863 2.651 2.863 fft_wrap_pw1pw2_140 451 13.1 0.454 0.499 2.619 2.708 potential_pw2rs 110 12.3 0.006 0.007 2.657 2.677 acc_transpose_blocks 49320 15.4 0.206 0.216 2.500 2.594 mp_waitany 14300 13.8 1.824 2.586 1.824 2.586 mp_alltoall_d11v 2046 13.8 2.011 2.492 2.011 2.492 grid_collocate_task_list 110 9.6 2.153 2.439 2.153 2.439 fft3d_ps 1111 14.6 0.795 0.877 2.262 2.311 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.957 1.977 make_images_data 4110 15.4 0.042 0.046 1.784 1.917 cp_fm_cholesky_invert 11 10.9 1.866 1.870 1.866 1.870 mp_sum_d 3889 11.9 1.282 1.858 1.282 1.858 hybrid_alltoall_any 4261 16.3 0.081 0.482 1.536 1.843 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.775 1.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.775000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.545455, yerr=2.016461 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 587.288576E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2636592. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.237 79.882 79.883 qs_mol_dyn_low 1 2.0 0.005 0.014 79.188 79.197 qs_forces 11 3.9 0.029 0.209 77.897 77.900 qs_energies 11 4.9 0.002 0.009 74.200 74.207 scf_env_do_scf 11 5.9 0.001 0.002 63.956 63.959 scf_env_do_scf_inner_loop 99 6.5 0.003 0.009 54.587 54.588 velocity_verlet 10 3.0 0.030 0.230 41.564 41.586 dbcsr_multiply_generic 2055 12.4 0.134 0.247 40.775 40.921 qs_scf_new_mos 99 7.5 0.001 0.001 36.520 36.646 qs_scf_loop_do_ot 99 8.5 0.001 0.001 36.520 36.645 ot_scf_mini 99 9.5 0.003 0.005 34.781 34.904 multiply_cannon 2055 13.4 0.222 0.243 32.844 34.007 multiply_cannon_loop 2055 14.4 0.923 0.949 31.324 32.029 ot_mini 99 10.5 0.001 0.001 19.847 19.986 multiply_cannon_multrec 24660 15.4 7.580 9.020 14.566 16.355 rebuild_ks_matrix 110 8.3 0.000 0.000 14.834 14.936 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 14.833 14.936 qs_ot_get_derivative 99 11.5 0.024 0.369 13.934 14.059 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.065 13.157 mp_waitall_1 176588 16.5 9.240 11.436 9.240 11.436 init_scf_loop 11 6.9 0.001 0.004 9.317 9.318 multiply_cannon_metrocomm3 24660 15.4 0.067 0.071 5.651 8.314 qs_ot_get_p 110 10.4 0.001 0.002 7.999 8.180 multiply_cannon_sync_h2d 24660 15.4 7.009 7.968 7.009 7.968 init_scf_run 11 5.9 0.000 0.001 7.705 7.706 scf_env_initial_rho_setup 11 6.9 0.001 0.006 7.705 7.706 sum_up_and_integrate 110 10.3 0.053 0.059 7.525 7.535 integrate_v_rspace 110 11.3 0.003 0.004 7.472 7.485 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.593 7.355 apply_single 110 13.6 0.000 0.001 6.593 7.354 dbcsr_mm_accdrv_process 52282 16.1 4.870 6.614 6.824 7.340 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.197 7.213 calculate_rho_elec 110 8.6 0.040 0.048 7.197 7.213 prepare_preconditioner 11 7.9 0.000 0.000 7.125 7.150 make_preconditioner 11 8.9 0.000 0.001 7.125 7.150 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.617 6.829 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.463 6.177 ot_diis_step 99 11.5 0.011 0.013 5.854 5.855 make_m2s 4110 13.4 0.057 0.060 5.302 5.835 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.773 5.797 make_images 4110 14.4 0.400 0.443 5.189 5.719 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.260 5.260 rs_pw_transfer 902 11.9 0.012 0.013 4.100 4.712 density_rs2pw 110 9.6 0.004 0.005 4.172 4.670 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.533 4.543 cp_fm_diag_elpa_base 48 14.0 4.475 4.503 4.530 4.541 pw_transfer 1331 11.6 0.067 0.074 4.294 4.481 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.186 4.375 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 4.214 4.217 wfi_extrapolate 11 7.9 0.003 0.019 3.853 3.853 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.657 3.793 calculate_first_density_matrix 1 7.0 0.001 0.006 3.739 3.742 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.547 3.616 cp_fm_cholesky_invert 11 10.9 3.494 3.501 3.494 3.501 fft_wrap_pw1pw2_140 451 13.1 0.519 0.534 3.314 3.499 make_images_data 4110 15.4 0.046 0.050 2.839 3.482 fft3d_ps 1111 14.6 1.111 1.334 3.159 3.373 grid_integrate_task_list 110 12.3 3.140 3.351 3.140 3.351 calculate_dm_sparse 110 9.5 0.001 0.001 3.277 3.306 hybrid_alltoall_any 4261 16.3 0.102 0.442 2.413 3.281 potential_pw2rs 110 12.3 0.008 0.009 3.225 3.273 jit_kernel_multiply 11 16.3 1.599 2.822 1.599 2.822 mp_sum_l 6594 12.7 1.994 2.695 1.994 2.695 mp_alltoall_d11v 2046 13.8 2.158 2.608 2.158 2.608 grid_collocate_task_list 110 9.6 2.102 2.532 2.102 2.532 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.141 2.160 mp_waitany 10164 13.8 1.445 1.991 1.445 1.991 mp_allgather_i34 2055 14.4 0.845 1.961 0.845 1.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 1.944 1.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.930 1.940 cp_fm_cholesky_decompose 22 10.9 1.864 1.870 1.864 1.870 mp_alltoall_z22v 1111 16.6 1.743 1.870 1.743 1.870 acc_transpose_blocks 24660 15.4 0.111 0.115 1.778 1.849 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.797 1.808 multiply_cannon_metrocomm4 22605 15.4 0.077 0.082 0.787 1.710 dbcsr_complete_redistribute 325 12.2 0.234 0.299 1.390 1.687 rs_pw_transfer_RS2PW_140 121 11.5 0.204 0.216 1.078 1.636 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=79.883000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.454545, yerr=5.821022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 661.172224E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.067 65.272 65.273 qs_mol_dyn_low 1 2.0 0.003 0.004 64.373 64.384 qs_forces 11 3.9 0.003 0.007 63.938 63.939 qs_energies 11 4.9 0.007 0.028 60.629 60.635 scf_env_do_scf 11 5.9 0.001 0.002 51.370 51.371 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.475 42.476 velocity_verlet 10 3.0 0.001 0.002 34.086 34.099 dbcsr_multiply_generic 2055 12.4 0.108 0.115 30.862 31.215 qs_scf_new_mos 99 7.5 0.001 0.001 26.879 26.990 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.879 26.989 ot_scf_mini 99 9.5 0.002 0.004 25.616 25.727 multiply_cannon 2055 13.4 0.214 0.223 24.016 25.060 multiply_cannon_loop 2055 14.4 0.617 0.634 22.756 23.934 ot_mini 99 10.5 0.001 0.001 14.726 14.851 rebuild_ks_matrix 110 8.3 0.000 0.000 12.769 12.934 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.769 12.933 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.224 11.376 mp_waitall_1 139946 16.5 8.038 10.938 8.038 10.938 multiply_cannon_multrec 16440 15.4 3.757 4.692 9.663 10.781 qs_ot_get_derivative 99 11.5 0.001 0.001 10.109 10.229 init_scf_loop 11 6.9 0.001 0.003 8.855 8.857 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 5.156 8.108 prepare_preconditioner 11 7.9 0.000 0.000 7.045 7.065 make_preconditioner 11 8.9 0.000 0.002 7.045 7.065 sum_up_and_integrate 110 10.3 0.060 0.062 6.805 6.821 init_scf_run 11 5.9 0.000 0.001 6.766 6.767 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.766 6.767 integrate_v_rspace 110 11.3 0.003 0.004 6.744 6.761 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.386 6.726 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.361 6.371 calculate_rho_elec 110 8.6 0.059 0.060 6.361 6.371 dbcsr_mm_accdrv_process 34862 16.1 4.726 5.471 5.762 6.237 qs_ot_get_p 110 10.4 0.001 0.001 5.936 6.088 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.249 5.862 apply_single 110 13.6 0.000 0.000 5.249 5.862 make_m2s 4110 13.4 0.050 0.052 4.365 4.731 density_rs2pw 110 9.6 0.004 0.005 3.384 4.643 make_images 4110 14.4 0.392 0.512 4.248 4.614 ot_diis_step 99 11.5 0.011 0.011 4.583 4.584 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.712 4.498 rs_pw_transfer 902 11.9 0.013 0.034 3.048 4.309 qs_ot_p2m_diag 48 11.0 0.042 0.045 4.209 4.216 multiply_cannon_sync_h2d 16440 15.4 3.611 4.167 3.611 4.167 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.855 3.856 pw_transfer 1331 11.6 0.066 0.073 3.841 3.851 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.734 3.745 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.614 3.615 calculate_first_density_matrix 1 7.0 0.000 0.002 3.505 3.506 grid_integrate_task_list 110 12.3 3.168 3.484 3.168 3.484 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.274 3.336 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.254 3.262 cp_fm_diag_elpa_base 48 14.0 3.187 3.218 3.252 3.261 fft_wrap_pw1pw2_140 451 13.1 0.640 0.646 3.149 3.165 wfi_extrapolate 11 7.9 0.001 0.001 3.049 3.049 make_images_data 4110 15.4 0.042 0.047 2.533 3.022 hybrid_alltoall_any 4261 16.3 0.105 0.375 2.295 2.935 cp_fm_cholesky_invert 11 10.9 2.846 2.852 2.846 2.852 calculate_dm_sparse 110 9.5 0.001 0.001 2.804 2.844 mp_waitany 17072 13.8 1.383 2.712 1.383 2.712 mp_sum_l 6594 12.7 1.867 2.603 1.867 2.603 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 2.500 2.576 grid_collocate_task_list 110 9.6 2.135 2.566 2.135 2.566 fft3d_ps 1111 14.6 1.093 1.104 2.553 2.560 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.918 2.537 potential_pw2rs 110 12.3 0.011 0.011 2.470 2.490 