=== 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: 7e52d8fe64174c63d4297b1f87b53e7f026adc7a ################# 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 2023.05.001, # HDF5 1.12.0, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, 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. # # Last update: 13.09.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.12.0 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/01 job id: 48939077 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/02 job id: 48939078 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/03 job id: 48939079 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/04 job id: 48939080 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/05 job id: 48939082 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/06 job id: 48939084 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/07 job id: 48939085 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/08 job id: 48939086 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/09 job id: 48939087 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/10 job id: 48939088 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/11 job id: 48939089 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/12 job id: 48939090 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/13 job id: 48939091 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/14 job id: 48939092 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/15 job id: 48939093 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/16 job id: 48939094 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/17 job id: 48939095 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/18 job id: 48939097 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/19 job id: 48939098 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/20 job id: 48939099 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/21 job id: 48939100 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/22 job id: 48939101 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/23 job id: 48939102 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/24 job id: 48939103 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/25 job id: 48939104 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/26 job id: 48939105 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/27 job id: 48939106 --- 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/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 0.032 0.042 141.939 141.940 farming_run 1 2.0 138.122 138.135 141.879 141.883 create_force_env 2 3.0 3.720 3.720 3.720 3.720 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.500455E+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 1134128. 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 0.007 0.026 116.384 116.385 qs_energies 1 2.0 0.000 0.000 116.073 116.083 mp2_main 1 3.0 0.000 0.000 113.393 113.403 mp2_gpw_main 1 4.0 0.039 0.057 112.027 112.037 mp2_ri_gpw_compute_in 1 5.0 0.178 0.196 92.430 92.611 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.591 54.762 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.164 40.962 45.815 get_2c_integrals 1 6.0 0.008 0.009 36.972 37.658 integrate_v_rspace 273 8.0 0.441 0.456 24.558 29.109 pw_transfer 6555 10.6 0.374 0.386 26.750 27.071 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 25.385 25.712 grid_integrate_task_list 273 9.0 20.476 25.446 20.476 25.446 fft_wrap_pw1pw2_100 2178 12.4 1.096 1.156 22.924 23.297 rpa_ri_compute_en 1 5.0 0.020 0.022 19.468 19.602 compute_2c_integrals 1 7.0 0.011 0.013 19.022 19.039 cp_fm_cholesky_decompose 12 8.2 18.325 19.035 18.325 19.035 compute_2c_integrals_loop_lm 1 8.0 0.007 0.009 18.549 18.606 mp2_eri_2c_integrate_gpw 1 9.0 2.377 2.413 18.542 18.598 cholesky_decomp 1 7.0 0.000 0.000 16.790 17.470 fft3d_s 5443 13.4 16.075 16.319 16.098 16.342 ao_to_mo_and_store_B_mult_1 272 7.0 10.784 15.328 10.784 15.328 calculate_wavefunction 272 8.0 5.400 5.486 12.317 13.037 rpa_num_int 1 6.0 0.001 0.011 11.125 11.133 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 11.075 11.108 calc_mat_Q 8 8.0 0.000 0.000 9.525 9.603 contract_S_to_Q 8 9.0 0.000 0.000 8.946 9.027 calc_potential_gpw 544 9.5 0.005 0.005 8.251 8.673 parallel_gemm_fm 14 9.1 0.000 0.000 8.544 8.643 parallel_gemm_fm_cosma 14 10.1 8.544 8.643 8.544 8.643 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.055 8.256 potential_pw2rs 545 10.0 0.106 0.109 7.431 8.109 create_integ_mat 1 6.0 0.022 0.029 8.079 8.080 collocate_single_gaussian 272 10.0 0.039 0.042 7.302 7.494 array2fm 1 7.0 0.000 0.000 6.642 7.123 pw_scatter_s 2720 13.7 4.362 4.615 4.362 4.615 pw_gather_s 2722 13.2 3.441 3.965 3.441 3.965 array2fm_buffer_send 1 8.0 2.903 3.156 2.903 3.156 scf_env_do_scf 1 3.0 0.000 0.000 2.382 2.382 pw_poisson_solve 545 10.5 1.116 1.166 2.189 2.333 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.037321, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2819.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 0.024 0.037 423.836 423.837 farming_run 1 2.0 421.550 421.556 423.799 423.801 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827128576 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788821 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.240388E+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 757 478553. 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 0.011 0.030 216.805 216.805 qs_energies 1 2.0 0.000 0.000 216.581 216.593 scf_env_do_scf 1 3.0 0.000 0.000 114.877 114.878 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.959 113.966 rebuild_ks_matrix 4 6.0 0.000 0.000 113.958 113.965 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 113.958 113.965 hfx_ks_matrix 4 8.0 0.001 0.001 113.598 113.602 integrate_four_center 4 9.0 0.154 0.471 113.597 113.601 integrate_four_center_main 4 10.0 0.098 0.618 102.043 105.226 integrate_four_center_bin 266 11.0 101.945 104.830 101.945 104.830 mp2_main 1 3.0 0.000 0.000 101.403 101.416 mp2_gpw_main 1 4.0 0.047 0.072 100.461 100.477 init_scf_loop 1 4.0 0.000 0.000 97.022 97.022 mp2_ri_gpw_compute_in 1 5.0 0.068 0.101 73.857 74.872 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.766 54.780 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.153 41.506 46.263 integrate_v_rspace 95 8.0 0.396 0.557 27.981 32.611 pw_transfer 2240 10.6 0.146 0.170 29.444 29.838 ao_to_mo_and_store_B_mult_1 91 7.0 10.592 28.971 10.592 28.971 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.467 28.848 grid_integrate_task_list 95 9.0 23.383 28.209 23.383 28.209 mp2_ri_gpw_compute_en 1 5.0 0.063 0.076 26.437 28.001 fft_wrap_pw1pw2_100 730 12.4 1.262 1.464 26.227 26.652 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.846 1.897 24.773 24.781 get_2c_integrals 1 6.0 0.000 0.000 20.001 20.029 compute_2c_integrals 1 7.0 0.007 0.016 18.988 18.997 fft3d_s 1823 13.4 18.483 18.844 18.497 18.859 compute_2c_integrals_loop_lm 1 8.0 0.007 0.017 18.538 18.828 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.823 18.531 18.812 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 17.853 17.853 calculate_wavefunction 91 8.0 2.034 2.071 9.634 9.862 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.581 8.783 9.087 potential_pw2rs 186 10.0 0.033 0.035 8.405 9.087 local_gemm 172 8.0 8.225 8.539 8.225 8.539 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.102 8.485 calc_potential_gpw 182 9.5 0.002 0.002 7.883 8.106 mp2_ri_gpw_compute_en_comm 22 7.0 0.505 0.523 7.753 8.083 collocate_single_gaussian 91 10.0 0.017 0.021 7.716 7.934 mp_sync 37 10.5 3.497 7.194 3.497 7.194 integrate_four_center_load 4 10.0 0.000 0.000 6.773 6.778 hfx_load_balance 1 11.0 0.000 0.000 6.773 6.778 mp2_ri_gpw_compute_en_ener 172 7.0 6.346 6.416 6.346 6.416 mp_sendrecv_dm3 2068 8.0 5.765 6.109 5.765 6.109 pw_gather_s 912 13.2 4.476 5.029 4.476 5.029 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.463534, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 452.841472E+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 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 947862. 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 0.008 0.026 59.282 59.284 qs_mol_dyn_low 1 2.0 0.003 0.003 58.915 58.924 qs_forces 11 3.9 0.002 0.003 58.816 58.816 qs_energies 11 4.9 0.001 0.002 57.145 57.157 scf_env_do_scf 11 5.9 0.000 0.001 50.150 50.150 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 47.626 47.626 qs_scf_new_mos 108 7.5 0.000 0.001 35.623 35.940 qs_scf_loop_do_ot 108 8.5 0.000 0.001 35.622 35.939 dbcsr_multiply_generic 2286 12.5 0.107 0.112 34.977 35.514 ot_scf_mini 108 9.5 0.002 0.002 33.869 34.062 velocity_verlet 10 3.0 0.004 0.014 29.040 29.041 multiply_cannon 2286 13.5 0.189 0.197 26.749 28.304 multiply_cannon_loop 2286 14.5 1.824 1.948 25.668 27.392 ot_mini 108 10.5 0.001 0.001 19.972 20.240 qs_ot_get_derivative 108 11.5 0.001 0.001 16.852 17.049 mp_waitall_1 245248 16.5 9.084 14.777 9.084 14.777 multiply_cannon_metrocomm3 54864 15.5 0.073 0.080 5.911 13.144 multiply_cannon_multrec 54864 15.5 3.658 5.707 7.746 11.170 qs_ot_get_p 119 10.4 0.001 0.001 9.036 9.355 rebuild_ks_matrix 119 8.3 0.000 0.000 8.943 9.100 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.017 8.943 9.100 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.889 8.030 mp_sum_l 7287 12.8 5.290 7.216 5.290 7.216 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.893 6.364 multiply_cannon_sync_h2d 54864 15.5 5.093 6.316 5.093 6.316 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.836 5.908 dbcsr_mm_accdrv_process 76910 16.1 1.801 2.875 4.002 5.722 init_scf_run 11 5.9 0.000 0.001 5.587 5.588 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.587 5.587 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.429 5.540 qs_rho_update_rho_low 119 7.7 0.000 0.001 5.305 5.428 calculate_rho_elec 119 8.7 0.011 0.017 5.304 5.427 sum_up_and_integrate 119 10.3 0.001 0.002 5.320 5.330 integrate_v_rspace 119 11.3 0.002 0.002 5.309 5.320 cp_dbcsr_syevd 50 12.0 0.002 0.003 4.950 4.951 cp_fm_diag_elpa 50 13.0 0.001 0.002 4.596 4.596 cp_fm_redistribute_end 50 14.0 2.339 4.557 2.354 4.562 cp_fm_diag_elpa_base 50 14.0 2.197 4.419 2.204 4.431 calculate_dm_sparse 119 9.5 0.000 0.000 3.468 3.608 acc_transpose_blocks 54864 15.5 0.226 0.256 2.284 3.265 density_rs2pw 119 9.7 0.004 0.004 3.161 3.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.001 3.213 apply_single 119 13.6 0.000 0.000 3.001 3.213 multiply_cannon_metrocomm1 54864 15.5 0.056 0.062 1.861 3.163 calculate_first_density_matrix 1 7.0 0.000 0.000 3.018 3.033 ot_diis_step 108 11.5 0.006 0.006 2.855 2.856 jit_kernel_multiply 13 15.8 2.138 2.811 2.138 2.811 make_m2s 4572 13.5 0.054 0.056 2.644 2.762 make_images 4572 14.5 0.134 0.139 2.560 2.678 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.586 2.652 pw_transfer 1439 11.6 0.051 0.057 2.522 2.650 potential_pw2rs 119 12.3 0.004 0.004 2.591 2.647 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.447 2.578 init_scf_loop 11 6.9 0.000 0.000 2.496 2.497 wfi_extrapolate 11 7.9 0.001 0.001 2.476 2.476 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.390 2.392 fft3d_ps 1201 14.6 0.374 0.485 2.164 2.294 grid_integrate_task_list 119 12.3 2.008 2.127 2.008 2.127 mp_sum_d 4135 12.0 1.419 2.076 1.419 2.076 transfer_rs2pw 487 10.6 0.005 0.006 1.934 2.049 fft_wrap_pw1pw2_140 487 13.2 0.138 0.153 1.867 1.997 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.939 1.993 mp_alltoall_d11v 2130 13.8 1.858 1.976 1.858 1.976 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.909 1.929 transfer_pw2rs 487 13.2 0.006 0.007 1.828 1.883 acc_transpose_blocks_kernels 54864 16.5 0.246 0.385 0.832 1.539 mp_waitany 12084 13.8 1.396 1.527 1.396 1.527 make_images_sizes 4572 15.5 0.004 0.005 1.184 1.472 mp_alltoall_i44 4572 16.5 1.180 1.468 1.180 1.468 acc_transpose_blocks_sync 164592 16.5 1.180 1.439 1.180 1.439 grid_collocate_task_list 119 9.7 1.350 1.427 1.350 1.427 mp_alltoall_z22v 1201 16.6 1.254 1.350 1.254 1.350 prepare_preconditioner 11 7.9 0.000 0.000 1.278 1.315 make_preconditioner 11 8.9 0.000 0.000 1.278 1.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.211 1.257 dbcsr_dot_sd 1205 11.9 0.048 0.060 0.791 1.195 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=59.284000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 488.144896E+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 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 1292196. 