acc_transpose_blocks 16440 15.4 0.073 0.076 2.256 2.475 mp_irecv_dv 48980 15.7 0.848 2.408 0.848 2.408 mp_alltoall_d11v 2046 13.8 1.951 2.351 1.951 2.351 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.180 1.064 2.323 acc_transpose_blocks_kernels 16440 16.4 0.207 0.214 1.818 2.029 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.006 2.007 dbcsr_complete_redistribute 325 12.2 0.322 0.348 1.484 1.949 jit_kernel_transpose 5 15.6 1.610 1.819 1.610 1.819 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.744 1.759 cp_fm_upper_to_full 70 14.2 1.378 1.751 1.378 1.751 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.733 1.745 mp_allgather_i34 2055 14.4 0.583 1.720 0.583 1.720 cp_fm_cholesky_decompose 22 10.9 1.685 1.710 1.685 1.710 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.353 1.501 jit_kernel_multiply 8 16.5 0.641 1.496 0.641 1.496 rs_gather_matrices 110 12.3 0.231 0.261 1.026 1.453 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.989 1.434 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.342 1.367 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.273000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.181818, yerr=9.053560 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 730.009600E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.027 70.051 70.051 qs_mol_dyn_low 1 2.0 0.003 0.003 69.647 69.657 qs_forces 11 3.9 0.003 0.004 69.578 69.578 qs_energies 11 4.9 0.001 0.002 66.092 66.095 scf_env_do_scf 11 5.9 0.000 0.001 56.019 56.022 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 44.134 44.135 velocity_verlet 10 3.0 0.001 0.001 37.883 37.885 dbcsr_multiply_generic 2055 12.4 0.115 0.121 32.353 32.623 qs_scf_new_mos 99 7.5 0.001 0.001 28.614 28.725 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.613 28.724 ot_scf_mini 99 9.5 0.003 0.003 26.942 27.037 multiply_cannon 2055 13.4 0.244 0.259 25.171 26.264 multiply_cannon_loop 2055 14.4 0.890 0.919 23.603 24.293 ot_mini 99 10.5 0.001 0.001 15.649 15.772 multiply_cannon_multrec 24660 15.4 4.218 6.791 13.078 14.245 rebuild_ks_matrix 110 8.3 0.000 0.000 12.522 12.616 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.521 12.616 init_scf_loop 11 6.9 0.000 0.000 11.842 11.843 qs_ot_get_derivative 99 11.5 0.001 0.001 11.440 11.542 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.039 11.120 prepare_preconditioner 11 7.9 0.000 0.000 10.060 10.077 make_preconditioner 11 8.9 0.000 0.000 10.060 10.077 dbcsr_mm_accdrv_process 52304 16.0 6.911 8.307 8.712 9.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.295 9.746 mp_waitall_1 121746 16.5 4.806 7.374 4.806 7.374 init_scf_run 11 5.9 0.000 0.001 6.857 6.857 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.856 6.857 sum_up_and_integrate 110 10.3 0.067 0.070 6.744 6.759 integrate_v_rspace 110 11.3 0.003 0.003 6.676 6.691 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.375 6.385 calculate_rho_elec 110 8.6 0.078 0.081 6.375 6.385 qs_ot_get_p 110 10.4 0.001 0.001 6.221 6.349 make_m2s 4110 13.4 0.059 0.062 5.461 5.877 make_images 4110 14.4 0.576 0.699 5.321 5.733 cp_fm_upper_to_full 70 14.2 3.314 4.806 3.314 4.806 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.621 4.671 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.887 4.495 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.332 4.349 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.078 4.202 apply_single 110 13.6 0.000 0.000 4.077 4.202 ot_diis_step 99 11.5 0.011 0.011 4.169 4.169 pw_transfer 1331 11.6 0.065 0.074 3.921 3.954 dbcsr_complete_redistribute 325 12.2 0.416 0.454 2.748 3.871 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.814 3.852 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.844 3.845 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.776 3.778 density_rs2pw 110 9.6 0.004 0.004 3.192 3.714 calculate_first_density_matrix 1 7.0 0.000 0.000 3.637 3.642 acc_transpose_blocks 24660 15.4 0.104 0.106 3.201 3.469 grid_integrate_task_list 110 12.3 3.288 3.449 3.288 3.449 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.377 3.411 multiply_cannon_sync_h2d 24660 15.4 3.186 3.358 3.186 3.358 rs_pw_transfer 902 11.9 0.010 0.011 2.597 3.323 calculate_dm_sparse 110 9.5 0.001 0.001 3.262 3.296 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.168 3.281 fft_wrap_pw1pw2_140 451 13.1 0.671 0.694 3.236 3.277 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.253 3.264 cp_fm_diag_elpa_base 48 14.0 3.101 3.162 3.251 3.262 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.152 3.212 make_images_data 4110 15.4 0.045 0.049 2.719 3.159 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.373 3.154 wfi_extrapolate 11 7.9 0.001 0.001 3.113 3.114 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.756 2.855 cp_fm_cholesky_invert 11 10.9 2.840 2.849 2.840 2.849 mp_alltoall_i22 605 13.7 1.643 2.804 1.643 2.804 acc_transpose_blocks_kernels 24660 16.4 0.306 0.311 2.538 2.800 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.680 2.681 fft3d_ps 1111 14.6 1.089 1.122 2.570 2.585 jit_kernel_transpose 5 15.6 2.233 2.493 2.233 2.493 grid_collocate_task_list 110 9.6 2.220 2.486 2.220 2.486 jit_kernel_multiply 10 15.8 1.468 2.402 1.468 2.402 mp_alltoall_d11v 2046 13.8 1.994 2.354 1.994 2.354 potential_pw2rs 110 12.3 0.013 0.013 2.307 2.317 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.836 1.853 mp_waitany 13376 13.8 1.141 1.843 1.141 1.843 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.730 1.762 cp_fm_cholesky_decompose 22 10.9 1.716 1.761 1.716 1.761 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.635 1.740 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.728 1.739 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.599 1.704 mp_allgather_i34 2055 14.4 0.668 1.701 0.668 1.701 mp_sum_l 6594 12.7 1.029 1.684 1.029 1.684 rs_pw_transfer_RS2PW_140 121 11.5 0.167 0.178 0.873 1.603 multiply_cannon_metrocomm4 20550 15.4 0.059 0.063 0.845 1.456 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=70.051000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.363636, yerr=9.068153 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 840.769536E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.034 59.190 59.190 qs_mol_dyn_low 1 2.0 0.003 0.003 58.481 58.491 qs_forces 11 3.9 0.003 0.003 58.409 58.410 qs_energies 11 4.9 0.002 0.002 54.672 54.676 scf_env_do_scf 11 5.9 0.000 0.001 45.071 45.071 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.054 37.055 velocity_verlet 10 3.0 0.001 0.001 31.956 31.960 dbcsr_multiply_generic 2055 12.4 0.105 0.109 25.120 25.264 qs_scf_new_mos 99 7.5 0.001 0.001 21.640 21.680 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.639 21.679 multiply_cannon 2055 13.4 0.246 0.263 19.458 20.732 ot_scf_mini 99 9.5 0.002 0.002 20.380 20.414 multiply_cannon_loop 2055 14.4 0.322 0.334 18.101 18.385 rebuild_ks_matrix 110 8.3 0.000 0.000 12.186 12.229 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.185 12.228 ot_mini 99 10.5 0.001 0.001 11.291 11.320 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.810 10.849 multiply_cannon_multrec 8220 15.4 3.245 4.682 8.234 9.288 mp_waitall_1 103326 16.6 6.432 8.242 6.432 8.242 init_scf_loop 11 6.9 0.000 0.000 7.968 7.969 qs_ot_get_derivative 99 11.5 0.001 0.001 7.422 7.454 sum_up_and_integrate 110 10.3 0.080 0.081 6.676 6.687 integrate_v_rspace 110 11.3 0.003 0.003 6.596 6.606 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.492 6.504 calculate_rho_elec 110 8.6 0.114 0.114 6.492 6.503 init_scf_run 11 5.9 0.000 0.001 6.465 6.465 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.464 6.465 prepare_preconditioner 11 7.9 0.000 0.000 6.246 6.252 make_preconditioner 11 8.9 0.000 0.000 6.246 6.252 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.833 5.905 dbcsr_mm_accdrv_process 17442 15.9 3.173 4.307 4.858 5.818 qs_ot_get_p 110 10.4 0.001 0.001 5.095 5.145 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 3.253 4.609 make_m2s 4110 13.4 0.039 0.040 4.324 4.578 make_images 4110 14.4 0.636 0.686 4.192 4.449 pw_transfer 1331 11.6 0.066 0.070 4.079 4.089 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.972 3.985 ot_diis_step 99 11.5 0.012 0.012 3.846 3.846 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.767 3.806 apply_single 110 13.6 0.000 0.000 3.767 3.806 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.654 3.657 density_rs2pw 110 9.6 0.004 0.004 3.170 3.549 grid_integrate_task_list 110 12.3 3.358 3.533 3.358 3.533 calculate_first_density_matrix 1 7.0 0.000 0.000 3.507 3.508 fft_wrap_pw1pw2_140 451 13.1 0.839 0.849 3.415 3.434 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.421 3.423 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.329 3.330 cp_fm_cholesky_invert 11 10.9 3.085 3.089 3.085 3.089 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.054 3.074 multiply_cannon_sync_h2d 8220 15.4 2.892 3.067 2.892 3.067 calculate_dm_sparse 110 9.5 0.001 0.001 2.868 