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 0.016 0.051 45.636 45.637 qs_mol_dyn_low 1 2.0 0.003 0.003 45.294 45.302 qs_forces 11 3.9 0.003 0.003 45.176 45.176 qs_energies 11 4.9 0.003 0.012 43.354 43.358 scf_env_do_scf 11 5.9 0.000 0.001 37.713 37.713 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 34.462 34.462 dbcsr_multiply_generic 2286 12.5 0.102 0.106 23.979 24.361 qs_scf_new_mos 108 7.5 0.001 0.001 23.620 23.849 qs_scf_loop_do_ot 108 8.5 0.001 0.001 23.619 23.848 ot_scf_mini 108 9.5 0.002 0.003 22.614 22.776 velocity_verlet 10 3.0 0.001 0.003 21.838 21.840 multiply_cannon 2286 13.5 0.209 0.218 17.786 19.379 multiply_cannon_loop 2286 14.5 1.199 1.261 16.337 17.949 ot_mini 108 10.5 0.001 0.001 13.657 13.892 mp_waitall_1 200699 16.5 7.243 12.084 7.243 12.084 qs_ot_get_derivative 108 11.5 0.001 0.001 11.179 11.342 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 4.562 9.757 multiply_cannon_multrec 27432 15.5 1.810 4.160 6.136 8.949 rebuild_ks_matrix 119 8.3 0.000 0.000 8.043 8.212 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.043 8.211 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.117 7.266 dbcsr_mm_accdrv_process 47894 16.0 3.529 5.858 4.245 6.601 qs_ot_get_p 119 10.4 0.001 0.001 5.792 6.015 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.489 5.269 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.863 4.896 calculate_rho_elec 119 8.7 0.021 0.024 4.863 4.895 sum_up_and_integrate 119 10.3 0.001 0.001 4.838 4.847 integrate_v_rspace 119 11.3 0.002 0.003 4.824 4.831 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.017 4.439 apply_single 119 13.6 0.000 0.000 3.016 4.439 mp_sum_l 7287 12.8 2.297 4.366 2.297 4.366 init_scf_run 11 5.9 0.000 0.001 4.282 4.283 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.282 4.282 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.709 3.733 make_m2s 4572 13.5 0.052 0.054 3.450 3.731 make_images 4572 14.5 0.206 0.243 3.361 3.645 cp_dbcsr_syevd 50 12.0 0.003 0.003 3.260 3.261 init_scf_loop 11 6.9 0.000 0.000 3.208 3.209 density_rs2pw 119 9.7 0.004 0.004 3.122 3.198 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.737 2.825 cp_fm_diag_elpa 50 13.0 0.000 0.001 2.790 2.791 cp_fm_redistribute_end 50 14.0 1.407 2.738 1.420 2.745 cp_fm_diag_elpa_base 50 14.0 1.282 2.596 1.316 2.652 potential_pw2rs 119 12.3 0.006 0.007 2.514 2.546 pw_transfer 1439 11.6 0.065 0.068 2.374 2.432 ot_diis_step 108 11.5 0.011 0.011 2.410 2.411 calculate_first_density_matrix 1 7.0 0.000 0.000 2.408 2.410 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.283 2.342 acc_transpose_blocks 27432 15.5 0.112 0.117 1.649 2.245 calculate_dm_sparse 119 9.5 0.000 0.001 2.159 2.243 prepare_preconditioner 11 7.9 0.000 0.000 2.169 2.197 make_preconditioner 11 8.9 0.000 0.000 2.169 2.197 multiply_cannon_sync_h2d 27432 15.5 1.673 2.189 1.673 2.189 transfer_rs2pw 487 10.6 0.005 0.005 2.051 2.165 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.040 2.103 make_images_data 4572 15.5 0.047 0.054 1.506 2.094 fft3d_ps 1201 14.6 0.520 0.579 1.920 1.980 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.964 1.967 grid_integrate_task_list 119 12.3 1.827 1.949 1.827 1.949 wfi_extrapolate 11 7.9 0.001 0.001 1.788 1.789 transfer_pw2rs 487 13.2 0.004 0.005 1.755 1.783 hybrid_alltoall_any 4725 16.4 0.053 0.116 1.310 1.765 mp_alltoall_d11v 2130 13.8 1.591 1.728 1.591 1.728 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.663 1.702 fft_wrap_pw1pw2_140 487 13.2 0.158 0.167 1.568 1.630 jit_kernel_multiply 8 16.5 0.656 1.506 0.656 1.506 make_images_sizes 4572 15.5 0.005 0.005 1.082 1.469 mp_alltoall_i44 4572 16.5 1.077 1.464 1.077 1.464 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.336 1.380 mp_allgather_i34 2286 14.5 0.870 1.364 0.870 1.364 grid_collocate_task_list 119 9.7 1.295 1.359 1.295 1.359 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.293 1.304 mp_sendrecv_dv 16779 12.7 1.225 1.259 1.225 1.259 acc_transpose_blocks_kernels 27432 16.5 0.189 0.280 0.700 1.180 mp_sum_d 4135 12.0 0.644 1.158 0.644 1.158 mp_alltoall_z22v 1201 16.6 1.050 1.152 1.050 1.152 cp_fm_cholesky_invert 11 10.9 1.020 1.025 1.020 1.025 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.988 1.008 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.974 0.975 dbcsr_complete_redistribute 329 12.2 0.082 0.112 0.864 0.947 acc_transpose_blocks_sync 82296 16.5 0.811 0.939 0.811 0.939 transfer_rs2pw_50 119 11.7 0.185 0.194 0.853 0.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=45.637000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.909091, yerr=1.880742 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 520.011776E+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 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. 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 0.007 0.028 34.727 34.727 qs_mol_dyn_low 1 2.0 0.003 0.003 34.358 34.394 qs_forces 11 3.9 0.003 0.011 34.296 34.297 qs_energies 11 4.9 0.001 0.002 32.712 32.716 scf_env_do_scf 11 5.9 0.000 0.001 27.535 27.535 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 24.918 24.919 dbcsr_multiply_generic 2286 12.5 0.097 0.099 17.978 18.133 qs_scf_new_mos 108 7.5 0.001 0.001 16.938 16.958 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.937 16.957 velocity_verlet 10 3.0 0.001 0.001 16.261 16.262 ot_scf_mini 108 9.5 0.002 0.002 16.133 16.152 multiply_cannon 2286 13.5 0.195 0.201 14.119 14.888 multiply_cannon_loop 2286 14.5 0.864 0.910 13.249 14.002 ot_mini 108 10.5 0.001 0.001 10.005 10.031 qs_ot_get_derivative 108 11.5 0.001 0.001 8.423 8.440 multiply_cannon_multrec 18288 15.5 1.866 2.897 7.024 7.375 dbcsr_mm_accdrv_process 38222 16.0 4.994 6.124 5.067 6.192 rebuild_ks_matrix 119 8.3 0.000 0.000 6.102 6.135 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.101 6.134 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.366 5.396 mp_waitall_1 158411 16.6 3.108 3.967 3.108 3.967 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.286 3.966 qs_ot_get_p 119 10.4 0.001 0.001 3.783 3.808 init_scf_run 11 5.9 0.000 0.001 3.787 3.787 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.786 3.786 sum_up_and_integrate 119 10.3 0.001 0.001 3.664 3.669 integrate_v_rspace 119 11.3 0.002 0.003 3.651 3.658 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.470 3.484 calculate_rho_elec 119 8.7 0.031 0.031 3.470 3.483 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.706 2.891 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.135 2.682 apply_single 119 13.6 0.000 0.000 2.135 2.682 init_scf_loop 11 6.9 0.000 0.000 2.598 2.599 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.460 2.469 calculate_first_density_matrix 1 7.0 0.000 0.000 2.454 2.455 make_m2s 4572 13.5 0.044 0.045 2.078 2.236 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.151 2.151 make_images 4572 14.5 0.193 0.204 1.993 2.149 density_rs2pw 119 9.7 0.004 0.004 2.016 2.081 mp_sum_l 7287 12.8 1.397 2.028 1.397 2.028 calculate_dm_sparse 119 9.5 0.000 0.001 1.909 1.919 pw_transfer 1439 11.6 0.065 0.069 1.884 1.894 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.884 1.893 grid_integrate_task_list 119 12.3 1.806 1.888 1.806 1.888 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.839 1.850 cp_fm_diag_elpa_base 50 14.0 1.814 1.827 1.837 1.847 prepare_preconditioner 11 7.9 0.000 0.000 1.836 1.839 make_preconditioner 11 8.9 0.000 0.000 1.836 1.839 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.792 1.803 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.677 1.769 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.751 1.753 acc_transpose_blocks 18288 15.5 0.079 0.080 1.638 1.744 ot_diis_step 108 11.5 0.011 0.011 1.542 1.542 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.516 1.527 potential_pw2rs 119 12.3 0.007 0.009 1.458 1.464 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.359 1.370 fft3d_ps 1201 14.6 0.521 0.539 1.360 1.369 grid_collocate_task_list 119 9.7 1.234 1.326 1.234 1.326 wfi_extrapolate 11 7.9 0.001 0.001 1.281 1.281 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.210 1.219 transfer_rs2pw 487 10.6 0.005 0.005 1.133 1.211 multiply_cannon_sync_h2d 18288 15.5 1.008 1.148 1.008 1.148 make_images_data 4572 15.5 0.047 0.051 0.882 1.067 qs_energies_init_hamiltonians 11 5.9 0.024 0.096 1.060 1.061 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.025 1.052 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.770 0.976 mp_alltoall_d11v 2130 13.8 0.785 0.913 0.785 0.913 transfer_pw2rs 487 13.2 0.004 0.004 0.909 0.911 acc_transpose_blocks_sync 54864 16.5 0.740 0.850 0.740 0.850 acc_transpose_blocks_kernels 18288 16.5 0.218 0.226 0.799 0.816 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.807 0.810 cp_fm_cholesky_invert 11 10.9 0.797 0.800 0.797 0.800 mp_alltoall_z22v 1201 16.6 0.672 0.739 0.672 0.739 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.664 0.727 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.727000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.545455, yerr=0.987525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 558.342144E+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 63489. 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 0.022 0.061 44.747 44.756 qs_mol_dyn_low 1 2.0 0.003 0.003 44.502 44.518 qs_forces 11 3.9 0.011 0.019 44.411 44.412 qs_energies 11 4.9 0.001 0.002 42.481 42.498 scf_env_do_scf 11 5.9 0.000 0.001 36.820 36.821 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 32.345 32.345 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.642 22.893 velocity_verlet 10 3.0 0.001 0.002 22.579 22.582 qs_scf_new_mos 108 7.5 0.001 0.001 22.304 22.377 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.303 22.377 ot_scf_mini 108 9.5 0.002 0.003 21.165 21.220 multiply_cannon 2286 13.5 0.218 0.225 17.234 18.111 multiply_cannon_loop 2286 14.5 1.547 1.626 15.916 16.665 ot_mini 108 10.5 0.001 0.001 12.710 12.777 qs_ot_get_derivative 108 11.5 0.001 0.001 10.668 10.723 multiply_cannon_multrec 27432 15.5 2.457 3.282 9.201 9.832 dbcsr_mm_accdrv_process 47916 15.9 6.197 7.989 6.641 8.070 rebuild_ks_matrix 119 8.3 0.000 0.000 7.576 7.652 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 7.575 7.652 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.691 6.760 qs_ot_get_p 119 10.4 0.001 0.001 5.334 5.421 sum_up_and_integrate 119 10.3 0.001 0.001 4.577 4.581 integrate_v_rspace 119 11.3 0.003 0.003 4.565 4.570 mp_waitall_1 137007 16.6 3.580 4.523 3.580 4.523 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.995 4.485 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.440 4.472 calculate_rho_elec 119 8.7 0.040 0.046 4.439 4.472 init_scf_loop 11 6.9 0.000 0.000 4.445 4.446 init_scf_run 11 5.9 0.000 0.001 4.213 4.214 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.213 4.213 make_m2s 4572 13.5 0.054 0.057 3.370 3.478 prepare_preconditioner 11 7.9 0.000 0.000 3.420 3.427 make_preconditioner 11 8.9 0.000 0.000 3.420 3.427 qs_ot_p2m_diag 50 11.0 0.015 0.023 3.376 3.391 make_images 4572 14.5 0.274 0.332 3.261 3.373 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.997 3.343 density_rs2pw 119 9.7 0.004 0.004 2.842 2.968 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.917 2.941 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.878 2.879 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.288 2.749 apply_single 119 13.6 0.000 0.000 2.288 2.749 acc_transpose_blocks 27432 15.5 0.118 0.122 2.416 2.547 pw_transfer 1439 11.6 0.065 0.069 2.402 2.461 cp_fm_diag_elpa 50 13.0 0.001 0.001 2.429 2.440 cp_fm_diag_elpa_base 50 14.0 2.383 2.404 2.425 2.436 calculate_first_density_matrix 1 7.0 0.000 0.000 2.392 2.395 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.310 2.373 calculate_dm_sparse 119 9.5 0.000 0.000 2.259 2.331 potential_pw2rs 119 12.3 0.009 0.010 2.307 2.325 mp_sum_l 7287 12.8 1.625 2.213 1.625 2.213 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 1.143 2.110 ot_diis_step 108 11.5 0.012 0.012 1.977 1.978 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.932 1.933 transfer_rs2pw 487 10.6 0.004 0.005 1.759 1.917 grid_integrate_task_list 119 12.3 1.812 1.913 1.812 1.913 fft3d_ps 1201 14.6 0.553 0.602 1.831 1.877 fft_wrap_pw1pw2_140 487 13.2 0.244 0.257 1.686 1.751 wfi_extrapolate 11 7.9 0.001 0.001 1.745 1.745 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.673 1.693 make_images_data 4572 15.5 0.047 0.052 1.373 1.635 acc_transpose_blocks_sync 82296 16.5 1.444 1.572 1.444 1.572 transfer_pw2rs 487 13.2 0.004 0.004 1.532 1.554 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.501 1.518 hybrid_alltoall_any 4725 16.4 0.065 0.153 1.132 1.384 mp_alltoall_d11v 2130 13.8 1.140 1.379 1.140 1.379 grid_collocate_task_list 119 9.7 1.249 1.332 1.249 1.332 dbcsr_complete_redistribute 329 12.2 0.115 0.141 1.027 1.315 cp_fm_cholesky_invert 11 10.9 1.289 1.295 1.289 1.295 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.266 1.279 cp_fm_upper_to_full 72 14.2 0.878 1.252 0.878 1.252 mp_alltoall_z22v 1201 16.6 1.160 1.207 1.160 1.207 jit_kernel_multiply 6 15.8 0.377 1.199 0.377 1.199 make_images_sizes 4572 15.5 0.005 0.005 0.865 1.162 mp_alltoall_i44 4572 16.5 0.860 1.157 0.860 1.157 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.102 1.112 mp_sendrecv_dv 8211 12.7 1.032 1.063 1.032 1.063 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.980 0.986 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.680 0.963 multiply_cannon_metrocomm1 27432 15.5 0.035 0.037 0.342 0.933 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=44.756000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=529.363636, yerr=3.723945 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 620.859392E+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 63488. 