2.904 make_images_data 4110 15.4 0.037 0.043 2.468 2.889 wfi_extrapolate 11 7.9 0.001 0.001 2.854 2.854 hybrid_alltoall_any 4261 16.3 0.200 0.866 2.398 2.799 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.775 2.782 cp_fm_diag_elpa_base 48 14.0 2.717 2.741 2.773 2.780 rs_pw_transfer 902 11.9 0.010 0.011 2.292 2.686 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.665 2.666 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.562 2.576 grid_collocate_task_list 110 9.6 2.313 2.572 2.313 2.572 fft3d_ps 1111 14.6 1.136 1.169 2.510 2.526 potential_pw2rs 110 12.3 0.016 0.016 2.159 2.162 jit_kernel_multiply 8 15.7 1.371 2.078 1.371 2.078 mp_alltoall_d11v 2046 13.8 1.764 2.037 1.764 2.037 acc_transpose_blocks 8220 15.4 0.035 0.036 1.987 1.995 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.769 1.991 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.856 1.878 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.819 1.828 acc_transpose_blocks_kernels 8220 16.4 0.109 0.111 1.747 1.752 cp_fm_cholesky_decompose 22 10.9 1.713 1.746 1.713 1.746 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.709 1.715 mp_allgather_i34 2055 14.4 0.540 1.663 0.540 1.663 jit_kernel_transpose 5 15.6 1.638 1.641 1.638 1.641 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.505 1.628 dbcsr_complete_redistribute 325 12.2 0.551 0.575 1.488 1.580 mp_waitany 9240 13.8 1.085 1.508 1.085 1.508 qs_create_task_list 11 7.9 0.000 0.000 1.217 1.320 generate_qs_task_list 11 8.9 0.375 0.443 1.216 1.319 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.480 1.257 rs_gather_matrices 110 12.3 0.322 0.365 0.988 1.223 rs_pw_transfer_RS2PW_140 121 11.5 0.162 0.166 0.812 1.219 mp_irecv_dv 24056 15.7 0.457 1.214 0.457 1.214 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.189 1.213 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.190000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=788.818182, yerr=14.465119 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 1.363796E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.040 100.179 100.179 qs_mol_dyn_low 1 2.0 0.003 0.003 99.846 99.854 qs_forces 11 3.9 0.003 0.003 99.773 99.774 qs_energies 11 4.9 0.002 0.002 95.410 95.416 scf_env_do_scf 11 5.9 0.001 0.001 83.222 83.222 velocity_verlet 10 3.0 0.001 0.001 62.512 62.519 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 51.778 51.779 dbcsr_multiply_generic 2055 12.4 0.122 0.128 33.870 34.349 qs_scf_new_mos 99 7.5 0.001 0.001 31.330 31.420 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.329 31.420 init_scf_loop 11 6.9 0.000 0.000 31.347 31.351 ot_scf_mini 99 9.5 0.002 0.002 29.260 29.325 prepare_preconditioner 11 7.9 0.000 0.000 29.108 29.131 make_preconditioner 11 8.9 0.000 0.000 29.108 29.131 make_full_inverse_cholesky 11 9.9 0.000 0.000 23.219 28.525 multiply_cannon 2055 13.4 0.350 0.372 24.504 24.902 multiply_cannon_loop 2055 14.4 0.343 0.344 22.279 23.494 cp_fm_upper_to_full 70 14.2 12.999 18.578 12.999 18.578 ot_mini 99 10.5 0.001 0.001 16.078 16.208 rebuild_ks_matrix 110 8.3 0.000 0.000 15.454 15.623 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 15.454 15.623 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.963 14.119 mp_waitall_1 84994 16.7 9.917 12.447 9.917 12.447 qs_ot_get_derivative 99 11.5 0.001 0.001 11.067 11.137 multiply_cannon_multrec 8220 15.4 4.401 4.591 10.897 11.051 dbcsr_complete_redistribute 325 12.2 1.006 1.017 7.709 10.990 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.527 9.816 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.873 9.146 qs_rho_update_rho_low 110 7.6 0.001 0.001 9.027 9.090 calculate_rho_elec 110 8.6 0.224 0.224 9.027 9.089 mp_alltoall_i22 605 13.7 5.590 8.823 5.590 8.823 sum_up_and_integrate 110 10.3 0.150 0.151 8.240 8.256 integrate_v_rspace 110 11.3 0.003 0.003 8.090 8.106 init_scf_run 11 5.9 0.000 0.001 7.784 7.784 scf_env_initial_rho_setup 11 6.9 0.000 0.000 7.783 7.784 qs_ot_get_p 110 10.4 0.001 0.001 7.368 7.533 make_m2s 4110 13.4 0.042 0.043 6.685 7.522 make_images 4110 14.4 0.878 0.927 6.494 7.328 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.810 6.997 cp_fm_cholesky_invert 11 10.9 6.911 6.917 6.911 6.917 dbcsr_mm_accdrv_process 11614 15.7 3.286 3.698 6.353 6.651 pw_transfer 1331 11.6 0.075 0.075 6.142 6.154 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.024 6.037 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.987 5.616 apply_single 110 13.6 0.000 0.000 4.987 5.616 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.231 5.243 make_images_data 4110 15.4 0.040 0.043 3.792 5.141 fft_wrap_pw1pw2_140 451 13.1 1.337 1.342 5.063 5.077 hybrid_alltoall_any 4261 16.3 0.256 0.550 3.639 4.938 ot_diis_step 99 11.5 0.016 0.016 4.917 4.918 calculate_dm_sparse 110 9.5 0.001 0.001 4.813 4.846 density_rs2pw 110 9.6 0.004 0.004 4.751 4.792 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.680 4.681 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.563 4.103 fft3d_ps 1111 14.6 1.303 1.315 3.952 3.967 multiply_cannon_sync_h2d 8220 15.4 3.950 3.954 3.950 3.954 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.923 3.924 cp_fm_diag_elpa_base 48 14.0 3.352 3.571 3.920 3.921 wfi_extrapolate 11 7.9 0.001 0.001 3.831 3.832 calculate_first_density_matrix 1 7.0 0.000 0.000 3.784 3.788 grid_integrate_task_list 110 12.3 3.706 3.771 3.706 3.771 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.711 3.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.203 3.208 rs_pw_transfer 902 11.9 0.010 0.011 3.149 3.176 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.116 3.162 potential_pw2rs 110 12.3 0.021 0.022 3.138 3.148 jit_kernel_multiply 10 15.3 2.861 3.116 2.861 3.116 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 3.067 3.085 cp_fm_cholesky_decompose 22 10.9 2.793 2.811 2.793 2.811 mp_alltoall_d11v 2046 13.8 2.650 2.773 2.650 2.773 grid_collocate_task_list 110 9.6 2.690 2.710 2.690 2.710 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.570 2.641 mp_sum_l 6594 12.7 1.779 2.527 1.779 2.527 mp_alltoall_z22v 1111 16.6 2.397 2.419 2.397 2.419 mp_allgather_i34 2055 14.4 0.993 2.418 0.993 2.418 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.235 2.287 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.091 2.189 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.094 2.109 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.997 2.095 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.973 2.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=100.179000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1210.090909, yerr=58.610354 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 630.464512E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 1765485. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.103 212.601 212.620 qs_mol_dyn_low 1 2.0 0.003 0.006 212.008 212.044 qs_forces 11 3.9 0.009 0.041 211.771 211.772 qs_energies 11 4.9 0.008 0.056 205.981 206.000 scf_env_do_scf 11 5.9 0.146 0.735 187.282 187.286 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 165.642 165.643 dbcsr_multiply_generic 2507 12.6 0.181 0.183 126.782 127.666 qs_scf_new_mos 117 7.6 0.001 0.001 125.154 125.474 qs_scf_loop_do_ot 117 8.6 0.001 0.001 125.153 125.474 velocity_verlet 10 3.0 0.023 0.111 125.269 125.271 ot_scf_mini 117 9.6 0.003 0.004 118.523 118.830 multiply_cannon 2507 13.6 0.237 0.251 103.110 104.857 multiply_cannon_loop 2507 14.6 2.075 2.132 100.920 102.537 ot_mini 117 10.6 0.001 0.002 66.517 66.827 multiply_cannon_multrec 60168 15.6 33.351 35.517 42.257 44.702 qs_ot_get_derivative 117 11.6 0.001 0.002 41.611 41.903 rebuild_ks_matrix 128 8.3 0.001 0.001 33.997 34.299 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.020 33.996 34.298 mp_waitall_1 267128 16.5 28.786 31.629 28.786 31.629 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.466 30.745 qs_ot_get_p 128 10.4 0.001 0.001 30.109 30.399 multiply_cannon_sync_h2d 60168 15.6 27.649 30.059 27.649 30.059 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.511 25.542 apply_single 128 13.6 0.001 0.001 24.511 25.541 ot_diis_step 117 11.6 0.009 0.015 24.694 24.696 qs_ot_p2m_diag 83 11.4 0.079 0.092 23.475 23.535 init_scf_loop 11 6.9 0.000 0.001 21.063 21.206 cp_dbcsr_syevd 83 12.4 0.005 0.005 21.075 21.076 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.683 19.932 cp_fm_diag_elpa 83 13.4 0.000 0.000 18.000 18.040 multiply_cannon_metrocomm3 60168 15.6 0.112 0.116 15.823 18.038 cp_fm_diag_elpa_base 83 14.4 17.926 17.969 17.996 18.034 prepare_preconditioner 11 7.9 0.000 0.000 16.139 16.206 make_preconditioner 11 8.9 0.000 0.000 16.139 16.206 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.376 15.553 sum_up_and_integrate 128 10.3 0.089 0.107 14.422 14.439 init_scf_run 11 5.9 0.000 0.001 14.389 14.390 scf_env_initial_rho_setup 11 6.9 0.000 0.002 14.389 14.390 integrate_v_rspace 128 11.3 0.004 0.005 14.332 14.350 make_m2s 5014 13.6 0.105 0.113 13.985 14.309 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.066 14.185 calculate_rho_elec 128 8.7 0.046 0.065 14.065 14.184 make_images 5014 14.6 0.399 0.417 13.801 14.136 density_rs2pw 128 9.7 0.006 0.007 7.311 10.556 mp_sum_l 7950 12.9 8.450 9.758 8.450 9.758 rs_pw_transfer 1046 11.9 0.020 0.029 5.987 9.305 cp_fm_cholesky_invert 11 10.9 9.236 9.244 9.236 9.244 wfi_extrapolate 11 7.9 0.003 0.017 9.225 9.225 dbcsr_mm_accdrv_process 124484 16.2 3.155 3.316 8.469 9.190 calculate_dm_sparse 128 9.5 0.001 0.001 8.701 8.803 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.443 8.690 multiply_cannon_metrocomm1 60168 15.6 0.086 0.091 6.286 8.344 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 8.156 8.169 pw_transfer 1547 11.6 0.074 0.086 7.963 8.146 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.867 7.960 fft_wrap_pw1pw2 1291 12.7 0.011 0.012 7.761 7.945 make_images_data 5014 15.6 0.064 0.069 6.846 7.749 grid_integrate_task_list 128 12.3 7.079 7.552 7.079 7.552 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 7.287 7.423 hybrid_alltoall_any 5200 16.5 0.289 2.261 5.989 7.290 fft_wrap_pw1pw2_140 523 13.2 1.273 1.314 6.744 6.935 mp_waitany 16020 13.9 2.757 6.135 2.757 6.135 grid_collocate_task_list 128 9.7 4.669 5.965 4.669 5.965 fft3d_ps 1291 14.7 2.142 2.776 5.527 5.834 mp_alltoall_d11v 2415 14.1 4.351 5.674 4.351 5.674 rs_pw_transfer_RS2PW_140 139 11.5 0.277 0.289 2.173 5.454 calculate_first_density_matrix 1 7.0 0.000 0.001 4.982 4.987 potential_pw2rs 128 12.3 0.009 0.011 4.738 4.757 cp_fm_cholesky_decompose 22 10.9 4.695 4.708 4.695 4.708 mp_sum_d 4465 12.1 3.527 4.260 3.527 4.260 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=212.620000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.818182, yerr=6.478190 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 832.495616E+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 4007 58147. MP_Allreduce 11096 960. MP_Sync 86 MP_Alltoall 1955 6220722. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.031 191.403 191.405 qs_mol_dyn_low 1 2.0 0.003 0.003 191.045 191.059 qs_forces 11 3.9 0.004 0.005 190.964 190.966 qs_energies 11 4.9 0.002 0.003 184.104 184.113 scf_env_do_scf 11 5.9 0.001 0.001 166.226 166.236 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 133.102 133.106 velocity_verlet 10 3.0 0.001 0.001 119.233 119.235 dbcsr_multiply_generic 2485 12.5 0.188 0.193 98.264 99.433 qs_scf_new_mos 116 7.6 0.001 0.001 93.136 93.659 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.135 93.658 ot_scf_mini 116 9.6 0.004 0.004 88.386 88.978 multiply_cannon 2485 13.5 0.471 0.527 78.349 82.404 multiply_cannon_loop 2485 14.5 1.233 1.270 75.011 77.687 ot_mini 116 10.6 0.001 0.001 49.657 50.201 mp_waitall_1 212858 16.6 24.581 39.862 24.581 39.862 multiply_cannon_multrec 29820 15.5 21.712 25.800 31.971 36.518 rebuild_ks_matrix 127 8.3 0.001 0.001 33.318 33.824 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 33.318 33.823 init_scf_loop 11 6.9 0.000 0.000 33.035 33.036 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.910 30.373 multiply_cannon_metrocomm3 29820 15.5 0.089 0.094 15.902 29.596 prepare_preconditioner 11 7.9 0.000 0.000 28.522 28.589 make_preconditioner 11 8.9 0.000 0.000 28.522 28.589 qs_ot_get_derivative 116 11.6 0.001 0.002 27.914 28.476 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.228 27.758 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.825 22.969 apply_single 127 13.6 0.001 0.001 21.824 22.968 ot_diis_step 116 11.6 0.014 0.015 21.563 21.573 qs_ot_get_p 127 10.4 0.001 0.001 20.622 21.262 multiply_cannon_sync_h2d 29820 15.5 18.975 21.231 18.975 21.231 cp_fm_cholesky_invert 11 10.9 16.765 16.777 16.765 16.777 make_m2s 4970 13.5 0.089 0.098 14.185 15.880 qs_ot_p2m_diag 82 11.4 0.185 0.212 15.728 15.760 make_images 4970 14.5 1.148 1.334 13.972 15.662 sum_up_and_integrate 127 10.3 0.115 0.132 14.718 14.750 integrate_v_rspace 127 11.3 0.003 0.004 14.603 14.636 cp_dbcsr_syevd 82 12.4 0.005 0.006 14.607 14.608 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.202 14.235 calculate_rho_elec 127 8.7 0.087 0.103 14.201 14.235 init_scf_run 11 5.9 0.000 0.001 13.052 13.053 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.052 13.053 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.395 11.422 cp_fm_diag_elpa_base 82 14.4 11.123 11.203 11.390 11.412 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 10.867 11.265 multiply_cannon_metrocomm4 27335 15.5 0.101 0.115 3.752 10.592 make_images_data 4970 15.5 0.062 0.072 8.579 10.591 dbcsr_mm_accdrv_process 61726 16.2 4.839 5.997 9.721 10.523 hybrid_alltoall_any 5155 16.4 0.338 1.493 7.381 10.245 mp_irecv_dv 68888 16.3 3.553 10.201 3.553 10.201 density_rs2pw 127 9.7 0.006 0.007 7.448 9.900 pw_transfer 1535 11.6 0.085 0.096 8.845 8.896 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 8.622 8.675 wfi_extrapolate 11 7.9 0.001 0.001 8.523 8.523 rs_pw_transfer 1038 11.9 0.014 0.017 5.694 8.152 fft_wrap_pw1pw2_140 519 13.2 1.329 1.354 7.600 7.674 grid_integrate_task_list 127 12.3 7.128 7.531 7.128 7.531 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.416 7.278 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.149 7.157 cp_fm_cholesky_decompose 22 10.9 6.914 6.986 6.914 6.986 calculate_dm_sparse 127 9.5 0.001 0.001 6.795 6.937 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.205 6.373 mp_sum_l 7884 12.9 3.956 6.125 3.956 6.125 fft3d_ps 1281 14.7 2.791 2.958 5.945 5.994 grid_collocate_task_list 127 9.7 4.746 5.929 4.746 5.929 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.284 5.348 potential_pw2rs 127 12.3 0.015 0.018 4.990 5.004 mp_alltoall_d11v 2401 14.1 4.383 5.001 4.383 5.001 mp_allgather_i34 2485 14.5 1.912 4.991 1.912 4.991 mp_waitany 11660 13.9 2.485 4.974 2.485 4.974 rs_pw_transfer_RS2PW_140 138 11.5 0.346 0.367 2.074 4.504 calculate_first_density_matrix 1 7.0 0.000 0.000 4.337 4.338 mp_sum_d 4454 12.1 2.718 4.141 2.718 4.141 dbcsr_complete_redistribute 393 12.7 0.828 0.921 3.190 4.051 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.795 3.837 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.405000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=793.636364, yerr=0.771389 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 946.937856E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.039 0.128 180.463 180.464 qs_mol_dyn_low 1 2.0 0.003 0.003 179.881 179.895 qs_forces 11 3.9 0.066 0.501 179.511 179.514 qs_energies 11 4.9 0.002 0.008 172.325 172.337 scf_env_do_scf 11 5.9 0.001 0.002 155.329 155.330 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.582 119.583 velocity_verlet 10 3.0 0.001 0.002 114.534 114.545 dbcsr_multiply_generic 2507 12.6 0.180 0.186 83.258 84.278 qs_scf_new_mos 117 7.6 0.001 0.001 80.936 81.301 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.936 81.300 ot_scf_mini 117 9.6 0.004 0.005 76.757 77.130 multiply_cannon 2507 13.6 0.500 0.520 62.954 67.429 multiply_cannon_loop 2507 14.6 0.864 0.894 59.999 62.567 ot_mini 117 10.6 0.001 0.001 42.860 43.248 mp_waitall_1 170520 16.6 25.632 36.134 25.632 36.134 init_scf_loop 11 6.9 0.001 0.009 35.644 35.645 rebuild_ks_matrix 128 8.3 0.001 0.001 31.003 31.481 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 31.002 31.480 prepare_preconditioner 11 7.9 0.000 0.001 31.134 31.188 make_preconditioner 11 8.9 0.000 0.001 31.134 31.188 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.765 30.176 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.271 28.763 multiply_cannon_multrec 20056 15.6 13.480 16.825 22.887 26.131 multiply_cannon_metrocomm3 20056 15.6 0.058 0.061 15.407 25.850 qs_ot_get_derivative 117 11.6 0.002 0.002 22.960 23.353 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.012 21.064 apply_single 128 13.6 0.001 0.001 20.012 21.064 ot_diis_step 117 11.6 0.018 0.018 19.796 19.797 qs_ot_get_p 128 10.4 0.001 0.002 19.225 19.681 multiply_cannon_sync_h2d 20056 15.6 14.107 15.965 14.107 15.965 make_m2s 5014 13.6 0.081 0.085 15.254 15.929 make_images 5014 14.6 1.171 1.270 15.017 15.692 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.994 15.006 sum_up_and_integrate 128 10.3 0.132 0.144 14.753 14.781 integrate_v_rspace 128 11.3 0.004 0.005 14.620 14.647 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.592 14.626 calculate_rho_elec 128 8.7 0.130 0.145 14.591 14.625 cp_fm_cholesky_invert 11 10.9 14.595 14.605 14.595 14.605 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.952 13.953 init_scf_run 11 5.9 0.000 0.001 11.925 11.926 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.925 11.926 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.732 10.747 cp_fm_diag_elpa_base 83 14.4 10.317 10.474 10.729 10.743 make_images_data 5014 15.6 0.058 0.067 9.559 10.666 hybrid_alltoall_any 5200 16.5 0.433 1.985 8.251 10.047 multiply_cannon_metrocomm4 17549 15.6 0.061 0.072 3.489 9.532 density_rs2pw 128 9.7 0.006 0.006 7.339 9.391 