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 0.022 0.058 36.281 36.289 qs_mol_dyn_low 1 2.0 0.003 0.003 35.803 35.849 qs_forces 11 3.9 0.010 0.032 35.730 35.735 qs_energies 11 4.9 0.001 0.002 33.863 33.875 scf_env_do_scf 11 5.9 0.000 0.001 28.187 28.188 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 24.787 24.788 velocity_verlet 10 3.0 0.010 0.010 18.316 18.319 dbcsr_multiply_generic 2286 12.5 0.095 0.098 15.610 15.730 qs_scf_new_mos 108 7.5 0.001 0.001 15.374 15.408 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.373 15.407 ot_scf_mini 108 9.5 0.002 0.002 14.525 14.558 multiply_cannon 2286 13.5 0.226 0.231 11.627 12.381 multiply_cannon_loop 2286 14.5 0.642 0.674 10.457 10.690 ot_mini 108 10.5 0.001 0.001 7.924 7.960 rebuild_ks_matrix 119 8.3 0.000 0.000 6.877 6.920 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 6.877 6.920 qs_ot_get_derivative 108 11.5 0.001 0.001 6.447 6.478 multiply_cannon_multrec 9144 15.5 1.735 1.921 6.241 6.470 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.137 6.175 dbcsr_mm_accdrv_process 12550 15.8 3.753 4.403 4.395 4.484 qs_ot_get_p 119 10.4 0.001 0.001 4.353 4.388 sum_up_and_integrate 119 10.3 0.001 0.001 4.353 4.357 integrate_v_rspace 119 11.3 0.003 0.003 4.343 4.347 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.274 4.290 calculate_rho_elec 119 8.7 0.060 0.061 4.274 4.290 init_scf_run 11 5.9 0.000 0.001 4.056 4.056 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.056 4.056 mp_waitall_1 115863 16.7 2.994 3.512 2.994 3.512 init_scf_loop 11 6.9 0.000 0.000 3.368 3.369 make_m2s 4572 13.5 0.034 0.035 2.733 2.952 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.908 2.911 make_images 4572 14.5 0.270 0.301 2.643 2.861 density_rs2pw 119 9.7 0.003 0.003 2.549 2.650 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.565 2.566 calculate_first_density_matrix 1 7.0 0.000 0.000 2.541 2.542 prepare_preconditioner 11 7.9 0.000 0.000 2.485 2.491 make_preconditioner 11 8.9 0.000 0.000 2.485 2.491 pw_transfer 1439 11.6 0.065 0.068 2.431 2.445 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.338 2.354 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.311 2.353 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.289 2.309 cp_fm_diag_elpa 50 13.0 0.017 0.017 2.138 2.146 cp_fm_diag_elpa_base 50 14.0 2.086 2.110 2.119 2.126 calculate_dm_sparse 119 9.5 0.000 0.000 2.011 2.037 potential_pw2rs 119 12.3 0.010 0.011 2.018 2.025 grid_integrate_task_list 119 12.3 1.922 1.993 1.922 1.993 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.846 1.848 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.823 1.841 fft3d_ps 1201 14.6 0.566 0.582 1.752 1.771 fft_wrap_pw1pw2_140 487 13.2 0.323 0.331 1.752 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.574 1.585 transfer_rs2pw 487 10.6 0.004 0.004 1.470 1.576 make_images_data 4572 15.5 0.042 0.045 1.240 1.574 acc_transpose_blocks 9144 15.5 0.041 0.042 1.521 1.571 hybrid_alltoall_any 4725 16.4 0.065 0.175 1.154 1.539 wfi_extrapolate 11 7.9 0.001 0.001 1.444 1.444 ot_diis_step 108 11.5 0.013 0.013 1.443 1.444 grid_collocate_task_list 119 9.7 1.343 1.420 1.343 1.420 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.338 1.381 apply_single 119 13.6 0.000 0.000 1.338 1.380 cp_fm_cholesky_invert 11 10.9 1.330 1.335 1.330 1.335 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.319 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.287 1.293 mp_alltoall_d11v 2130 13.8 1.189 1.267 1.189 1.267 transfer_pw2rs 487 13.2 0.003 0.003 1.250 1.257 jit_kernel_multiply 6 15.4 0.602 1.221 0.602 1.221 mp_allgather_i34 2286 14.5 0.482 1.092 0.482 1.092 mp_alltoall_z22v 1201 16.6 1.058 1.088 1.058 1.088 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.070 1.080 make_images_sizes 4572 15.5 0.005 0.005 0.713 1.069 mp_alltoall_i44 4572 16.5 0.708 1.064 0.708 1.064 mp_sum_l 7287 12.8 0.835 0.989 0.835 0.989 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.839 0.978 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.875 0.922 dbcsr_complete_redistribute 329 12.2 0.159 0.166 0.837 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.868 0.872 qs_env_update_s_mstruct 11 6.9 0.011 0.013 0.768 0.822 mp_sendrecv_dv 5355 12.7 0.768 0.778 0.768 0.778 acc_transpose_blocks_sync 27432 16.5 0.712 0.759 0.712 0.759 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.752 0.755 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=36.289000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.181818, yerr=4.763845 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 762.757120E+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 63723. 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 0.026 0.064 49.261 49.280 qs_mol_dyn_low 1 2.0 0.003 0.003 48.853 48.861 qs_forces 11 3.9 0.003 0.004 48.764 48.766 qs_energies 11 4.9 0.002 0.002 46.649 46.653 scf_env_do_scf 11 5.9 0.001 0.001 40.434 40.434 scf_env_do_scf_inner_loop 108 6.5 0.004 0.020 31.614 31.617 velocity_verlet 10 3.0 0.014 0.015 27.343 27.351 dbcsr_multiply_generic 2286 12.5 0.102 0.105 21.192 21.337 qs_scf_new_mos 108 7.5 0.001 0.001 20.535 20.628 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.534 20.627 ot_scf_mini 108 9.5 0.002 0.002 19.264 19.361 multiply_cannon 2286 13.5 0.304 0.316 15.824 16.869 multiply_cannon_loop 2286 14.5 0.870 0.886 14.357 15.378 ot_mini 108 10.5 0.001 0.001 11.616 11.736 qs_ot_get_derivative 108 11.5 0.001 0.001 9.437 9.537 multiply_cannon_multrec 9144 15.5 3.483 5.038 8.936 9.223 init_scf_loop 11 6.9 0.000 0.000 8.778 8.784 rebuild_ks_matrix 119 8.3 0.000 0.000 8.102 8.246 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 8.102 8.245 prepare_preconditioner 11 7.9 0.000 0.000 7.697 7.711 make_preconditioner 11 8.9 0.000 0.000 7.697 7.711 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.185 7.575 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.289 7.418 dbcsr_mm_accdrv_process 12550 15.8 4.596 6.535 5.314 6.837 mp_waitall_1 94719 16.7 3.720 5.104 3.720 5.104 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.960 4.967 calculate_rho_elec 119 8.7 0.118 0.121 4.959 4.967 cp_fm_upper_to_full 72 14.2 3.209 4.661 3.209 4.661 qs_ot_get_p 119 10.4 0.001 0.001 4.447 4.592 sum_up_and_integrate 119 10.3 0.001 0.001 4.568 4.573 integrate_v_rspace 119 11.3 0.003 0.003 4.557 4.563 init_scf_run 11 5.9 0.000 0.001 4.020 4.020 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.020 4.020 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.435 3.891 make_m2s 4572 13.5 0.038 0.038 3.421 3.634 make_images 4572 14.5 0.354 0.385 3.300 3.513 pw_transfer 1439 11.6 0.068 0.069 3.174 3.183 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 3.076 3.085 dbcsr_complete_redistribute 329 12.2 0.282 0.288 2.234 3.084 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.344 2.863 apply_single 119 13.6 0.000 0.000 2.344 2.863 density_rs2pw 119 9.7 0.003 0.003 2.782 2.797 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.684 2.754 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.707 2.710 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.806 2.638 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.489 2.540 fft_wrap_pw1pw2_140 487 13.2 0.576 0.580 2.450 2.459 mp_sum_l 7287 12.8 1.401 2.433 1.401 2.433 calculate_dm_sparse 119 9.5 0.000 0.000 2.360 2.374 mp_alltoall_i22 627 13.8 1.501 2.373 1.501 2.373 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.507 2.334 calculate_first_density_matrix 1 7.0 0.000 0.000 2.327 2.328 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.303 2.304 grid_integrate_task_list 119 12.3 2.166 2.185 2.166 2.185 fft3d_ps 1201 14.6 0.600 0.610 2.171 2.175 ot_diis_step 108 11.5 0.014 0.014 2.140 2.141 make_images_data 4572 15.5 0.046 0.049 1.678 2.095 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.627 2.063 cp_fm_cholesky_invert 11 10.9 1.952 1.955 1.952 1.955 cp_fm_diag_elpa 50 13.0 0.001 0.001 1.924 1.924 cp_fm_diag_elpa_base 50 14.0 1.763 1.825 1.920 1.921 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.911 1.912 potential_pw2rs 119 12.3 0.014 0.014 1.895 1.898 acc_transpose_blocks 9144 15.5 0.043 0.044 1.831 1.866 qs_energies_init_hamiltonians 11 5.9 0.010 0.020 1.816 1.817 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.598 1.642 wfi_extrapolate 11 7.9 0.001 0.001 1.628 1.628 grid_collocate_task_list 119 9.7 1.581 1.596 1.581 1.596 mp_alltoall_d11v 2130 13.8 1.573 1.586 1.573 1.586 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.443 1.468 mp_alltoall_z22v 1201 16.6 1.433 1.454 1.433 1.454 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.276 transfer_rs2pw 487 10.6 0.005 0.005 1.167 1.193 qs_env_update_s_mstruct 11 6.9 0.008 0.010 1.133 1.153 acc_transpose_blocks_sync 27432 16.5 1.111 1.132 1.111 1.132 make_images_sizes 4572 15.5 0.005 0.005 0.641 1.075 mp_alltoall_i44 4572 16.5 0.637 1.071 0.637 1.071 jit_kernel_multiply 5 15.4 0.688 1.063 0.688 1.063 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.029 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.991 1.014 transfer_pw2rs 487 13.2 0.003 0.003 1.003 1.007 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=49.280000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=714.272727, yerr=14.697501 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.547008E+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 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. 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 0.011 0.031 94.122 94.123 qs_mol_dyn_low 1 2.0 0.003 0.003 93.759 93.768 qs_forces 11 3.9 0.003 0.003 93.682 93.683 qs_energies 11 4.9 0.002 0.002 90.601 90.613 scf_env_do_scf 11 5.9 0.000 0.001 80.111 80.114 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 72.799 72.800 dbcsr_multiply_generic 2055 12.4 0.109 0.111 54.198 54.452 qs_scf_new_mos 99 7.5 0.000 0.001 53.077 53.191 qs_scf_loop_do_ot 99 8.5 0.000 0.001 53.077 53.190 ot_scf_mini 99 9.5 0.002 0.002 50.548 50.646 velocity_verlet 10 3.0 0.002 0.010 48.316 48.317 multiply_cannon 2055 13.4 0.186 0.191 43.896 44.835 multiply_cannon_loop 2055 14.4 1.796 1.839 42.484 43.411 ot_mini 99 10.5 0.001 0.001 28.165 28.272 qs_ot_get_derivative 99 11.5 0.001 0.001 21.284 21.376 multiply_cannon_multrec 49320 15.4 11.231 11.952 17.863 18.523 rebuild_ks_matrix 110 8.3 0.000 0.001 16.093 16.205 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 16.093 16.205 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.170 14.273 qs_ot_get_p 110 10.4 0.001 0.001 13.327 13.489 mp_waitall_1 220248 16.4 12.562 13.381 12.562 13.381 multiply_cannon_sync_h2d 49320 15.4 9.352 9.960 9.352 9.960 qs_ot_p2m_diag 48 11.0 0.012 0.019 9.279 9.330 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.938 8.460 sum_up_and_integrate 110 10.3 0.002 0.003 8.353 8.368 integrate_v_rspace 110 11.3 0.003 0.003 8.328 8.349 cp_dbcsr_syevd 48 12.0 0.002 0.003 8.319 8.320 init_scf_run 11 5.9 0.000 0.001 8.263 8.263 scf_env_initial_rho_setup 11 6.9 0.000 0.001 8.263 8.263 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.808 7.942 calculate_rho_elec 110 8.6 0.021 0.024 7.807 7.942 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.296 7.751 apply_single 110 13.6 0.000 0.000 7.296 7.750 multiply_cannon_metrocomm3 49320 15.4 0.082 0.085 6.513 7.636 cp_fm_diag_elpa 48 13.0 0.000 0.001 7.396 7.422 cp_fm_diag_elpa_base 48 14.0 7.361 7.391 7.392 7.418 init_scf_loop 11 6.9 0.000 0.000 7.271 7.272 dbcsr_mm_accdrv_process 87628 16.1 3.020 3.130 6.503 6.828 ot_diis_step 99 11.5 0.006 0.006 6.675 6.676 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.077 6.123 mp_sum_l 6594 12.7 4.518 5.632 4.518 5.632 make_m2s 4110 13.4 0.062 0.065 5.336 5.435 make_images 4110 14.4 0.178 0.191 5.236 5.336 density_rs2pw 110 9.6 0.004 0.005 4.801 5.005 prepare_preconditioner 11 7.9 0.000 0.000 4.946 4.969 make_preconditioner 11 8.9 0.000 0.000 4.946 4.968 make_full_inverse_cholesky 11 9.9 0.000 0.000 4.697 4.752 pw_transfer 1331 11.6 0.056 0.064 4.494 4.610 wfi_extrapolate 11 7.9 0.001 0.001 4.579 4.579 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.404 4.525 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 3.056 4.392 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 4.381 4.386 calculate_dm_sparse 110 9.5 0.001 0.001 3.954 4.068 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.903 3.958 potential_pw2rs 110 12.3 0.005 0.006 3.891 3.934 fft3d_ps 1111 14.6 0.792 0.887 3.705 3.795 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.679 3.749 fft_wrap_pw1pw2_140 451 13.1 0.385 0.423 3.456 3.580 calculate_first_density_matrix 1 7.0 0.000 0.000 3.540 3.549 grid_integrate_task_list 110 12.3 3.233 3.451 3.233 3.451 jit_kernel_multiply 13 15.9 3.203 3.247 3.203 3.247 mp_alltoall_d11v 2046 13.8 2.457 3.018 2.457 3.018 acc_transpose_blocks 49320 15.4 0.214 0.225 2.568 2.899 transfer_rs2pw 451 10.6 0.005 0.006 2.660 2.847 cp_fm_cholesky_invert 11 10.9 2.803 2.808 2.803 2.808 mp_alltoall_z22v 1111 16.6 2.425 2.601 2.425 2.601 transfer_pw2rs 451 13.1 0.006 0.007 2.528 2.566 make_images_data 4110 15.4 0.044 0.047 2.190 2.409 grid_collocate_task_list 110 9.6 2.155 2.285 2.155 2.285 hybrid_alltoall_any 4261 16.3 0.084 0.476 1.871 2.149 mp_waitany 14300 13.8 1.935 2.138 1.935 2.138 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.058 2.076 make_images_sizes 4110 15.4 0.004 0.005 1.527 1.994 mp_alltoall_i44 4110 16.4 1.522 1.990 1.522 1.990 mp_sum_d 3889 11.9 1.458 1.944 1.458 1.944 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=94.123000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.000000, yerr=2.923261 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 591.687680E+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 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 2905426. 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 0.008 0.028 79.843 79.844 qs_mol_dyn_low 1 2.0 0.003 0.003 79.526 79.536 qs_forces 11 3.9 0.003 0.003 79.450 79.456 qs_energies 11 4.9 0.002 0.002 76.009 76.028 scf_env_do_scf 11 5.9 0.000 0.001 66.703 66.706 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 57.086 57.087 dbcsr_multiply_generic 2055 12.4 0.118 0.122 41.093 41.699 velocity_verlet 10 3.0 0.004 0.009 41.126 41.127 qs_scf_new_mos 99 7.5 0.001 0.001 39.245 39.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 39.244 39.393 ot_scf_mini 99 9.5 0.003 0.003 37.394 37.522 multiply_cannon 2055 13.4 0.224 0.244 32.389 33.695 multiply_cannon_loop 2055 14.4 1.171 1.205 30.629 32.492 ot_mini 99 10.5 0.001 0.001 21.028 21.184 multiply_cannon_multrec 24660 15.4 6.823 8.622 13.969 15.558 qs_ot_get_derivative 99 11.5 0.001 0.001 15.023 15.151 rebuild_ks_matrix 110 8.3 0.000 0.001 14.656 14.781 qs_ks_build_kohn_sham_matrix 110 9.3 0.019 0.033 14.656 14.781 mp_waitall_1 176588 16.5 9.273 13.973 9.273 13.973 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.976 13.083 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.810 10.206 init_scf_loop 11 6.9 0.000 0.000 9.568 9.569 qs_ot_get_p 110 10.4 0.001 0.001 9.275 9.433 prepare_preconditioner 11 7.9 0.000 0.000 7.344 7.362 make_preconditioner 11 8.9 0.000 0.000 7.344 7.362 sum_up_and_integrate 110 10.3 0.001 0.002 7.343 7.353 integrate_v_rspace 110 11.3 0.002 0.003 7.316 7.327 dbcsr_mm_accdrv_process 52282 16.1 5.450 6.104 6.981 7.299 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.537 7.274 apply_single 110 13.6 0.000 0.001 6.537 7.274 multiply_cannon_sync_h2d 24660 15.4 6.178 7.262 6.178 7.262 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.014 7.060 calculate_rho_elec 110 8.6 0.039 0.047 7.013 7.059 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.870 7.045 init_scf_run 11 5.9 0.000 0.001 6.802 6.802 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.801 6.802 qs_ot_p2m_diag 48 11.0 0.029 0.044 6.696 6.727 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.136 6.136 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.339 6.018 make_m2s 4110 13.4 0.057 0.059 5.278 5.932 ot_diis_step 99 11.5 0.010 0.010 5.868 5.869 make_images 4110 14.4 0.407 0.462 5.166 5.817 cp_fm_diag_elpa 48 13.0 0.004 0.010 5.278 5.293 cp_fm_diag_elpa_base 48 14.0 5.206 5.224 5.266 5.279 pw_transfer 1331 11.6 0.066 0.074 4.327 4.508 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.220 4.403 density_rs2pw 110 9.6 0.004 0.004 4.180 4.344 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.261 4.325 wfi_extrapolate 11 7.9 0.001 0.001 3.928 3.929 mp_sum_l 6594 12.7 2.796 3.652 2.796 3.652 cp_fm_cholesky_invert 11 10.9 3.599 3.612 3.599 3.612 fft3d_ps 1111 14.6 1.112 1.339 3.322 3.519 make_images_data 4110 15.4 0.049 0.053 2.724 3.407 grid_integrate_task_list 110 12.3 3.134 3.373 3.134 3.373 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.369 3.372 fft_wrap_pw1pw2_140 451 13.1 0.455 0.471 3.114 3.302 calculate_dm_sparse 110 9.5 0.001 0.001 3.222 3.261 hybrid_alltoall_any 4261 16.3 0.106 0.454 2.340 3.235 potential_pw2rs 110 12.3 0.008 0.009 3.167 3.196 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.934 2.983 calculate_first_density_matrix 1 7.0 0.000 0.000 2.745 2.749 mp_alltoall_d11v 2046 13.8 2.168 2.500 2.168 2.500 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.361 2.407 grid_collocate_task_list 110 9.6 2.169 2.310 2.169 2.310 transfer_rs2pw 451 10.6 0.006 0.007 2.091 2.253 mp_allgather_i34 2055 14.4 1.067 2.197 1.067 2.197 cp_fm_cholesky_decompose 22 10.9 2.033 2.063 2.033 2.063 acc_transpose_blocks 24660 15.4 0.113 0.117 1.992 2.030 jit_kernel_multiply 10 16.4 1.162 2.014 1.162 2.014 mp_alltoall_z22v 1111 16.6 1.870 1.970 1.870 1.970 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.824 1.918 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.882 1.888 transfer_pw2rs 451 13.1 0.005 0.006 1.857 1.883 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.841 1.856 dbcsr_complete_redistribute 325 12.2 0.233 0.307 1.473 1.764 mp_irecv_dv 57340 16.2 0.697 1.691 0.697 1.691 mp_sum_d 3889 11.9 1.217 1.662 1.217 1.662 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.636 1.651 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=79.844000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.363636, yerr=6.799368 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 666.464256E+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 66421. 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 0.030 0.066 61.087 61.088 qs_mol_dyn_low 1 2.0 0.003 0.003 60.510 60.525 qs_forces 11 3.9 0.003 0.004 60.425 60.426 qs_energies 11 4.9 0.002 0.002 57.251 57.256 scf_env_do_scf 11 5.9 0.000 0.001 48.885 48.885 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 40.228 40.228 velocity_verlet 10 3.0 0.004 0.009 32.926 32.928 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.874 29.116 qs_scf_new_mos 99 7.5 0.001 0.001 25.378 25.472 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.377 25.472 ot_scf_mini 99 9.5 0.002 0.002 24.124 24.233 multiply_cannon 2055 13.4 0.213 0.225 22.562 23.729 multiply_cannon_loop 2055 14.4 0.822 0.845 21.365 22.373 ot_mini 99 10.5 0.001 0.001 13.804 13.914 rebuild_ks_matrix 110 8.3 0.000 0.000 12.237 12.358 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.236 12.358 multiply_cannon_multrec 16440 15.4 3.651 4.771 9.920 10.974 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.778 10.883 mp_waitall_1 139946 16.5 6.961 10.171 6.961 10.171 qs_ot_get_derivative 99 11.5 0.001 0.001 9.381 9.488 init_scf_loop 11 6.9 0.000 0.000 8.621 8.622 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.283 7.246 prepare_preconditioner 11 7.9 0.000 0.000 6.881 6.902 make_preconditioner 11 8.9 0.000 0.000 6.881 6.902 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.188 6.538 sum_up_and_integrate 110 10.3 0.001 0.002 6.370 6.386 integrate_v_rspace 110 11.3 0.003 0.003 6.344 6.360 dbcsr_mm_accdrv_process 34862 16.1 5.411 5.758 6.113 6.203 init_scf_run 11 5.9 0.000 0.001 5.894 5.894 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.894 5.894 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.865 5.873 calculate_rho_elec 110 8.6 0.058 0.059 5.865 5.873 qs_ot_get_p 110 10.4 0.001 0.001 5.480 5.636 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.934 5.394 apply_single 110 13.6 0.000 0.000 4.934 5.394 make_m2s 4110 13.4 0.049 0.051 4.224 4.617 make_images 4110 14.4 0.400 0.524 4.109 4.502 ot_diis_step 99 11.5 0.010 0.011 4.394 4.394 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.186 3.839 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.803 3.806 multiply_cannon_sync_h2d 16440 15.4 3.231 3.765 3.231 3.765 pw_transfer 1331 11.6 0.065 0.072 3.638 3.646 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.531 3.541 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.431 3.431 grid_integrate_task_list 110 12.3 3.196 3.415 3.196 3.415 density_rs2pw 110 9.6 0.004 0.005 3.067 3.244 fft_wrap_pw1pw2_140 451 13.1 0.579 0.585 3.027 3.038 wfi_extrapolate 11 7.9 0.001 0.001 3.015 3.016 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.923 2.924 cp_fm_diag_elpa 48 13.0 0.001 0.001 2.898 2.909 cp_fm_diag_elpa_base 48 14.0 2.828 2.865 2.896 2.906 make_images_data 4110 15.4 0.045 0.050 2.363 2.879 hybrid_alltoall_any 4261 16.3 0.108 0.384 2.125 2.874 calculate_first_density_matrix 1 7.0 0.000 0.000 2.784 2.785 cp_fm_cholesky_invert 11 10.9 2.694 2.700 2.694 2.700 calculate_dm_sparse 110 9.5 0.001 0.001 2.565 2.596 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.501 2.548 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.475 2.535 fft3d_ps 1111 14.6 1.091 1.101 2.474 2.481 multiply_cannon_metrocomm4 14385 15.4 0.049 0.053 0.884 2.449 grid_collocate_task_list 110 9.6 2.219 2.407 2.219 2.407 mp_irecv_dv 48980 15.7 0.808 2.310 0.808 2.310 potential_pw2rs 110 12.3 0.011 0.012 2.198 2.203 mp_alltoall_d11v 2046 13.8 1.883 2.136 1.883 2.136 dbcsr_complete_redistribute 325 12.2 0.340 0.400 1.550 2.020 mp_sum_l 6594 12.7 1.466 1.981 1.466 1.981 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.968 1.970 cp_fm_upper_to_full 70 14.2 1.384 1.772 1.384 1.772 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.653 1.670 cp_fm_cholesky_decompose 22 10.9 1.628 1.651 1.628 1.651 acc_transpose_blocks 16440 15.4 0.076 0.078 1.574 1.625 mp_allgather_i34 2055 14.4 0.508 1.531 0.508 1.531 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.065 1.517 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.499 1.512 transfer_rs2pw 451 10.6 0.005 0.006 1.337 1.507 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.362 1.455 mp_waitany 17072 13.8 1.147 1.299 1.147 1.299 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.277 1.285 transfer_pw2rs 451 13.1 0.005 0.005 1.226 1.230 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.088000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.090909, yerr=8.073014 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 726.884352E+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 66419. 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 0.025 0.070 78.628 78.629 qs_mol_dyn_low 1 2.0 0.003 0.003 78.186 78.196 qs_forces 11 3.9 0.003 0.004 78.026 78.027 qs_energies 11 4.9 0.017 0.076 74.422 74.427 scf_env_do_scf 11 5.9 0.000 0.001 64.964 64.966 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 50.284 50.284 velocity_verlet 10 3.0 0.020 0.031 43.967 43.972 dbcsr_multiply_generic 2055 12.4 0.116 0.122 33.632 34.038 qs_scf_new_mos 99 7.5 0.001 0.001 33.530 33.662 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.529 33.661 ot_scf_mini 99 9.5 0.002 0.003 31.725 31.849 multiply_cannon 2055 13.4 0.236 0.250 24.510 25.785 multiply_cannon_loop 2055 14.4 1.422 1.504 22.683 23.731 ot_mini 99 10.5 0.001 0.001 17.336 17.488 init_scf_loop 11 6.9 0.000 0.000 14.620 14.621 multiply_cannon_multrec 24660 15.4 4.067 6.661 13.023 14.142 rebuild_ks_matrix 110 8.3 0.000 0.000 13.243 13.367 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.242 13.366 qs_ot_get_derivative 99 11.5 0.001 0.001 12.898 13.026 prepare_preconditioner 11 7.9 0.000 0.000 12.687 12.705 make_preconditioner 11 8.9 0.000 0.000 12.687 12.705 make_full_inverse_cholesky 11 9.9 0.000 0.000 10.795 12.292 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.716 11.826 dbcsr_mm_accdrv_process 52304 16.0 7.787 9.181 8.800 9.831 qs_ot_get_p 110 10.4 0.001 0.001 8.837 9.012 mp_waitall_1 121746 16.5 6.121 8.590 6.121 8.590 sum_up_and_integrate 110 10.3 0.001 0.002 7.293 7.303 integrate_v_rspace 110 11.3 0.003 0.003 7.267 7.279 make_m2s 4110 13.4 0.060 0.062 6.843 7.116 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.103 7.115 calculate_rho_elec 110 8.6 0.078 0.081 7.102 7.114 make_images 4110 14.4 0.576 0.694 6.699 6.972 init_scf_run 11 5.9 0.000 0.001 6.469 6.469 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.468 6.469 qs_ot_p2m_diag 48 11.0 0.055 0.064 6.323 6.343 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.664 5.665 cp_fm_upper_to_full 70 14.2 3.482 5.027 3.482 5.027 cp_fm_diag_elpa 48 13.0 0.001 0.001 4.842 4.853 cp_fm_diag_elpa_base 48 14.0 4.665 4.727 4.838 4.849 pw_transfer 1331 11.6 0.065 0.075 4.409 4.459 cp_fm_cholesky_invert 11 10.9 4.430 4.441 4.430 4.441 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.303 4.357 ot_diis_step 99 11.5 0.011 0.011 4.356 4.356 dbcsr_complete_redistribute 325 12.2 0.410 0.488 3.180 4.340 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.192 4.330 apply_single 110 13.6 0.000 0.000 4.192 4.330 density_rs2pw 110 9.6 0.004 0.004 4.101 4.274 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.184 4.237 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.115 4.186 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.800 4.093 make_images_data 4110 15.4 0.049 0.053 3.455 3.863 hybrid_alltoall_any 4261 16.3 0.122 0.461 2.903 3.655 wfi_extrapolate 11 7.9 0.001 0.001 3.649 3.649 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.428 3.583 fft_wrap_pw1pw2_140 451 13.1 0.608 0.629 3.413 3.470 grid_integrate_task_list 110 12.3 3.270 3.414 3.270 3.414 fft3d_ps 1111 14.6 1.086 1.116 3.188 3.224 calculate_dm_sparse 110 9.5 0.001 0.001 3.139 3.167 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.883 3.027 potential_pw2rs 110 12.3 0.013 0.014 3.001 3.022 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.017 3.018 mp_alltoall_i22 605 13.7 1.786 2.969 1.786 2.969 calculate_first_density_matrix 1 7.0 0.000 0.000 2.714 2.716 mp_alltoall_d11v 2046 13.8 2.430 2.649 2.430 2.649 acc_transpose_blocks 24660 15.4 0.112 0.115 2.431 2.580 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.472 2.514 cp_fm_cholesky_decompose 22 10.9 2.477 2.507 2.477 2.507 multiply_cannon_sync_h2d 24660 15.4 2.336 2.430 2.336 2.430 grid_collocate_task_list 110 9.6 2.267 2.404 2.267 2.404 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.308 2.309 transfer_rs2pw 451 10.6 0.005 0.006 1.982 2.196 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.102 2.128 mp_sum_l 6594 12.7 1.598 2.043 1.598 2.043 mp_allgather_i34 2055 14.4 0.915 1.991 0.915 1.991 mp_alltoall_z22v 1111 16.6 1.884 1.956 1.884 1.956 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.776 1.788 transfer_pw2rs 451 13.1 0.004 0.005 1.735 1.758 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.613 1.711 make_images_sizes 4110 15.4 0.005 0.005 1.049 1.645 mp_alltoall_i44 4110 16.4 1.044 1.640 1.044 1.640 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.605 1.619 mp_waitany 13376 13.8 1.317 1.596 1.317 1.596 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=78.629000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=690.727273, yerr=6.743224 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 861.880320E+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 66417. 