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 9.060 9.330 mp_irecv_dv 50230 16.2 3.369 9.284 3.369 9.284 pw_transfer 1547 11.6 0.086 0.104 9.020 9.119 dbcsr_mm_accdrv_process 41502 16.2 4.945 5.934 8.861 8.991 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 8.796 8.900 fft_wrap_pw1pw2_140 523 13.2 1.419 1.445 7.763 7.877 grid_integrate_task_list 128 12.3 7.237 7.793 7.237 7.793 wfi_extrapolate 11 7.9 0.001 0.001 7.485 7.485 cp_fm_cholesky_decompose 22 10.9 7.427 7.444 7.427 7.444 rs_pw_transfer 1046 11.9 0.014 0.015 5.327 7.421 cp_fm_upper_to_full 105 14.8 5.686 7.231 5.686 7.231 dbcsr_complete_redistribute 395 12.7 1.165 1.189 4.617 6.358 calculate_dm_sparse 128 9.5 0.001 0.001 6.244 6.330 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.271 6.287 fft3d_ps 1291 14.7 2.744 2.971 5.908 6.004 grid_collocate_task_list 128 9.7 4.974 5.888 4.974 5.888 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.365 5.496 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.577 5.260 mp_alltoall_d11v 2415 14.1 4.495 5.193 4.495 5.193 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.415 5.149 potential_pw2rs 128 12.3 0.021 0.024 4.815 4.832 mp_allgather_i34 2507 14.6 1.470 4.768 1.470 4.768 mp_sum_l 7950 12.9 3.284 4.663 3.284 4.663 mp_waitany 11748 13.9 2.428 4.619 2.428 4.619 calculate_first_density_matrix 1 7.0 0.000 0.002 4.246 4.247 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.066 4.084 transfer_fm_to_dbcsr 11 9.9 0.019 0.027 2.350 4.076 rs_pw_transfer_RS2PW_140 139 11.5 0.328 0.349 1.977 4.064 mp_alltoall_i22 716 14.1 1.993 3.854 1.993 3.854 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 3.780 3.781 mp_sum_d 4468 12.1 2.627 3.658 2.627 3.658 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.606 3.644 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=180.464000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=898.000000, yerr=10.962747 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/18/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 4.353788E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.171448E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770319E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379241840 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58354. MP_Allreduce 11057 1083. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.030 192.089 192.090 qs_mol_dyn_low 1 2.0 0.003 0.003 191.638 191.676 qs_forces 11 3.9 0.004 0.005 191.489 191.496 qs_energies 11 4.9 0.002 0.003 184.199 184.209 scf_env_do_scf 11 5.9 0.001 0.002 166.549 166.556 velocity_verlet 10 3.0 0.001 0.001 126.767 126.771 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 119.671 119.672 qs_scf_new_mos 117 7.6 0.001 0.001 81.657 81.943 qs_scf_loop_do_ot 117 8.6 0.001 0.001 81.657 81.942 dbcsr_multiply_generic 2507 12.6 0.187 0.194 80.836 81.484 ot_scf_mini 117 9.6 0.003 0.003 77.107 77.359 multiply_cannon 2507 13.6 0.557 0.590 55.997 58.234 multiply_cannon_loop 2507 14.6 1.185 1.211 52.321 53.825 init_scf_loop 11 6.9 0.000 0.000 46.752 46.753 ot_mini 117 10.6 0.001 0.001 42.794 43.042 prepare_preconditioner 11 7.9 0.000 0.000 42.634 42.656 make_preconditioner 11 8.9 0.000 0.000 42.634 42.656 make_full_inverse_cholesky 11 9.9 0.012 0.022 36.234 41.254 multiply_cannon_multrec 30084 15.6 14.085 18.907 26.156 30.785 rebuild_ks_matrix 128 8.3 0.001 0.001 30.219 30.445 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.218 30.444 mp_waitall_1 147882 16.7 17.698 27.554 17.698 27.554 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.180 27.375 qs_ot_get_derivative 117 11.6 0.002 0.002 23.202 23.459 make_m2s 5014 13.6 0.095 0.100 20.384 21.370 make_images 5014 14.6 1.944 2.270 20.077 21.064 qs_ot_get_p 128 10.4 0.001 0.001 20.450 20.727 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.012 19.557 apply_single 128 13.6 0.001 0.001 19.011 19.557 ot_diis_step 117 11.6 0.017 0.018 19.457 19.458 cp_fm_upper_to_full 105 14.8 11.441 16.855 11.441 16.855 cp_fm_cholesky_invert 11 10.9 16.338 16.346 16.338 16.346 qs_ot_p2m_diag 83 11.4 0.342 0.389 16.250 16.303 multiply_cannon_metrocomm3 30084 15.6 0.047 0.050 6.680 15.372 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.958 14.960 sum_up_and_integrate 128 10.3 0.140 0.154 14.897 14.919 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.872 14.912 calculate_rho_elec 128 8.7 0.173 0.189 14.872 14.911 integrate_v_rspace 128 11.3 0.004 0.004 14.757 14.783 dbcsr_complete_redistribute 395 12.7 1.496 1.619 9.056 12.851 multiply_cannon_sync_h2d 30084 15.6 11.730 12.725 11.730 12.725 make_images_data 5014 15.6 0.061 0.065 11.026 12.540 dbcsr_mm_accdrv_process 62264 16.2 7.399 8.500 11.644 12.275 init_scf_run 11 5.9 0.000 0.001 11.782 11.783 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.782 11.783 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.755 11.770 cp_fm_diag_elpa_base 83 14.4 10.780 11.107 11.747 11.761 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.648 11.403 hybrid_alltoall_any 5200 16.5 0.525 2.212 9.867 11.288 transfer_fm_to_dbcsr 11 9.9 0.001 0.007 6.377 10.047 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.494 9.689 pw_transfer 1547 11.6 0.086 0.100 9.373 9.447 mp_alltoall_i22 716 14.1 5.557 9.355 5.557 9.355 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.147 9.228 density_rs2pw 128 9.7 0.006 0.006 7.164 8.514 fft_wrap_pw1pw2_140 523 13.2 1.564 1.595 8.126 8.218 grid_integrate_task_list 128 12.3 7.526 7.980 7.526 7.980 cp_fm_cholesky_decompose 22 10.9 7.740 7.832 7.740 7.832 wfi_extrapolate 11 7.9 0.001 0.001 7.804 7.805 multiply_cannon_metrocomm4 25070 15.6 0.079 0.091 2.777 7.342 mp_irecv_dv 76098 16.2 2.629 7.066 2.629 7.066 calculate_dm_sparse 128 9.5 0.001 0.001 6.353 6.443 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.195 6.247 rs_pw_transfer 1046 11.9 0.013 0.014 4.798 6.226 fft3d_ps 1291 14.7 2.846 2.906 6.064 6.108 mp_alltoall_d11v 2415 14.1 5.077 5.934 5.077 5.934 grid_collocate_task_list 128 9.7 5.155 5.860 5.155 5.860 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.109 5.183 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.691 4.759 potential_pw2rs 128 12.3 0.024 0.024 4.648 4.665 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.425 4.426 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.239 4.298 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=192.090000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1094.454545, yerr=22.157960 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.539809E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+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 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11085 1167. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.030 177.529 177.530 qs_mol_dyn_low 1 2.0 0.003 0.003 177.136 177.149 qs_forces 11 3.9 0.004 0.004 177.027 177.031 qs_energies 11 4.9 0.002 0.003 169.401 169.411 scf_env_do_scf 11 5.9 0.001 0.002 150.361 150.365 velocity_verlet 10 3.0 0.001 0.001 115.287 115.292 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 114.004 114.006 dbcsr_multiply_generic 2485 12.5 0.180 0.192 75.967 76.508 qs_scf_new_mos 116 7.6 0.001 0.001 75.392 75.476 qs_scf_loop_do_ot 116 8.6 0.001 0.001 75.392 75.476 ot_scf_mini 116 9.6 0.003 0.004 70.927 71.035 multiply_cannon 2485 13.5 0.573 0.606 56.058 60.137 multiply_cannon_loop 2485 14.5 0.443 0.456 51.404 52.454 ot_mini 116 10.6 0.001 0.001 39.970 40.059 init_scf_loop 11 6.9 0.000 0.000 36.208 36.209 mp_waitall_1 124680 16.7 26.283 33.631 26.283 33.631 prepare_preconditioner 11 7.9 0.000 0.000 32.192 32.224 make_preconditioner 11 8.9 0.000 0.000 32.192 32.224 make_full_inverse_cholesky 11 9.9 0.017 0.028 30.088 30.352 rebuild_ks_matrix 127 8.3 0.001 0.001 30.002 30.095 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.001 30.094 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.153 27.234 multiply_cannon_multrec 9940 15.5 10.404 15.755 18.557 22.353 qs_ot_get_derivative 116 11.6 0.002 0.002 19.952 20.052 ot_diis_step 116 11.6 0.019 0.020 19.947 19.947 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.588 19.932 apply_single 127 13.6 0.001 0.001 19.588 19.932 multiply_cannon_metrocomm3 9940 15.5 0.023 0.024 12.731 19.793 cp_fm_cholesky_invert 11 10.9 18.468 18.474 18.468 18.474 make_m2s 4970 13.5 0.065 0.071 16.166 18.439 qs_ot_get_p 127 10.4 0.001 0.001 18.036 18.155 make_images 4970 14.5 2.280 2.762 15.860 18.135 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.454 15.500 calculate_rho_elec 127 8.7 0.254 0.265 15.453 15.500 sum_up_and_integrate 127 10.3 0.180 0.190 15.158 15.205 integrate_v_rspace 127 11.3 0.004 0.004 14.978 15.033 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.250 14.264 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.127 13.129 make_images_data 4970 15.5 0.049 0.059 9.843 12.486 hybrid_alltoall_any 5155 16.4 0.836 3.776 9.724 12.218 init_scf_run 11 5.9 0.000 0.001 12.205 12.205 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.204 12.205 multiply_cannon_sync_h2d 9940 15.5 11.434 12.052 11.434 12.052 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.100 10.112 cp_fm_diag_elpa_base 82 14.4 9.865 9.940 10.097 10.108 pw_transfer 1535 11.6 0.085 0.094 9.587 9.607 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.364 9.386 dbcsr_mm_accdrv_process 20590 16.1 2.692 3.489 7.786 8.462 fft_wrap_pw1pw2_140 519 13.2 1.913 1.958 8.274 8.303 cp_fm_cholesky_decompose 22 10.9 8.035 8.148 8.035 8.148 grid_integrate_task_list 127 12.3 7.742 8.084 7.742 8.084 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.920 7.993 density_rs2pw 127 9.7 0.006 0.006 6.943 7.852 wfi_extrapolate 11 7.9 0.001 0.001 7.740 7.740 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.401 7.607 mp_allgather_i34 2485 14.5 2.860 7.084 2.860 7.084 calculate_dm_sparse 127 9.5 0.001 0.001 6.612 6.689 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.251 6.271 grid_collocate_task_list 127 9.7 5.424 6.090 5.424 6.090 mp_alltoall_d11v 2401 14.1 5.244 5.905 5.244 5.905 fft3d_ps 1281 14.7 2.749 2.819 5.847 5.884 dbcsr_complete_redistribute 393 12.7 2.107 2.154 5.236 5.634 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.290 5.291 multiply_cannon_metrocomm4 7455 15.5 0.023 0.026 1.865 5.277 mp_irecv_dv 28618 15.9 1.830 5.194 1.830 5.194 rs_pw_transfer 1038 11.9 0.012 0.013 4.248 5.176 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.112 5.150 potential_pw2rs 127 12.3 0.027 0.027 4.552 4.570 calculate_first_density_matrix 1 7.0 0.000 0.000 4.241 4.245 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.561 3.825 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.744 3.793 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.422 3.759 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.746 3.757 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.599 3.681 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.551 3.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=177.530000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1425.181818, yerr=63.262538 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.135640E+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 11.342272E+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 716108613552 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 59820. MP_Allreduce 10965 1506. MP_Sync 86 MP_Alltoall 1700 36954374. 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 0.021 0.039 301.583 301.584 qs_mol_dyn_low 1 2.0 0.003 0.003 300.634 301.049 qs_forces 11 3.9 0.005 0.005 300.485 300.489 qs_energies 11 4.9 0.019 0.058 291.045 291.055 scf_env_do_scf 11 5.9 0.002 0.015 267.149 267.163 velocity_verlet 10 3.0 0.001 0.001 215.681 215.689 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 141.011 141.013 init_scf_loop 11 6.9 0.000 0.000 125.865 125.866 prepare_preconditioner 11 7.9 0.000 0.000 120.792 120.813 make_preconditioner 11 8.9 0.000 0.000 120.792 120.813 make_full_inverse_cholesky 11 9.9 0.036 0.039 96.716 117.943 qs_scf_new_mos 116 7.6 0.001 0.001 90.458 90.545 qs_scf_loop_do_ot 116 8.6 0.001 0.001 90.457 90.544 ot_scf_mini 116 9.6 0.004 0.004 85.646 85.688 dbcsr_multiply_generic 2485 12.5 0.211 0.222 83.769 84.367 cp_fm_upper_to_full 104 14.8 53.758 76.838 53.758 76.838 multiply_cannon 2485 13.5 0.695 0.735 58.905 59.472 multiply_cannon_loop 2485 14.5 0.473 0.479 55.315 56.756 ot_mini 116 10.6 0.001 0.001 44.063 44.096 dbcsr_complete_redistribute 393 12.7 3.975 4.042 29.855 43.312 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.413 39.853 rebuild_ks_matrix 127 8.3 0.001 0.001 38.287 38.339 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.017 38.286 38.339 transfer_fm_to_dbcsr 11 9.9 0.030 0.034 24.032 37.289 mp_alltoall_i22 712 14.1 21.858 35.491 21.858 35.491 qs_ks_update_qs_env 127 7.6 0.001 0.001 35.197 35.243 cp_fm_cholesky_invert 11 10.9 33.762 33.769 33.762 33.769 mp_waitall_1 102768 16.8 28.327 31.837 28.327 31.837 qs_ot_get_p 127 10.4 0.001 0.001 26.066 26.122 qs_ot_get_derivative 116 11.6 0.002 0.002 24.111 24.154 qs_ot_p2m_diag 82 11.4 0.868 0.872 21.973 22.003 make_m2s 4970 13.5 0.074 0.077 20.208 21.056 make_images 4970 14.5 3.755 3.884 19.730 20.577 qs_rho_update_rho_low 127 7.7 0.001 0.001 20.480 20.499 calculate_rho_elec 127 8.7 0.476 0.477 20.479 20.498 cp_dbcsr_syevd 82 12.4 0.006 0.006 20.262 20.265 ot_diis_step 116 11.6 0.022 0.022 19.916 19.916 multiply_cannon_metrocomm3 9940 15.5 0.023 0.025 18.540 19.792 sum_up_and_integrate 127 10.3 0.318 0.320 19.643 19.714 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.182 19.407 apply_single 127 13.6 0.001 0.001 19.182 19.407 integrate_v_rspace 127 11.3 0.004 0.004 19.324 19.394 multiply_cannon_multrec 9940 15.5 10.496 12.202 17.948 18.009 cp_fm_diag_elpa 82 13.4 0.000 0.000 17.038 17.039 cp_fm_diag_elpa_base 82 14.4 12.632 14.301 17.034 17.034 multiply_cannon_sync_h2d 9940 15.5 15.528 15.543 15.528 15.543 init_scf_run 11 5.9 0.000 0.001 13.340 13.340 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.340 13.340 hybrid_alltoall_any 5155 16.4 1.288 3.012 11.053 12.766 make_images_data 4970 15.5 0.057 0.061 10.943 12.652 pw_transfer 1535 11.6 0.092 0.093 11.422 11.430 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 11.188 11.194 fft_wrap_pw1pw2_140 519 13.2 3.053 3.097 9.938 9.946 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 9.365 9.401 wfi_extrapolate 11 7.9 0.001 0.001 9.333 9.333 mp_alltoall_d11v 2401 14.1 8.403 9.230 8.403 9.230 cp_fm_cholesky_decompose 22 10.9 9.026 9.072 9.026 9.072 dbcsr_mm_accdrv_process 20590 16.0 3.791 5.605 7.216 9.057 grid_integrate_task_list 127 12.3 8.536 8.714 8.536 8.714 qs_energies_init_hamiltonians 11 5.9 0.052 0.060 8.407 8.408 density_rs2pw 127 9.7 0.005 0.005 8.002 8.171 calculate_dm_sparse 127 9.5 0.001 0.001 7.303 7.365 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.551 6.615 grid_collocate_task_list 127 9.7 6.392 6.453 6.392 6.453 fft3d_ps 1281 14.7 2.824 2.840 6.297 6.325 rs_scatter_matrices 138 9.7 3.734 4.617 6.064 6.294 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.147 6.246 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=301.584000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2839.272727, yerr=158.005806 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 1.259463E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 36123232. 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 0.037 85.288 85.289 qs_energies 1 2.0 0.000 0.000 84.774 84.786 ls_scf 1 3.0 0.000 0.000 83.894 83.904 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.833 73.004 multiply_cannon 111 7.7 0.018 0.020 56.123 57.288 multiply_cannon_loop 111 8.7 0.205 0.218 52.631 53.836 ls_scf_main 1 4.0 0.000 0.000 52.409 52.420 density_matrix_trs4 2 5.0 0.002 0.003 46.839 46.963 ls_scf_init_scf 1 4.0 0.001 0.003 28.422 28.423 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.276 27.337 mp_waitall_1 11031 10.9 22.493 25.443 22.493 25.443 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.124 25.149 multiply_cannon_multrec 2664 9.7 8.186 9.012 15.609 17.524 multiply_cannon_sync_h2d 2664 9.7 13.705 15.507 13.705 15.507 make_m2s 222 7.7 0.008 0.010 13.166 13.692 make_images 222 8.7 0.098 0.108 13.145 13.673 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.593 12.213 make_images_data 222 9.7 0.004 0.005 7.783 8.461 dbcsr_mm_accdrv_process 4760 10.4 0.511 0.638 7.043 8.065 hybrid_alltoall_any 227 10.6 0.216 1.840 6.671 7.861 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.449 7.653 dbcsr_mm_accdrv_process_sort 4760 11.4 6.331 7.257 6.331 7.257 calculate_norms 4752 9.8 5.525 6.202 5.525 6.202 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.031 5.188 mp_sum_l 887 5.1 3.021 4.310 3.021 4.310 make_images_sizes 222 9.7 0.000 0.000 0.800 3.662 mp_alltoall_i44 222 10.7 0.800 3.662 0.800 3.662 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.057 3.340 mp_irecv_dv 6231 10.9 2.041 3.312 2.041 3.312 arnoldi_extremal 4 6.8 0.000 0.000 3.222 3.252 arnoldi_normal_ev 4 7.8 0.001 0.002 3.222 3.252 build_subspace 16 8.4 0.009 0.012 3.124 3.126 ls_scf_post 1 4.0 0.000 0.000 3.063 3.074 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.217 3.042 ls_scf_store_result 1 5.0 0.000 0.000 2.881 2.922 dbcsr_special_finalize 555 9.7 0.005 0.006 2.234 2.743 dbcsr_merge_single_wm 555 10.7 0.466 0.635 2.226 2.735 make_images_pack 222 9.7 2.205 2.626 2.206 2.627 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.325 2.563 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.353 2.459 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.457 2.067 2.459 dbcsr_sort_data 658 11.4 2.018 2.433 2.018 2.433 buffer_matrices_ensure_size 222 8.7 1.789 2.245 1.789 2.245 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.791 1.793 rebuild_ks_matrix 3 7.3 0.000 0.000 1.782 1.783 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.782 