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 0.045 0.051 66.214 66.219 qs_mol_dyn_low 1 2.0 0.003 0.003 65.711 65.721 qs_forces 11 3.9 0.003 0.003 65.619 65.622 qs_energies 11 4.9 0.002 0.002 61.725 61.733 scf_env_do_scf 11 5.9 0.000 0.001 52.246 52.246 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.150 42.151 velocity_verlet 10 3.0 0.002 0.002 37.005 37.009 dbcsr_multiply_generic 2055 12.4 0.118 0.121 26.050 26.244 qs_scf_new_mos 99 7.5 0.001 0.001 25.834 25.896 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.834 25.895 ot_scf_mini 99 9.5 0.002 0.002 24.442 24.472 multiply_cannon 2055 13.4 0.237 0.249 18.892 20.602 multiply_cannon_loop 2055 14.4 0.605 0.623 17.305 17.811 ot_mini 99 10.5 0.001 0.001 12.902 12.918 rebuild_ks_matrix 110 8.3 0.000 0.000 12.687 12.709 qs_ks_build_kohn_sham_matrix 110 9.3 0.015 0.016 12.686 12.709 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.231 11.253 init_scf_loop 11 6.9 0.000 0.000 10.024 10.026 qs_ot_get_derivative 99 11.5 0.001 0.001 9.036 9.065 mp_waitall_1 103326 16.6 7.388 8.844 7.388 8.844 multiply_cannon_multrec 8220 15.4 3.228 4.403 7.820 8.727 prepare_preconditioner 11 7.9 0.000 0.000 8.249 8.252 make_preconditioner 11 8.9 0.000 0.000 8.249 8.252 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.767 7.881 qs_ot_get_p 110 10.4 0.001 0.001 7.290 7.318 sum_up_and_integrate 110 10.3 0.001 0.002 7.101 7.114 integrate_v_rspace 110 11.3 0.003 0.003 7.074 7.087 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.969 6.982 calculate_rho_elec 110 8.6 0.114 0.114 6.969 6.982 init_scf_run 11 5.9 0.000 0.001 6.088 6.089 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.088 6.088 make_m2s 4110 13.4 0.038 0.040 5.284 5.664 make_images 4110 14.4 0.650 0.709 5.152 5.532 dbcsr_mm_accdrv_process 17442 15.9 3.182 3.922 4.452 5.457 qs_ot_p2m_diag 48 11.0 0.081 0.084 5.231 5.243 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.791 4.792 pw_transfer 1331 11.6 0.066 0.070 4.606 4.627 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.499 4.523 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.839 4.281 cp_fm_cholesky_invert 11 10.9 4.093 4.099 4.093 4.099 cp_fm_diag_elpa 48 13.0 0.022 0.043 3.992 4.000 cp_fm_diag_elpa_base 48 14.0 3.873 3.912 3.946 3.954 density_rs2pw 110 9.6 0.004 0.004 3.826 3.927 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.843 3.886 apply_single 110 13.6 0.000 0.000 3.842 3.886 ot_diis_step 99 11.5 0.012 0.012 3.814 3.815 fft_wrap_pw1pw2_140 451 13.1 0.776 0.788 3.701 3.723 grid_integrate_task_list 110 12.3 3.389 3.506 3.389 3.506 make_images_data 4110 15.4 0.041 0.046 2.745 3.367 wfi_extrapolate 11 7.9 0.001 0.001 3.308 3.308 hybrid_alltoall_any 4261 16.3 0.200 0.858 2.645 3.286 fft3d_ps 1111 14.6 1.142 1.182 3.164 3.194 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.893 2.894 calculate_dm_sparse 110 9.5 0.001 0.001 2.706 2.742 potential_pw2rs 110 12.3 0.015 0.016 2.700 2.711 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.705 2.708 calculate_first_density_matrix 1 7.0 0.000 0.000 2.673 2.675 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.642 2.674 cp_fm_cholesky_decompose 22 10.9 2.607 2.628 2.607 2.628 mp_alltoall_d11v 2046 13.8 2.395 2.576 2.395 2.576 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.551 2.570 multiply_cannon_sync_h2d 8220 15.4 2.372 2.470 2.372 2.470 grid_collocate_task_list 110 9.6 2.374 2.447 2.374 2.447 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.322 2.332 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.582 2.291 dbcsr_complete_redistribute 325 12.2 0.555 0.600 1.957 2.084 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 1.993 mp_allgather_i34 2055 14.4 0.754 1.961 0.754 1.961 mp_alltoall_z22v 1111 16.6 1.784 1.825 1.784 1.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.792 1.796 transfer_rs2pw 451 10.6 0.005 0.005 1.660 1.777 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.538 1.651 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.590 1.601 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.552 1.580 make_images_sizes 4110 15.4 0.005 0.005 0.919 1.565 mp_alltoall_i44 4110 16.4 0.914 1.561 0.914 1.561 acc_transpose_blocks 8220 15.4 0.039 0.040 1.402 1.427 transfer_pw2rs 451 13.1 0.004 0.004 1.381 1.391 mp_sum_l 6594 12.7 1.121 1.384 1.121 1.384 mp_waitany 9240 13.8 1.235 1.378 1.235 1.378 qs_create_task_list 11 7.9 0.000 0.001 1.242 1.340 generate_qs_task_list 11 8.9 0.373 0.438 1.241 1.340 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=66.219000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=810.818182, yerr=12.960206 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 1.420653E+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 67098. 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 0.039 0.078 97.473 97.487 qs_mol_dyn_low 1 2.0 0.003 0.003 96.903 96.928 qs_forces 11 3.9 0.003 0.003 96.805 96.807 qs_energies 11 4.9 0.008 0.025 92.455 92.463 scf_env_do_scf 11 5.9 0.001 0.001 81.457 81.457 velocity_verlet 10 3.0 0.010 0.020 61.672 61.680 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 50.407 50.408 dbcsr_multiply_generic 2055 12.4 0.121 0.128 32.674 32.990 qs_scf_new_mos 99 7.5 0.001 0.001 31.255 31.311 qs_scf_loop_do_ot 99 8.5 0.001 0.001 31.254 31.310 init_scf_loop 11 6.9 0.000 0.000 30.970 30.972 ot_scf_mini 99 9.5 0.002 0.002 29.256 29.303 prepare_preconditioner 11 7.9 0.000 0.000 28.815 28.830 make_preconditioner 11 8.9 0.000 0.000 28.815 28.830 make_full_inverse_cholesky 11 9.9 0.000 0.000 22.931 28.248 multiply_cannon 2055 13.4 0.337 0.355 23.797 24.396 multiply_cannon_loop 2055 14.4 0.831 0.838 21.634 22.489 cp_fm_upper_to_full 70 14.2 12.840 18.425 12.840 18.425 ot_mini 99 10.5 0.001 0.001 16.146 16.227 rebuild_ks_matrix 110 8.3 0.001 0.001 14.839 14.948 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.015 14.838 14.948 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.402 13.500 mp_waitall_1 84994 16.7 9.525 11.353 9.525 11.353 qs_ot_get_derivative 99 11.5 0.001 0.001 11.135 11.185 dbcsr_complete_redistribute 325 12.2 1.010 1.041 7.735 11.013 multiply_cannon_multrec 8220 15.4 4.090 4.237 9.856 9.966 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.540 9.811 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.868 9.098 mp_alltoall_i22 605 13.7 5.546 8.883 5.546 8.883 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.095 8.136 calculate_rho_elec 110 8.6 0.225 0.225 8.095 8.135 sum_up_and_integrate 110 10.3 0.002 0.002 7.576 7.589 integrate_v_rspace 110 11.3 0.003 0.003 7.548 7.561 qs_ot_get_p 110 10.4 0.001 0.001 7.249 7.366 make_m2s 4110 13.4 0.044 0.044 6.552 7.302 make_images 4110 14.4 0.887 0.943 6.358 7.108 cp_fm_cholesky_invert 11 10.9 6.763 6.769 6.763 6.769 init_scf_run 11 5.9 0.000 0.001 6.477 6.477 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.476 6.477 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 5.728 6.456 dbcsr_mm_accdrv_process 11614 15.7 3.985 4.160 5.615 5.833 pw_transfer 1331 11.6 0.074 0.074 5.773 5.790 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.657 5.674 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.030 5.565 apply_single 110 13.6 0.000 0.000 5.030 5.565 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.211 5.221 ot_diis_step 99 11.5 0.015 0.016 4.953 4.955 fft_wrap_pw1pw2_140 451 13.1 1.281 1.288 4.829 4.845 make_images_data 4110 15.4 0.045 0.048 3.567 4.740 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.624 4.625 hybrid_alltoall_any 4261 16.3 0.263 0.562 3.431 4.533 density_rs2pw 110 9.6 0.004 0.004 4.390 4.414 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.475 3.942 cp_fm_diag_elpa 48 13.0 0.003 0.019 3.918 3.919 cp_fm_diag_elpa_base 48 14.0 3.338 3.556 3.897 3.897 wfi_extrapolate 11 7.9 0.001 0.001 3.886 3.886 grid_integrate_task_list 110 12.3 3.723 3.814 3.723 3.814 qs_energies_init_hamiltonians 11 5.9 0.018 0.032 3.795 3.798 fft3d_ps 1111 14.6 1.300 1.318 3.706 3.731 calculate_dm_sparse 110 9.5 0.001 0.001 3.419 3.449 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.259 3.286 multiply_cannon_sync_h2d 8220 15.4 3.124 3.141 3.124 3.141 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.117 3.121 potential_pw2rs 110 12.3 0.022 0.022 2.889 2.898 cp_fm_cholesky_decompose 22 10.9 2.777 2.798 2.777 2.798 grid_collocate_task_list 110 9.6 2.703 2.730 2.703 2.730 mp_alltoall_d11v 2046 13.8 2.623 2.717 2.623 2.717 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.494 2.536 calculate_first_density_matrix 1 7.0 0.000 0.000 2.459 2.460 qs_env_update_s_mstruct 11 6.9 0.001 0.002 2.285 2.348 mp_allgather_i34 2055 14.4 0.907 2.261 0.907 2.261 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.107 2.186 mp_alltoall_z22v 1111 16.6 2.154 2.179 2.154 2.179 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.987 2.057 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.042 2.054 qs_ot_get_orbitals 99 10.5 0.000 0.000 1.979 2.007 qs_create_task_list 11 7.9 0.012 0.024 1.923 1.966 generate_qs_task_list 11 8.9 0.730 0.786 1.911 1.966 mp_sum_l 6594 12.7 1.449 1.964 1.449 1.964 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=97.487000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1269.363636, yerr=58.012965 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 631.517184E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 2582268. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.091 231.413 231.417 qs_mol_dyn_low 1 2.0 0.005 0.018 230.613 230.634 qs_forces 11 3.9 0.005 0.009 230.452 230.453 qs_energies 11 4.9 0.002 0.008 224.537 224.562 scf_env_do_scf 11 5.9 0.001 0.001 206.898 206.902 scf_env_do_scf_inner_loop 118 6.6 0.004 0.037 181.929 181.930 qs_scf_new_mos 118 7.6 0.001 0.001 138.664 138.958 qs_scf_loop_do_ot 118 8.6 0.001 0.001 138.663 138.957 velocity_verlet 10 3.0 0.007 0.011 137.657 137.658 ot_scf_mini 118 9.6 0.003 0.003 131.881 132.144 dbcsr_multiply_generic 2527 12.6 0.184 0.187 130.246 130.955 multiply_cannon 2527 13.6 0.243 0.249 103.691 105.178 multiply_cannon_loop 2527 14.6 2.449 2.524 101.024 102.416 ot_mini 118 10.6 0.001 0.001 70.490 70.723 qs_ot_get_derivative 118 11.6 0.001 0.001 45.192 45.440 multiply_cannon_multrec 60648 15.6 31.688 33.959 41.569 43.871 qs_ot_get_p 129 10.4 0.001 0.001 38.658 39.001 rebuild_ks_matrix 129 8.3 0.001 0.001 35.699 35.908 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 35.698 35.907 mp_waitall_1 269254 16.5 32.375 35.690 32.375 35.690 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.161 32.360 qs_ot_p2m_diag 83 11.4 0.082 0.101 31.175 31.233 cp_dbcsr_syevd 83 12.4 0.011 0.030 28.227 28.232 multiply_cannon_sync_h2d 60648 15.6 26.067 28.078 26.067 28.078 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.534 25.148 apply_single 129 13.6 0.001 0.001 24.534 25.148 ot_diis_step 118 11.6 0.008 0.008 24.944 24.945 init_scf_loop 11 6.9 0.000 0.000 24.884 24.885 cp_fm_diag_elpa 83 13.4 0.001 0.002 24.292 24.332 cp_fm_diag_elpa_base 83 14.4 24.170 24.219 24.285 24.324 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 21.775 22.003 prepare_preconditioner 11 7.9 0.000 0.000 20.057 20.099 make_preconditioner 11 8.9 0.000 0.000 20.057 20.099 make_full_inverse_cholesky 11 9.9 0.000 0.000 19.231 19.440 multiply_cannon_metrocomm3 60648 15.6 0.118 0.125 16.569 18.503 make_m2s 5054 13.6 0.106 0.112 15.469 15.869 sum_up_and_integrate 129 10.3 0.002 0.004 15.855 15.868 integrate_v_rspace 129 11.3 0.003 0.004 15.796 15.813 make_images 5054 14.6 0.404 0.422 15.281 15.689 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.406 15.485 calculate_rho_elec 129 8.7 0.045 0.064 15.405 15.485 init_scf_run 11 5.9 0.000 0.001 13.269 13.269 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.268 13.269 cp_fm_cholesky_invert 11 10.9 11.684 11.696 11.684 11.696 mp_sum_l 8010 12.9 9.859 10.965 9.859 10.965 dbcsr_mm_accdrv_process 125468 16.2 4.770 4.972 9.441 9.993 wfi_extrapolate 11 7.9 0.001 0.001 9.720 9.720 density_rs2pw 129 9.7 0.006 0.007 8.948 9.589 multiply_cannon_metrocomm1 60648 15.6 0.095 0.098 7.544 9.413 pw_transfer 1559 11.6 0.076 0.090 8.816 9.094 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.966 9.066 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 8.612 8.881 calculate_dm_sparse 129 9.5 0.001 0.001 8.626 8.693 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.404 8.519 make_images_data 5054 15.6 0.068 0.072 7.408 8.387 grid_integrate_task_list 129 12.3 7.072 7.773 7.072 7.773 hybrid_alltoall_any 5240 16.5 0.298 2.300 6.378 7.463 fft3d_ps 1301 14.7 2.212 2.867 6.615 7.038 fft_wrap_pw1pw2_140 527 13.2 1.119 1.179 6.728 6.987 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.785 6.798 potential_pw2rs 129 12.3 0.009 0.010 6.203 6.287 cp_fm_cholesky_decompose 22 10.9 6.079 6.094 6.079 6.094 mp_alltoall_d11v 2423 14.1 4.756 6.091 4.756 6.091 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.839 5.955 grid_collocate_task_list 129 9.7 4.877 5.248 4.877 5.248 transfer_rs2pw 527 10.6 0.007 0.009 4.588 5.