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.289000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 2.104824E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.078 94.411 94.412 qs_energies 1 2.0 0.000 0.001 93.858 93.871 ls_scf 1 3.0 0.062 0.358 92.464 92.479 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.640 77.959 multiply_cannon 111 7.7 0.029 0.043 54.690 59.356 ls_scf_main 1 4.0 0.008 0.038 55.751 55.788 multiply_cannon_loop 111 8.7 0.116 0.123 51.396 55.050 density_matrix_trs4 2 5.0 0.002 0.003 49.857 50.047 ls_scf_init_scf 1 4.0 0.001 0.002 32.855 32.948 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.398 31.504 mp_waitall_1 9105 10.9 22.118 30.898 22.118 30.898 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.625 27.634 multiply_cannon_multrec 1332 9.7 13.101 17.027 22.914 27.498 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 12.261 22.669 make_m2s 222 7.7 0.006 0.008 15.559 16.159 make_images 222 8.7 1.575 1.979 15.529 16.131 dbcsr_mm_accdrv_process 4041 10.4 0.416 1.729 9.424 11.582 dbcsr_mm_accdrv_process_sort 4041 11.4 8.531 10.020 8.531 10.020 make_images_data 222 9.7 0.004 0.004 8.998 9.821 hybrid_alltoall_any 227 10.6 0.523 2.501 8.370 9.486 mp_sum_l 887 5.1 6.100 9.262 6.100 9.262 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.231 7.722 mp_irecv_dv 3311 11.0 3.211 7.661 3.211 7.661 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.467 7.107 apply_matrix_preconditioner 6 5.3 0.000 0.000 6.689 6.952 calculate_norms 2376 9.8 6.007 6.770 6.007 6.770 multiply_cannon_sync_h2d 1332 9.7 4.774 5.728 4.774 5.728 arnoldi_extremal 4 6.8 0.000 0.000 4.929 4.946 arnoldi_normal_ev 4 7.8 0.001 0.005 4.929 4.945 build_subspace 16 8.4 0.014 0.021 4.674 4.677 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.478 4.224 ls_scf_post 1 4.0 0.002 0.016 3.796 3.855 ls_scf_store_result 1 5.0 0.000 0.000 3.545 3.664 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.280 3.502 dbcsr_matrix_vector_mult_local 304 10.0 2.762 3.243 2.764 3.245 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.662 2.750 mp_allgather_i34 111 8.7 0.988 2.601 0.988 2.601 make_images_pack 222 9.7 2.040 2.456 2.042 2.458 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.076 2.170 rebuild_ks_matrix 3 7.3 0.000 0.000 2.063 2.157 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.021 2.063 2.157 dbcsr_sort_data 436 11.2 1.825 2.089 1.825 2.089 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.412000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1760.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 2.769846E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.050 0.087 95.298 95.301 qs_energies 1 2.0 0.000 0.000 94.533 94.539 ls_scf 1 3.0 0.000 0.001 93.131 93.138 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.284 77.545 multiply_cannon 111 7.7 0.041 0.107 53.382 57.838 ls_scf_main 1 4.0 0.000 0.001 57.817 57.822 multiply_cannon_loop 111 8.7 0.100 0.106 49.792 52.984 density_matrix_trs4 2 5.0 0.030 0.264 51.705 51.909 mp_waitall_1 7281 11.0 23.954 33.761 23.954 33.761 ls_scf_init_scf 1 4.0 0.001 0.004 31.471 31.472 ls_scf_init_matrix_S 1 5.0 0.000 0.000 30.221 30.270 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.835 27.848 multiply_cannon_multrec 888 9.7 12.638 15.223 21.367 24.348 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.052 22.526 make_m2s 222 7.7 0.006 0.007 17.253 18.461 make_images 222 8.7 1.974 2.300 17.215 18.423 make_images_data 222 9.7 0.003 0.004 9.950 11.028 hybrid_alltoall_any 227 10.6 0.619 2.865 9.572 10.876 dbcsr_mm_accdrv_process 3754 10.4 0.303 0.693 8.244 9.693 dbcsr_mm_accdrv_process_sort 3754 11.4 7.805 9.000 7.805 9.000 mp_sum_l 887 5.1 5.099 8.340 5.099 8.340 multiply_cannon_sync_h2d 888 9.7 5.993 7.201 5.993 7.201 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.450 7.000 mp_irecv_dv 2335 11.1 2.435 6.949 2.435 6.949 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.793 6.473 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.571 6.147 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.002 5.242 arnoldi_extremal 4 6.8 0.000 0.000 5.200 5.219 arnoldi_normal_ev 4 7.8 0.001 0.004 5.200 5.218 build_subspace 16 8.4 0.014 0.020 4.886 4.895 calculate_norms 1584 9.8 4.334 4.773 4.334 4.773 mp_allgather_i34 111 8.7 1.413 3.869 1.413 3.869 ls_scf_post 1 4.0 0.000 0.001 3.844 3.851 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.500 3.845 dbcsr_matrix_vector_mult_local 304 10.0 3.039 3.613 3.041 3.616 ls_scf_store_result 1 5.0 0.000 0.000 3.293 3.387 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.883 2.978 dbcsr_sort_data 325 11.1 1.923 2.203 1.923 2.203 make_images_pack 222 9.7 1.810 2.116 1.813 2.119 make_images_sizes 222 9.7 0.000 0.000 0.946 2.048 mp_alltoall_i44 222 10.7 0.946 2.048 0.946 2.048 dbcsr_data_release 9322 10.9 1.322 1.983 1.322 1.983 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.969 1.970 rebuild_ks_matrix 3 7.3 0.000 0.000 1.950 1.951 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.950 1.951 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.301000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2163.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 3.391582E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.096 0.116 97.625 97.626 qs_energies 1 2.0 0.000 0.000 96.902 96.916 ls_scf 1 3.0 0.015 0.060 95.254 95.267 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.539 78.778 ls_scf_main 1 4.0 0.001 0.019 58.791 58.807 multiply_cannon 111 7.7 0.058 0.156 51.614 56.098 density_matrix_trs4 2 5.0 0.003 0.019 52.542 52.668 multiply_cannon_loop 111 8.7 0.114 0.125 46.522 49.020 ls_scf_init_scf 1 4.0 0.010 0.019 33.179 33.181 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.859 31.926 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.248 29.270 mp_waitall_1 6369 11.0 22.503 29.166 22.503 29.166 multiply_cannon_multrec 1332 9.7 14.191 17.290 22.150 24.540 make_m2s 222 7.7 0.007 0.008 21.152 22.628 make_images 222 8.7 3.143 3.603 21.102 22.581 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.017 16.836 make_images_data 222 9.7 0.004 0.004 11.784 13.435 hybrid_alltoall_any 227 10.6 0.797 3.761 11.119 12.879 dbcsr_mm_accdrv_process 3641 10.4 0.237 0.596 7.599 9.114 dbcsr_mm_accdrv_process_sort 3641 11.4 7.216 8.694 7.216 8.694 mp_sum_l 887 5.1 4.195 7.689 4.195 7.689 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.291 6.191 multiply_cannon_sync_h2d 1332 9.7 5.498 6.071 5.498 6.071 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.067 5.948 mp_irecv_dv 3229 10.9 2.043 5.871 2.043 5.871 arnoldi_extremal 4 6.8 0.000 0.000 5.306 5.332 arnoldi_normal_ev 4 7.8 0.001 0.005 5.306 5.332 build_subspace 16 8.4 0.014 0.021 4.933 4.940 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.602 4.813 mp_allgather_i34 111 8.7 2.214 4.708 2.214 4.708 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.530 4.672 calculate_norms 2376 9.8 4.190 4.487 4.190 4.487 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.618 3.920 dbcsr_matrix_vector_mult_local 304 10.0 3.220 3.695 3.222 3.697 dbcsr_sort_data 658 11.4 3.073 3.449 3.073 3.449 ls_scf_post 1 4.0 0.004 0.016 3.269 3.284 dbcsr_special_finalize 555 9.7 0.006 0.007 2.819 3.228 dbcsr_merge_single_wm 555 10.7 0.539 0.662 2.810 3.220 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.055 3.116 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.051 dbcsr_data_release 10477 10.7 1.588 2.374 1.588 2.374 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.016 2.018 rebuild_ks_matrix 3 7.3 0.000 0.000 1.993 1.995 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.016 1.993 1.994 dbcsr_finalize 304 7.8 0.050 0.061 1.808 1.968 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.626000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2694.