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=231.417000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.818182, yerr=7.183544 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 838.770688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6003809. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.067 193.680 193.699 qs_mol_dyn_low 1 2.0 0.003 0.003 193.194 193.221 qs_forces 11 3.9 0.006 0.018 193.109 193.110 qs_energies 11 4.9 0.001 0.002 186.333 186.346 scf_env_do_scf 11 5.9 0.001 0.001 169.530 169.539 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 135.349 135.352 velocity_verlet 10 3.0 0.001 0.002 121.661 121.662 dbcsr_multiply_generic 2507 12.6 0.197 0.211 97.891 99.158 qs_scf_new_mos 117 7.6 0.001 0.001 96.569 97.119 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.568 97.118 ot_scf_mini 117 9.6 0.004 0.004 91.730 92.363 multiply_cannon 2507 13.6 0.503 0.563 77.463 81.691 multiply_cannon_loop 2507 14.6 1.579 1.652 74.030 77.103 ot_mini 117 10.6 0.001 0.001 50.715 51.295 mp_waitall_1 214728 16.6 23.992 37.533 23.992 37.533 multiply_cannon_multrec 30084 15.6 21.059 26.087 31.653 36.912 init_scf_loop 11 6.9 0.000 0.000 34.070 34.070 rebuild_ks_matrix 128 8.3 0.001 0.001 32.371 33.185 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.371 33.184 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.114 29.867 prepare_preconditioner 11 7.9 0.000 0.000 29.707 29.771 make_preconditioner 11 8.9 0.000 0.000 29.707 29.771 qs_ot_get_derivative 117 11.6 0.001 0.002 28.638 29.257 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.408 28.946 multiply_cannon_metrocomm3 30084 15.6 0.096 0.101 15.645 28.298 qs_ot_get_p 128 10.4 0.001 0.001 23.015 23.788 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.136 23.324 apply_single 128 13.6 0.001 0.001 22.136 23.324 ot_diis_step 117 11.6 0.014 0.015 21.893 21.895 multiply_cannon_sync_h2d 30084 15.6 18.082 20.395 18.082 20.395 qs_ot_p2m_diag 83 11.4 0.189 0.217 17.901 17.949 cp_fm_cholesky_invert 11 10.9 17.384 17.397 17.384 17.397 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.622 16.623 make_m2s 5014 13.6 0.089 0.094 14.187 15.736 make_images 5014 14.6 1.181 1.383 13.975 15.519 sum_up_and_integrate 128 10.3 0.002 0.004 14.127 14.160 integrate_v_rspace 128 11.3 0.003 0.004 14.067 14.104 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.553 13.594 calculate_rho_elec 128 8.7 0.087 0.104 13.553 13.594 cp_fm_diag_elpa 83 13.4 0.001 0.001 13.331 13.364 cp_fm_diag_elpa_base 83 14.4 13.057 13.161 13.324 13.352 init_scf_run 11 5.9 0.000 0.001 11.932 11.934 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.932 11.934 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.448 11.869 multiply_cannon_metrocomm4 27577 15.6 0.105 0.120 3.828 10.783 dbcsr_mm_accdrv_process 62242 16.2 5.381 6.167 10.052 10.630 mp_irecv_dv 69486 16.3 3.622 10.368 3.622 10.368 make_images_data 5014 15.6 0.067 0.076 8.165 9.945 hybrid_alltoall_any 5200 16.5 0.353 1.546 6.909 9.416 pw_transfer 1547 11.6 0.085 0.097 8.586 8.654 wfi_extrapolate 11 7.9 0.001 0.001 8.505 8.505 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.363 8.435 density_rs2pw 128 9.7 0.006 0.007 7.301 7.737 cp_fm_cholesky_decompose 22 10.9 7.449 7.555 7.449 7.555 grid_integrate_task_list 128 12.3 7.172 7.494 7.172 7.494 fft_wrap_pw1pw2_140 523 13.2 1.213 1.233 7.283 7.377 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.358 7.175 calculate_dm_sparse 128 9.5 0.001 0.001 6.499 6.657 mp_sum_l 7950 12.9 4.443 6.477 4.443 6.477 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.165 6.172 fft3d_ps 1291 14.7 2.837 3.005 5.925 5.982 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.469 5.555 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.237 5.397 grid_collocate_task_list 128 9.7 5.004 5.360 5.004 5.360 mp_allgather_i34 2507 14.6 1.948 4.796 1.948 4.796 potential_pw2rs 128 12.3 0.016 0.019 4.679 4.701 mp_alltoall_d11v 2415 14.1 4.085 4.485 4.085 4.485 mp_sum_d 4465 12.1 2.717 4.087 2.717 4.087 dbcsr_complete_redistribute 395 12.7 0.771 0.855 3.217 4.078 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=193.699000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=799.727273, yerr=0.862439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/17/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 2.906045E+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 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 980.533248E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.032 186.242 186.243 qs_mol_dyn_low 1 2.0 0.003 0.003 185.617 185.733 qs_forces 11 3.9 0.004 0.004 185.490 185.492 qs_energies 11 4.9 0.001 0.002 178.846 178.854 scf_env_do_scf 11 5.9 0.001 0.001 162.710 162.711 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 123.743 123.745 velocity_verlet 10 3.0 0.001 0.002 119.194 119.197 qs_scf_new_mos 116 7.6 0.001 0.001 85.489 85.803 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.488 85.803 dbcsr_multiply_generic 2485 12.5 0.180 0.186 82.193 83.153 ot_scf_mini 116 9.6 0.003 0.004 81.226 81.604 multiply_cannon 2485 13.5 0.500 0.523 61.612 65.534 multiply_cannon_loop 2485 14.5 1.132 1.208 58.530 61.111 ot_mini 116 10.6 0.001 0.001 43.204 43.611 init_scf_loop 11 6.9 0.000 0.000 38.856 38.857 mp_waitall_1 169034 16.6 24.881 34.880 24.881 34.880 prepare_preconditioner 11 7.9 0.000 0.000 34.779 34.828 make_preconditioner 11 8.9 0.000 0.000 34.779 34.828 make_full_inverse_cholesky 11 9.9 0.000 0.000 32.143 33.734 rebuild_ks_matrix 127 8.3 0.001 0.001 30.774 31.212 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.018 30.773 31.211 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.733 28.122 multiply_cannon_multrec 19880 15.5 13.134 15.690 22.435 24.940 qs_ot_get_derivative 116 11.6 0.001 0.002 23.861 24.247 qs_ot_get_p 127 10.4 0.001 0.001 23.439 23.952 multiply_cannon_metrocomm3 19880 15.5 0.061 0.065 14.584 23.902 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.344 20.305 apply_single 127 13.6 0.001 0.001 19.343 20.304 ot_diis_step 116 11.6 0.017 0.018 19.226 19.227 qs_ot_p2m_diag 82 11.4 0.262 0.269 18.870 18.883 cp_dbcsr_syevd 82 12.4 0.005 0.005 17.678 17.679 cp_fm_cholesky_invert 11 10.9 16.631 16.644 16.631 16.644 make_m2s 4970 13.5 0.079 0.082 15.343 16.423 make_images 4970 14.5 1.154 1.246 15.111 16.185 multiply_cannon_sync_h2d 19880 15.5 13.611 15.300 13.611 15.300 sum_up_and_integrate 127 10.3 0.002 0.003 14.763 14.788 integrate_v_rspace 127 11.3 0.003 0.004 14.704 14.730 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.373 14.422 calculate_rho_elec 127 8.7 0.129 0.144 14.372 14.421 cp_fm_diag_elpa 82 13.4 0.001 0.001 14.014 14.041 cp_fm_diag_elpa_base 82 14.4 13.547 13.736 14.009 14.037 init_scf_run 11 5.9 0.000 0.001 11.093 11.093 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.092 11.093 make_images_data 4970 15.5 0.062 0.071 8.922 10.553 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.532 9.816 hybrid_alltoall_any 5155 16.4 0.447 2.045 7.618 9.651 pw_transfer 1535 11.6 0.085 0.102 8.912 9.017 dbcsr_mm_accdrv_process 41158 16.2 5.607 5.933 8.764 8.908 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.690 8.802 cp_fm_cholesky_decompose 22 10.9 8.608 8.662 8.608 8.662 multiply_cannon_metrocomm4 17395 15.5 0.066 0.076 3.398 8.624 mp_irecv_dv 49801 16.2 3.270 8.374 3.270 8.374 density_rs2pw 127 9.7 0.006 0.006 7.769 8.096 wfi_extrapolate 11 7.9 0.001 0.001 7.895 7.895 grid_integrate_task_list 127 12.3 7.280 7.747 7.280 7.747 cp_fm_upper_to_full 104 14.8 5.906 7.680 5.906 7.680 fft_wrap_pw1pw2_140 519 13.2 1.293 1.322 7.408 7.526 dbcsr_complete_redistribute 393 12.7 1.183 1.219 5.118 7.016 fft3d_ps 1281 14.7 2.737 2.954 6.085 6.170 calculate_dm_sparse 127 9.5 0.001 0.001 5.856 5.927 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.823 5.719 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.539 5.547 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.864 5.499 grid_collocate_task_list 127 9.7 5.124 5.392 5.124 5.392 mp_alltoall_d11v 2401 14.1 4.616 5.215 4.616 5.215 potential_pw2rs 127 12.3 0.021 0.023 5.192 5.212 mp_sum_l 7884 12.9 3.418 4.883 3.418 4.883 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.550 4.675 mp_allgather_i34 2485 14.5 1.589 4.508 1.589 4.508 transfer_fm_to_dbcsr 11 9.9 0.018 0.023 2.617 4.466 mp_alltoall_i22 712 14.1 2.181 4.173 2.181 4.173 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.052 4.079 transfer_rs2pw 519 10.6 0.006 0.007 3.454 3.730 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.724 3.725 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=186.243000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=926.272727, yerr=11.740303 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1973537472512 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.243766E+12 0.0% 0.0% 100.0% flops max/rank 4.387745E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806547424 0.0% 0.0% 100.0% number of processed stacks 6032640 0.0% 0.0% 100.0% average stack size 0.0 0.0 1128.3 marketing flops 145.662591E+12 ------------------------------------------------------------------------------- # multiplications 2534 max memory usage/rank 1.154216E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1155504 MPI messages size (bytes): total size 2.039825E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.765312E+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 323336 36533174272 131072 < size <= 4194304 723016 792311627776 4194304 < size <= 16777216 70800 669922351760 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57626. MP_Allreduce 11184 1079. MP_Sync 88 MP_Alltoall 1724 12509574. MP_SendRecv 5934 75008. MP_ISendRecv 5934 75008. MP_Wait 22612 MP_ISend 15064 244788. MP_IRecv 15064 244788. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.083 0.138 209.785 209.786 qs_mol_dyn_low 1 2.0 0.003 0.003 209.001 209.044 qs_forces 11 3.9 0.004 0.004 208.883 208.891 qs_energies 11 4.9 0.008 0.011 201.606 201.614 scf_env_do_scf 11 5.9 0.001 0.001 183.855 183.865 velocity_verlet 10 3.0 0.008 0.013 138.687 138.690 scf_env_do_scf_inner_loop 118 6.6 0.004 0.045 132.406 132.408 qs_scf_new_mos 118 7.6 0.001 0.001 94.143 94.508 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.142 94.508 ot_scf_mini 118 9.6 0.003 0.003 89.258 89.597 dbcsr_multiply_generic 2534 12.6 0.191 0.197 85.935 86.882 multiply_cannon 2534 13.6 0.553 0.586 57.532 60.457 multiply_cannon_loop 2534 14.6 1.864 1.950 52.761 55.263 init_scf_loop 11 6.9 0.000 0.000 51.312 51.313 ot_mini 118 10.6 0.001 0.001 47.006 47.355 prepare_preconditioner 11 7.9 0.000 0.000 47.160 47.185 make_preconditioner 11 8.9 0.000 0.000 47.160 47.185 make_full_inverse_cholesky 11 9.9 0.011 0.023 40.630 45.757 multiply_cannon_multrec 30408 15.6 13.887 18.689 26.967 31.406 rebuild_ks_matrix 129 8.3 0.001 0.001 30.337 30.691 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.019 30.336 30.691 mp_waitall_1 149452 16.7 18.658 30.479 18.658 30.479 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.395 27.713 qs_ot_get_p 129 10.4 0.001 0.001 27.319 27.695 qs_ot_get_derivative 118 11.6 0.001 0.002 26.728 27.072 make_m2s 5068 13.6 0.096 0.099 23.102 24.548 make_images 5068 14.6 2.001 2.752 22.791 24.241 qs_ot_p2m_diag 84 11.4 0.347 0.396 22.346 22.419 cp_dbcsr_syevd 84 12.4 0.005 0.006 20.727 20.730 ot_diis_step 118 11.6 0.018 0.019 20.119 20.122 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.598 20.102 apply_single 129 13.6 0.001 0.001 19.597 20.102 cp_fm_cholesky_invert 11 10.9 19.145 19.154 19.145 19.154 cp_fm_upper_to_full 106 14.8 11.772 17.317 11.772 17.317 cp_fm_diag_elpa 84 13.4 0.002 0.010 16.896 16.915 cp_fm_diag_elpa_base 84 14.4 15.858 16.224 16.879 16.896 multiply_cannon_metrocomm3 30408 15.6 0.050 0.054 6.317 15.909 qs_rho_update_rho_low 129 7.7 0.001 0.001 14.896 14.953 calculate_rho_elec 129 8.7 0.174 0.190 14.896 14.953 sum_up_and_integrate 129 10.3 0.002 0.003 14.903 14.929 integrate_v_rspace 129 11.3 0.003 0.004 14.843 14.870 make_images_data 5068 15.6 0.067 0.074 11.912 13.772 dbcsr_complete_redistribute 397 12.7 1.497 1.638 9.589 13.493 dbcsr_mm_accdrv_process 62840 16.2 8.527 9.457 12.643 13.183 hybrid_alltoall_any 5255 16.5 0.538 2.232 10.497 12.956 copy_fm_to_dbcsr 210 11.7 0.001 0.002 8.039 11.941 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 11.637 11.888 init_scf_run 11 5.9 0.000 0.001 11.666 11.667 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.666 11.667 multiply_cannon_sync_h2d 30408 15.6 10.522 11.014 10.522 11.014 transfer_fm_to_dbcsr 11 9.9 0.001 0.008 6.508 10.291 mp_alltoall_i22 720 14.1 5.845 9.809 5.845 9.809 pw_transfer 1559 11.6 0.087 0.102 9.231 9.346 cp_fm_cholesky_decompose 22 10.9 9.049 9.141 9.049 9.141 fft_wrap_pw1pw2 1301 12.7 0.010 0.012 9.002 9.123 wfi_extrapolate 11 7.9 0.001 0.001 8.233 8.233 density_rs2pw 129 9.7 0.006 0.006 7.657 8.016 grid_integrate_task_list 129 12.3 7.580 7.960 7.580 7.960 fft_wrap_pw1pw2_140 527 13.2 1.456 1.502 7.814 7.944 multiply_cannon_metrocomm4 25340 15.6 0.086 0.096 2.815 7.161 mp_alltoall_d11v 2429 14.1 5.844 6.947 5.844 6.947 mp_irecv_dv 76866 16.2 2.659 6.875 2.659 6.875 calculate_dm_sparse 129 9.5 0.001 0.001 6.639 6.722 fft3d_ps 1301 14.7 2.860 2.914 6.159 6.247 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.699 5.774 grid_collocate_task_list 129 9.7 5.331 5.710 5.331 5.710 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.074 5.164 potential_pw2rs 129 12.3 0.023 0.023 5.026 5.070 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.986 5.056 mp_allgather_i34 2534 14.6 2.830 4.773 2.830 4.773 mp_sum_l 8031 12.9 3.378 4.646 3.378 4.646 qs_energies_init_hamiltonians 11 5.9 0.010 0.021 4.522 4.522 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.396 4.516 make_images_sizes 5068 15.6 0.007 0.007 2.673 4.323 mp_alltoall_i44 5068 16.6 2.667 4.316 2.667 4.316 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=209.786000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1087.090909, yerr=18.889413 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 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 1.582162E+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 716108490000 ------------------------------------------------------------------------------- - - - 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 3981 58503. MP_Allreduce 11030 1168. MP_Sync 86 MP_Alltoall 1700 18828162. 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 0.061 0.082 190.292 190.301 qs_mol_dyn_low 1 2.0 0.003 0.003 189.670 189.692 qs_forces 11 3.9 0.004 0.004 189.526 189.530 qs_energies 11 4.9 0.002 0.002 181.918 181.923 scf_env_do_scf 11 5.9 0.001 0.001 163.703 163.716 velocity_verlet 10 3.0 0.010 0.014 123.930 123.937 scf_env_do_scf_inner_loop 116 6.6 0.004 0.031 121.513 121.515 qs_scf_new_mos 116 7.6 0.001 0.001 83.421 83.540 qs_scf_loop_do_ot 116 8.6 0.001 0.001 83.421 83.540 ot_scf_mini 116 9.6 0.003 0.004 78.920 78.990 dbcsr_multiply_generic 2485 12.5 0.187 0.202 74.211 74.644 multiply_cannon 2485 13.5 0.553 0.578 54.053 56.902 multiply_cannon_loop 2485 14.5 0.809 0.843 50.610 51.570 init_scf_loop 11 6.9 0.000 0.000 42.024 42.027 ot_mini 116 10.6 0.001 0.001 39.765 39.875 prepare_preconditioner 11 7.9 0.000 0.000 38.102 38.126 make_preconditioner 11 8.9 0.000 0.000 38.102 38.126 make_full_inverse_cholesky 11 9.9 0.017 0.025 35.804 36.165 mp_waitall_1 124680 16.7 26.118 33.143 26.118 33.143 rebuild_ks_matrix 127 8.3 0.001 0.001 29.707 29.807 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 29.706 29.806 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.001 27.096 qs_ot_get_p 127 10.4 0.001 0.001 26.251 26.365 qs_ot_get_derivative 116 11.6 0.001 0.002 22.594 22.669 cp_fm_cholesky_invert 11 10.9 22.386 22.394 22.386 22.394 qs_ot_p2m_diag 82 11.4 0.490 0.495 21.882 21.903 multiply_cannon_multrec 9940 15.5 10.317 14.536 17.878 20.883 cp_dbcsr_syevd 82 12.4 0.005 0.006 20.522 20.524 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.719 20.005 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.179 17.300 apply_single 127 13.6 0.001 0.001 17.178 17.300 ot_diis_step 116 11.6 0.020 0.020 17.064 17.064 cp_fm_diag_elpa 82 13.4 0.001 0.002 16.954 16.962 cp_fm_diag_elpa_base 82 14.4 16.677 16.787 16.947 16.955 make_m2s 4970 13.5 0.066 0.072 15.774 16.762 make_images 4970 14.5 2.166 2.575 15.469 16.457 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.999 15.033 calculate_rho_elec 127 8.7 0.254 0.265 14.999 15.033 sum_up_and_integrate 127 10.3 0.002 0.002 14.901 14.946 integrate_v_rspace 127 11.3 0.004 0.004 14.841 14.888 init_scf_run 11 5.9 0.000 0.001 11.224 11.225 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.224 11.225 multiply_cannon_sync_h2d 9940 15.5 10.685 11.044 10.685 11.044 make_images_data 4970 15.5 0.055 0.064 8.976 10.595 hybrid_alltoall_any 5155 16.4 0.837 3.751 8.652 10.047 cp_fm_cholesky_decompose 22 10.9 9.772 9.936 9.772 9.936 pw_transfer 1535 11.6 0.085 0.093 9.731 9.776 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.510 9.561 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.973 9.026 fft_wrap_pw1pw2_140 519 13.2 1.792 1.827 8.214 8.270 grid_integrate_task_list 127 12.3 7.719 8.102 7.719 8.102 wfi_extrapolate 11 7.9 0.001 0.001 7.963 7.963 dbcsr_mm_accdrv_process 20590 16.1 3.221 4.423 7.219 7.901 density_rs2pw 127 9.7 0.005 0.006 7.525 7.876 multiply_cannon_metrocomm1 9940 15.5 0.030 0.030 4.740 7.427 mp_alltoall_d11v 2401 14.1 5.468 6.308 5.468 6.308 fft3d_ps 1281 14.7 2.741 2.809 6.264 6.303 calculate_dm_sparse 127 9.5 0.001 0.001 6.183 6.259 dbcsr_complete_redistribute 393 12.7 2.102 2.209 5.824 6.170 grid_collocate_task_list 127 9.7 5.519 5.810 5.519 5.810 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.304 5.332 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.263 5.264 mp_allgather_i34 2485 14.5 1.638 5.220 1.638 5.220 potential_pw2rs 127 12.3 0.026 0.028 4.964 4.985 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.766 4.153 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.072 4.112 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.030 4.106 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.883 3.924 multiply_cannon_metrocomm4 7455 15.5 0.027 0.029 1.814 3.923 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.579 3.872 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.761 3.866 mp_irecv_dv 28618 15.9 1.775 3.848 1.775 3.848 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=190.301000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1480.363636, yerr=58.703063 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 3.228287E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+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 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58586. MP_Allreduce 11179 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.080 311.083 311.102 qs_mol_dyn_low 1 2.0 0.003 0.003 310.385 310.408 qs_forces 11 3.9 0.004 0.004 310.253 310.257 qs_energies 11 4.9 0.002 0.002 301.196 301.215 scf_env_do_scf 11 5.9 0.001 0.001 277.562 277.569 velocity_verlet 10 3.0 0.027 0.031 223.810 223.818 scf_env_do_scf_inner_loop 118 6.6 0.004 0.019 146.184 146.187 init_scf_loop 11 6.9 0.000 0.000 131.073 131.077 prepare_preconditioner 11 7.9 0.000 0.000 125.998 126.028 make_preconditioner 11 8.9 0.000 0.000 125.998 126.028 make_full_inverse_cholesky 11 9.9 0.038 0.039 101.268 123.120 qs_scf_new_mos 118 7.6 0.001 0.001 99.990 100.095 qs_scf_loop_do_ot 118 8.6 0.001 0.001 99.989 100.094 ot_scf_mini 118 9.6 0.004 0.004 94.897 95.006 dbcsr_multiply_generic 2535 12.6 0.245 0.258 86.275 86.548 cp_fm_upper_to_full 106 14.8 53.727 77.497 53.727 77.497 multiply_cannon 2535 13.6 0.689 0.781 61.483 64.285 multiply_cannon_loop 2535 14.6 1.067 1.089 57.174 57.735 ot_mini 118 10.6 0.001 0.001 47.513 47.678 dbcsr_complete_redistribute 397 12.7 3.999 4.053 31.026 44.430 copy_fm_to_dbcsr 210 11.7 0.001 0.002 27.371 40.745 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.687 37.924 cp_fm_cholesky_invert 11 10.9 36.889 36.896 36.889 36.896 rebuild_ks_matrix 129 8.3 0.001 0.001 35.703 35.891 qs_ks_build_kohn_sham_matrix 129 9.3 0.026 0.028 35.702 35.891 mp_alltoall_i22 720 14.1 22.479 35.771 22.479 35.771 qs_ks_update_qs_env 129 7.6 0.001 0.001 33.121 33.293 mp_waitall_1 104820 16.8 30.299 33.286 30.299 33.286 qs_ot_get_p 129 10.4 0.001 0.001 31.714 31.926 qs_ot_get_derivative 118 11.6 0.002 0.002 27.024 27.131 qs_ot_p2m_diag 84 11.4 0.891 0.896 26.992 27.013 cp_dbcsr_syevd 84 12.4 0.006 0.006 25.046 25.050 cp_fm_diag_elpa 84 13.4 0.008 0.016 21.354 21.355 cp_fm_diag_elpa_base 84 14.4 16.694 18.343 21.333 21.334 multiply_cannon_metrocomm3 10140 15.6 0.025 0.026 20.542 21.065 make_m2s 5070 13.6 0.078 0.079 19.389 20.701 ot_diis_step 118 11.6 0.023 0.024 20.393 20.393 make_images 5070 14.6 3.104 3.281 18.900 20.213 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.407 19.845 apply_single 129 13.6 0.001 0.001 19.406 19.845 multiply_cannon_multrec 10140 15.6 10.456 12.348 18.492 18.568 qs_rho_update_rho_low 129 7.7 0.001 0.001 17.891 17.917 calculate_rho_elec 129 8.7 0.485 0.485 17.890 17.917 sum_up_and_integrate 129 10.3 0.002 0.002 17.294 17.383 integrate_v_rspace 129 11.3 0.004 0.006 17.231 17.320 multiply_cannon_sync_h2d 10140 15.6 14.424 14.460 14.424 14.460 init_scf_run 11 5.9 0.000 0.001 13.237 13.238 scf_env_initial_rho_setup 11 6.9 0.000 0.000 13.237 13.238 hybrid_alltoall_any 5257 16.5 1.324 3.121 10.821 12.849 pw_transfer 1559 11.6 0.095 0.096 12.672 12.692 make_images_data 5070 15.6 0.064 0.070 10.743 12.675 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 12.432 12.451 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.916 10.995 fft_wrap_pw1pw2_140 527 13.2 3.069 3.098 10.921 10.936 cp_fm_cholesky_decompose 22 10.9 10.709 10.741 10.709 10.741 dbcsr_mm_accdrv_process 20958 16.1 4.297 6.241 7.782 9.684 wfi_extrapolate 11 7.9 0.001 0.001 9.506 9.506 grid_integrate_task_list 129 12.3 8.633 8.804 8.633 8.804 density_rs2pw 129 9.7 0.005 0.005 8.616 8.668 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.149 8.150 mp_alltoall_d11v 2429 14.1 7.902 8.070 7.902 8.070 fft3d_ps 1301 14.7 2.877 2.888 7.642 7.682 calculate_dm_sparse 129 9.5 0.001 0.001 7.038 7.094 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.727 6.765 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.398 6.557 grid_collocate_task_list 129 9.7 6.474 6.544 6.474 6.544 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=311.102000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2917.272727, yerr=162.181767 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 1.259921E+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 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.026 0.057 87.594 87.596 qs_energies 1 2.0 0.000 0.000 86.947 86.970 ls_scf 1 3.0 0.002 0.015 86.041 86.058 dbcsr_multiply_generic 111 6.7 0.014 0.015 74.141 74.354 multiply_cannon 111 7.7 0.017 0.020 57.456 58.552 multiply_cannon_loop 111 8.7 0.225 0.244 53.955 55.351 ls_scf_main 1 4.0 0.001 0.017 52.618 52.627 density_matrix_trs4 2 5.0 0.003 0.020 46.964 47.062 ls_scf_init_scf 1 4.0 0.003 0.021 30.301 30.309 ls_scf_init_matrix_S 1 5.0 0.002 0.013 29.012 29.070 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.782 26.808 mp_waitall_1 11031 10.9 22.107 25.813 22.107 25.813 multiply_cannon_multrec 2664 9.7 8.190 8.941 15.692 17.503 multiply_cannon_sync_h2d 2664 9.7 13.796 15.742 13.796 15.742 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.430 13.641 make_m2s 222 7.7 0.009 0.010 12.986 13.553 make_images 222 8.7 0.099 0.109 12.964 13.532 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.458 8.811 dbcsr_mm_accdrv_process 4760 10.4 0.594 0.715 7.117 8.131 make_images_data 222 9.7 0.004 0.005 7.535 8.031 hybrid_alltoall_any 227 10.6 0.216 1.835 6.463 7.569 dbcsr_mm_accdrv_process_sort 4760 11.4 6.322 7.261 6.322 7.261 calculate_norms 4752 9.8 5.571 6.316 5.571 6.316 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.000 5.250 mp_sum_l 887 5.1 3.193 4.397 3.193 4.397 arnoldi_extremal 4 6.8 0.002 0.012 3.609 3.647 arnoldi_normal_ev 4 7.8 0.008 0.060 3.607 3.638 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.041 3.544 mp_irecv_dv 6231 10.9 2.024 3.519 2.024 3.519 build_subspace 16 8.4 0.009 0.015 3.388 3.396 make_images_sizes 222 9.7 0.000 0.000 0.720 3.229 mp_alltoall_i44 222 10.7 0.720 3.228 0.720 3.228 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.323 3.151 ls_scf_post 1 4.0 0.001 0.008 3.120 3.137 ls_scf_store_result 1 5.0 0.000 0.000 2.846 2.892 dbcsr_special_finalize 555 9.7 0.005 0.007 2.380 2.771 dbcsr_merge_single_wm 555 10.7 0.478 0.651 2.371 2.763 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.430 2.670 make_images_pack 222 9.7 2.203 2.606 2.205 2.607 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.580 2.584 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.415 2.566 dbcsr_matrix_vector_mult_local 304 10.0 2.069 2.480 2.071 2.482 dbcsr_sort_data 658 11.4 2.152 2.463 2.152 2.463 buffer_matrices_ensure_size 222 8.7 1.782 2.219 1.782 2.219 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.811 1.820 rebuild_ks_matrix 3 7.3 0.000 0.000 1.802 1.811 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.015 1.802 1.811 acc_transpose_blocks 2664 9.7 0.018 0.020 1.631 1.788 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=87.596000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 2.132779E+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 266673. 