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 4.690305E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.055 100.506 100.507 qs_energies 1 2.0 0.000 0.000 99.741 99.752 ls_scf 1 3.0 0.000 0.000 97.680 97.689 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.540 78.706 ls_scf_main 1 4.0 0.000 0.000 62.409 62.410 multiply_cannon 111 7.7 0.100 0.164 55.658 60.976 density_matrix_trs4 2 5.0 0.002 0.003 55.132 55.213 multiply_cannon_loop 111 8.7 0.069 0.079 51.094 52.549 mp_waitall_1 5436 11.0 26.765 32.489 26.765 32.489 ls_scf_init_scf 1 4.0 0.003 0.005 31.662 31.668 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.977 30.023 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.814 27.821 multiply_cannon_multrec 444 9.7 14.039 16.873 21.173 24.678 make_m2s 222 7.7 0.004 0.005 18.003 20.503 make_images 222 8.7 3.720 4.408 17.941 20.444 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.208 15.963 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.120 14.719 make_images_data 222 9.7 0.003 0.004 10.235 12.663 hybrid_alltoall_any 227 10.6 0.788 3.771 9.891 12.543 multiply_cannon_sync_h2d 444 9.7 6.505 8.929 6.505 8.929 dbcsr_mm_accdrv_process 3003 10.4 0.181 0.430 6.842 7.976 dbcsr_mm_accdrv_process_sort 3003 11.4 6.522 7.632 6.522 7.632 mp_allgather_i34 111 8.7 2.792 7.058 2.792 7.058 arnoldi_extremal 4 6.8 0.000 0.000 5.950 5.959 arnoldi_normal_ev 4 7.8 0.001 0.004 5.950 5.959 build_subspace 16 8.4 0.015 0.020 5.558 5.566 mp_sum_l 887 5.1 2.922 5.439 2.922 5.439 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.613 4.790 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.248 4.471 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.635 4.311 mp_irecv_dv 1241 11.2 1.617 4.290 1.617 4.290 dbcsr_matrix_vector_mult_local 304 10.0 3.741 4.200 3.743 4.203 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.049 4.164 make_images_sizes 222 9.7 0.000 0.000 1.120 3.763 mp_alltoall_i44 222 10.7 1.120 3.763 1.120 3.763 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.625 3.734 calculate_norms 792 9.8 3.542 3.661 3.542 3.661 ls_scf_post 1 4.0 0.000 0.000 3.608 3.619 ls_scf_store_result 1 5.0 0.000 0.000 3.378 3.421 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.485 2.487 rebuild_ks_matrix 3 7.3 0.000 0.000 2.452 2.453 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.452 2.453 dbcsr_finalize 304 7.8 0.062 0.078 2.200 2.284 dbcsr_merge_all 275 8.9 0.473 0.528 2.049 2.123 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.046 2.046 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.507000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3725.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_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 8.754156E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.094 0.112 111.137 111.137 qs_energies 1 2.0 0.000 0.000 109.134 109.157 ls_scf 1 3.0 0.000 0.000 106.171 106.192 dbcsr_multiply_generic 111 6.7 0.024 0.027 79.463 79.599 ls_scf_main 1 4.0 0.000 0.000 66.617 66.618 density_matrix_trs4 2 5.0 0.002 0.003 57.312 57.382 multiply_cannon 111 7.7 0.507 0.676 51.252 53.059 multiply_cannon_loop 111 8.7 0.068 0.071 47.484 48.759 ls_scf_init_scf 1 4.0 0.001 0.001 35.824 35.825 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.248 34.262 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.400 31.412 mp_waitall_1 4527 11.1 22.911 26.651 22.911 26.651 make_m2s 222 7.7 0.005 0.005 24.124 25.043 make_images 222 8.7 4.588 4.964 24.018 24.936 multiply_cannon_multrec 444 9.7 17.857 18.520 22.535 23.162 hybrid_alltoall_any 227 10.6 1.663 3.654 13.040 15.980 make_images_data 222 9.7 0.003 0.003 13.335 15.741 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.987 12.064 multiply_cannon_sync_h2d 444 9.7 8.843 8.886 8.843 8.886 arnoldi_extremal 4 6.8 0.000 0.000 7.616 7.628 arnoldi_normal_ev 4 7.8 0.003 0.009 7.616 7.628 build_subspace 16 8.4 0.026 0.035 7.060 7.069 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.625 5.782 dbcsr_matrix_vector_mult_local 304 10.0 5.112 5.431 5.115 5.434 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.097 5.341 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.202 5.294 dbcsr_mm_accdrv_process 1814 10.4 0.281 0.325 4.505 4.649 dbcsr_mm_accdrv_process_sort 1814 11.4 4.201 4.329 4.201 4.329 ls_scf_post 1 4.0 0.000 0.000 3.730 3.749 make_images_sizes 222 9.7 0.000 0.000 1.460 3.685 mp_alltoall_i44 222 10.7 1.460 3.684 1.460 3.684 ls_scf_store_result 1 5.0 0.000 0.000 3.460 3.472 mp_allgather_i34 111 8.7 1.050 3.347 1.050 3.347 calculate_norms 792 9.8 3.244 3.278 3.244 3.278 dbcsr_finalize 304 7.8 0.082 0.089 3.090 3.165 dbcsr_merge_all 275 8.9 0.890 0.917 2.876 2.946 qs_energies_init_hamiltonians 1 3.0 0.002 0.005 2.932 2.934 dbcsr_complete_redistribute 5 7.6 1.433 1.474 2.773 2.881 dbcsr_data_release 12724 10.6 2.329 2.864 2.329 2.864 matrix_ls_to_qs 2 6.0 0.000 0.000 2.413 2.537 dbcsr_sort_data 325 11.1 2.446 2.504 2.446 2.504 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.482 2.484 mp_sum_l 887 5.1 1.747 2.479 1.747 2.479 rebuild_ks_matrix 3 7.3 0.000 0.000 2.416 2.418 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 2.416 2.418 dbcsr_new_transposed 4 7.5 0.243 0.252 2.301 2.315 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.185 2.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=111.137000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6840.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/fe1068d0649e3e881f064d86a0513eff1752c2c8_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.180352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 76559. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 840876907. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.118 266.591 266.593 qs_mol_dyn_low 1 2.0 0.004 0.045 265.030 265.314 qs_forces 5 3.8 0.008 0.119 264.822 264.824 qs_energies 5 4.8 0.005 0.064 261.481 261.508 scf_env_do_scf 5 5.8 0.002 0.075 238.479 238.482 scf_env_do_scf_inner_loop 105 6.6 0.004 0.075 205.796 205.813 qs_scf_new_mos 105 7.6 0.000 0.001 157.362 157.513 qs_scf_loop_do_ot 105 8.6 0.001 0.001 157.361 157.513 ot_scf_mini 105 9.6 0.003 0.017 147.343 147.458 dbcsr_multiply_generic 1445 12.2 0.128 0.136 141.798 142.192 multiply_cannon 1445 13.2 0.273 0.283 121.206 123.516 multiply_cannon_loop 1445 14.2 2.370 2.504 118.918 120.268 velocity_verlet 4 3.0 0.002 0.054 115.554 115.555 ot_mini 105 10.6 0.001 0.015 60.853 60.986 qs_ot_get_p 112 10.4 0.001 0.001 56.255 56.509 multiply_cannon_multrec 69360 15.2 31.702 37.226 39.466 44.835 qs_ot_p2m_diag 40 11.0 0.020 0.031 44.315 44.408 mp_waitall_1 488190 16.1 36.948 43.405 36.948 43.405 cp_dbcsr_syevd 40 12.0 0.002 0.002 40.974 40.976 qs_ot_get_derivative 55 11.6 0.001 0.001 39.194 39.325 cp_fm_syevd 40 13.0 0.000 0.000 35.302 35.452 multiply_cannon_sync_h2d 69360 15.2 30.595 34.152 30.595 34.152 rebuild_ks_matrix 110 8.4 0.000 0.000 32.631 32.836 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.076 32.631 32.835 init_scf_loop 7 6.6 0.001 0.043 32.635 32.642 multiply_cannon_metrocomm3 69360 15.2 0.195 0.207 25.276 31.955 qs_ks_update_qs_env 112 7.6 0.001 0.001 30.129 30.335 cp_fm_redistribute_end 40 14.0 15.099 30.110 15.106 30.113 cp_fm_syevd_base 40 14.0 14.994 30.013 14.994 30.013 prepare_preconditioner 7 7.6 0.000 0.008 26.796 26.831 make_preconditioner 7 8.6 0.001 0.040 26.796 26.830 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 22.965 23.164 apply_single 62 13.6 0.000 0.000 22.965 23.164 qs_rho_update_rho_low 110 7.6 0.001 0.016 22.709 23.007 calculate_rho_elec 110 8.6 0.030 0.034 22.708 23.006 ot_new_cg_direction 55 11.6 0.001 0.031 20.956 20.963 init_scf_run 5 5.8 0.000 0.001 20.121 20.122 scf_env_initial_rho_setup 5 6.8 0.001 0.027 20.121 20.122 rs_pw_transfer 690 11.5 0.011 0.028 18.561 19.870 make_full_inverse_cholesky 7 9.6 0.000 0.000 18.715 18.783 density_rs2pw 110 9.6 0.006 0.007 16.440 17.619 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.966 15.121 calculate_first_density_matrix 1 7.0 0.001 0.021 14.694 14.710 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 14.681 14.696 acc_transpose_blocks 69360 15.2 0.349 0.374 13.692 14.303 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.049 14.162 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 14.097 14.145 pw_transfer 1645 12.4 0.083 0.106 13.182 13.504 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 13.040 13.367 mp_sum_l 4764 12.2 12.152 12.910 12.152 12.910 cp_fm_cholesky_invert 7 10.6 12.670 12.684 12.670 12.684 acc_transpose_blocks_kernels 69360 16.2 0.840 0.898 11.805 12.265 calculate_dm_sparse 110 9.5 0.000 0.001 11.662 11.824 sum_up_and_integrate 60 10.3 0.028 0.046 11.545 11.562 integrate_v_rspace 60 11.3 0.002 0.018 11.517 11.535 jit_kernel_transpose 5 15.0 10.965 11.385 10.965 11.385 fft_wrap_pw1pw2_240 915 15.0 1.198 1.281 10.987 11.268 qs_vxc_create 110 10.4 0.002 0.005 10.888 10.937 check_diag 80 13.5 8.586 8.857 10.298 10.445 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.357 10.440 fft3d_pb 915 16.0 2.374 2.663 9.049 9.423 dbcsr_mm_accdrv_process 154766 15.8 3.992 4.208 7.633 8.583 make_m2s 2890 13.2 0.079 0.087 8.012 8.554 make_images 2890 14.2 0.239 0.258 7.904 8.444 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 7.582 7.839 make_full_single_inverse 7 9.6 0.001 0.002 7.749 7.785 potential_pw2rs 60 12.3 0.003 0.003 7.452 7.538 multiply_cannon_metrocomm1 69360 15.2 0.089 0.096 4.663 7.522 xc_vxc_pw_create 60 11.3 0.039 0.050 7.322 7.371 rs_pw_transfer_RS2PW_30 110 11.6 1.592 1.686 6.291 6.990 xc_pw_derive 510 13.4 0.006 0.007 6.601 6.679 mp_alltoall_z22v 2340 17.7 6.165 6.511 6.165 6.511 mp_sendrecv_dv 168740 12.6 6.009 6.186 6.009 6.186 mp_waitany 7680 13.5 4.710 5.905 4.710 5.905 cp_fm_cholesky_decompose 14 10.2 5.500 5.508 5.500 5.508 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=266.593000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: fe1068d0649e3e881f064d86a0513eff1752c2c8 Summary: empty Status: OK