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 0.046 0.123 92.049 92.050 qs_energies 1 2.0 0.002 0.010 91.278 91.312 ls_scf 1 3.0 0.003 0.014 89.933 89.948 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.230 75.540 multiply_cannon 111 7.7 0.027 0.039 54.047 58.322 multiply_cannon_loop 111 8.7 0.137 0.166 51.249 54.342 ls_scf_main 1 4.0 0.002 0.018 54.196 54.210 density_matrix_trs4 2 5.0 0.003 0.018 48.483 48.707 ls_scf_init_scf 1 4.0 0.001 0.002 32.066 32.078 ls_scf_init_matrix_S 1 5.0 0.005 0.021 30.579 30.661 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.206 28.215 mp_waitall_1 9105 10.9 19.896 27.792 19.896 27.792 multiply_cannon_multrec 1332 9.7 13.346 16.501 22.742 27.114 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.045 19.849 make_m2s 222 7.7 0.006 0.008 14.500 15.023 make_images 222 8.7 1.370 1.694 14.469 14.993 dbcsr_mm_accdrv_process 4041 10.4 0.361 0.649 8.990 10.482 dbcsr_mm_accdrv_process_sort 4041 11.4 8.473 9.954 8.473 9.954 make_images_data 222 9.7 0.004 0.005 8.140 8.951 hybrid_alltoall_any 227 10.6 0.546 2.584 7.545 8.821 mp_sum_l 887 5.1 5.274 7.987 5.274 7.987 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.183 7.790 mp_irecv_dv 3311 11.0 3.163 7.735 3.163 7.735 calculate_norms 2376 9.8 6.102 6.723 6.102 6.723 multiply_cannon_sync_h2d 1332 9.7 4.960 6.266 4.960 6.266 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.936 6.261 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.928 5.147 arnoldi_extremal 4 6.8 0.000 0.000 5.070 5.087 arnoldi_normal_ev 4 7.8 0.001 0.005 5.070 5.087 build_subspace 16 8.4 0.014 0.021 4.819 4.822 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.353 4.538 ls_scf_post 1 4.0 0.002 0.007 3.667 3.683 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.343 3.577 ls_scf_store_result 1 5.0 0.000 0.000 3.372 3.487 dbcsr_matrix_vector_mult_local 304 10.0 2.741 3.228 2.743 3.229 mp_allgather_i34 111 8.7 0.782 2.760 0.782 2.760 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.595 2.670 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.579 2.584 dbcsr_data_new 4174 10.1 2.115 2.399 2.115 2.399 make_images_pack 222 9.7 1.818 2.131 1.821 2.133 dbcsr_sort_data 436 11.2 1.839 2.074 1.839 2.074 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.883 1.897 rebuild_ks_matrix 3 7.3 0.000 0.000 1.870 1.884 qs_ks_build_kohn_sham_matrix 3 8.3 0.005 0.018 1.870 1.883 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=92.050000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 2.901328E+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 265448. 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 0.041 0.062 95.455 95.455 qs_energies 1 2.0 0.000 0.000 94.748 94.753 ls_scf 1 3.0 0.000 0.000 93.330 93.337 dbcsr_multiply_generic 111 6.7 0.016 0.016 77.409 77.751 multiply_cannon 111 7.7 0.030 0.041 54.245 58.409 ls_scf_main 1 4.0 0.000 0.000 56.867 56.872 multiply_cannon_loop 111 8.7 0.117 0.138 51.515 55.678 density_matrix_trs4 2 5.0 0.002 0.003 50.841 50.997 mp_waitall_1 7281 11.0 23.745 33.267 23.745 33.267 ls_scf_init_scf 1 4.0 0.004 0.007 32.710 32.713 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.506 31.596 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.056 29.070 multiply_cannon_multrec 888 9.7 12.597 15.136 21.247 24.406 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.367 23.282 make_m2s 222 7.7 0.006 0.007 16.228 16.890 make_images 222 8.7 1.586 1.859 16.190 16.853 make_images_data 222 9.7 0.004 0.004 9.166 10.380 hybrid_alltoall_any 227 10.6 0.642 2.959 8.818 9.964 dbcsr_mm_accdrv_process 3754 10.4 0.320 0.511 8.180 9.436 dbcsr_mm_accdrv_process_sort 3754 11.4 7.742 8.926 7.742 8.926 mp_sum_l 887 5.1 5.237 8.851 5.237 8.851 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.849 8.063 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.518 7.327 mp_irecv_dv 2335 11.1 2.503 7.270 2.503 7.270 multiply_cannon_sync_h2d 888 9.7 5.992 7.069 5.992 7.069 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.026 7.049 arnoldi_extremal 4 6.8 0.000 0.000 5.427 5.439 arnoldi_normal_ev 4 7.8 0.002 0.005 5.427 5.439 build_subspace 16 8.4 0.014 0.020 5.128 5.132 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.822 5.074 calculate_norms 1584 9.8 4.368 4.743 4.368 4.743 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.566 3.849 ls_scf_post 1 4.0 0.000 0.000 3.752 3.756 dbcsr_matrix_vector_mult_local 304 10.0 3.023 3.598 3.025 3.599 ls_scf_store_result 1 5.0 0.000 0.000 3.478 3.577 mp_allgather_i34 111 8.7 0.837 3.530 0.837 3.530 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.905 3.022 compute_matrix_preconditioner 1 6.0 0.000 0.001 2.750 2.755 make_images_sizes 222 9.7 0.000 0.000 1.262 2.547 mp_alltoall_i44 222 10.7 1.261 2.547 1.261 2.547 dbcsr_data_new 4116 9.9 2.108 2.447 2.108 2.447 dbcsr_sort_data 325 11.1 1.852 2.143 1.852 2.143 acc_transpose_blocks 888 9.7 0.005 0.006 1.918 2.103 acc_transpose_blocks_kernels 888 10.7 0.013 0.015 1.854 2.036 jit_kernel_transpose 1 13.0 1.841 2.022 1.841 2.022 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.455000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2215.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 3.346133E+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 266673. 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 0.030 0.093 100.009 100.009 qs_energies 1 2.0 0.000 0.000 99.192 99.197 ls_scf 1 3.0 0.000 0.000 97.537 97.543 dbcsr_multiply_generic 111 6.7 0.017 0.018 80.889 81.197 ls_scf_main 1 4.0 0.000 0.000 59.346 59.359 multiply_cannon 111 7.7 0.039 0.096 53.416 57.780 density_matrix_trs4 2 5.0 0.002 0.003 53.188 53.342 multiply_cannon_loop 111 8.7 0.154 0.170 48.239 52.164 ls_scf_init_scf 1 4.0 0.007 0.012 34.927 34.937 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.679 33.750 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 31.042 31.056 mp_waitall_1 6369 11.0 22.516 30.000 22.516 30.000 multiply_cannon_multrec 1332 9.7 14.330 17.568 22.320 24.816 make_m2s 222 7.7 0.007 0.008 21.031 22.518 make_images 222 8.7 3.142 3.608 20.980 22.470 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.046 16.770 make_images_data 222 9.7 0.004 0.004 11.678 13.388 hybrid_alltoall_any 227 10.6 0.795 3.735 10.940 12.784 mp_sum_l 887 5.1 4.898 9.157 4.898 9.157 dbcsr_mm_accdrv_process 3641 10.4 0.302 0.480 7.637 9.153 dbcsr_mm_accdrv_process_sort 3641 11.4 7.194 8.665 7.194 8.665 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.765 7.375 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.115 6.157 mp_irecv_dv 3229 10.9 2.089 6.073 2.089 6.073 multiply_cannon_sync_h2d 1332 9.7 5.418 5.879 5.418 5.879 arnoldi_extremal 4 6.8 0.000 0.000 5.372 5.392 arnoldi_normal_ev 4 7.8 0.001 0.005 5.372 5.392 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.621 5.212 build_subspace 16 8.4 0.014 0.021 5.035 5.042 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.643 4.847 mp_allgather_i34 111 8.7 2.290 4.832 2.290 4.832 calculate_norms 2376 9.8 4.183 4.480 4.183 4.480 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.654 3.939 dbcsr_matrix_vector_mult_local 304 10.0 3.201 3.701 3.203 3.703 dbcsr_sort_data 658 11.4 3.005 3.436 3.005 3.436 ls_scf_post 1 4.0 0.000 0.000 3.264 3.268 dbcsr_special_finalize 555 9.7 0.006 0.007 2.762 3.203 dbcsr_merge_single_wm 555 10.7 0.536 0.659 2.753 3.194 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.997 3.081 ls_scf_store_result 1 5.0 0.000 0.000 2.982 3.064 compute_matrix_preconditioner 1 6.0 0.001 0.001 2.555 2.562 dbcsr_data_release 10477 10.7 1.578 2.438 1.578 2.438 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=100.009000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2756.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 4.761723E+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 265536. 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 0.052 0.094 93.870 93.870 qs_energies 1 2.0 0.000 0.000 92.868 92.873 ls_scf 1 3.0 0.000 0.000 90.916 90.921 dbcsr_multiply_generic 111 6.7 0.017 0.018 72.326 72.514 multiply_cannon 111 7.7 0.086 0.179 54.710 58.241 ls_scf_main 1 4.0 0.000 0.000 56.301 56.308 multiply_cannon_loop 111 8.7 0.093 0.118 52.184 53.530 density_matrix_trs4 2 5.0 0.002 0.003 49.319 49.383 ls_scf_init_scf 1 4.0 0.000 0.001 31.203 31.206 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.955 29.983 mp_waitall_1 5436 11.0 24.008 28.285 24.008 28.285 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.862 27.875 multiply_cannon_multrec 444 9.7 13.704 16.353 20.869 21.952 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.436 15.830 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.633 13.985 make_m2s 222 7.7 0.005 0.005 12.971 13.830 make_images 222 8.7 2.042 2.476 12.904 13.760 make_images_data 222 9.7 0.003 0.004 7.762 9.122 hybrid_alltoall_any 227 10.6 0.806 3.831 7.744 8.890 multiply_cannon_sync_h2d 444 9.7 6.768 8.132 6.768 8.132 dbcsr_mm_accdrv_process 3003 10.4 0.363 0.406 6.854 7.992 dbcsr_mm_accdrv_process_sort 3003 11.4 6.491 7.593 6.491 7.593 arnoldi_extremal 4 6.8 0.000 0.000 6.002 6.017 arnoldi_normal_ev 4 7.8 0.002 0.005 6.002 6.017 build_subspace 16 8.4 0.015 0.019 5.605 5.619 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.431 4.615 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.252 4.447 mp_sum_l 887 5.1 2.662 4.219 2.662 4.219 dbcsr_matrix_vector_mult_local 304 10.0 3.713 4.180 3.715 4.182 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.570 3.872 mp_irecv_dv 1241 11.2 1.555 3.849 1.555 3.849 calculate_norms 792 9.8 3.635 3.727 3.635 3.727 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.532 3.603 ls_scf_post 1 4.0 0.000 0.000 3.412 3.417 mp_allgather_i34 111 8.7 1.084 3.335 1.084 3.335 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.827 3.232 ls_scf_store_result 1 5.0 0.000 0.000 3.200 3.227 make_images_sizes 222 9.7 0.000 0.000 0.816 2.947 mp_alltoall_i44 222 10.7 0.816 2.947 0.816 2.947 compute_matrix_preconditioner 1 6.0 0.001 0.002 2.582 2.592 dbcsr_finalize 304 7.8 0.062 0.077 2.198 2.303 dbcsr_data_new 4608 9.7 1.797 2.258 1.797 2.258 acc_transpose_blocks 444 9.7 0.003 0.003 2.123 2.162 dbcsr_merge_all 275 8.9 0.481 0.524 2.056 2.134 acc_transpose_blocks_kernels 444 10.7 0.007 0.007 2.053 2.056 jit_kernel_transpose 1 13.0 2.046 2.049 2.046 2.049 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.031 2.034 rebuild_ks_matrix 3 7.3 0.000 0.000 1.999 2.002 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.010 1.999 2.002 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 1.936 1.936 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=93.870000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3683.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 8.875917E+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 284089. 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 0.110 0.150 109.497 109.497 qs_energies 1 2.0 0.000 0.000 107.947 107.949 ls_scf 1 3.0 0.000 0.000 104.969 104.972 dbcsr_multiply_generic 111 6.7 0.024 0.028 78.540 78.683 ls_scf_main 1 4.0 0.000 0.000 63.910 63.918 density_matrix_trs4 2 5.0 0.002 0.003 54.954 55.010 multiply_cannon 111 7.7 0.155 0.233 50.590 52.691 multiply_cannon_loop 111 8.7 0.098 0.100 47.470 49.304 ls_scf_init_scf 1 4.0 0.011 0.013 37.277 37.279 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.756 35.780 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 33.040 33.048 mp_waitall_1 4527 11.1 22.113 25.854 22.113 25.854 make_m2s 222 7.7 0.005 0.005 22.725 23.919 make_images 222 8.7 3.577 3.871 22.618 23.810 multiply_cannon_multrec 444 9.7 17.877 18.457 22.585 23.093 hybrid_alltoall_any 227 10.6 1.653 3.626 12.858 15.714 make_images_data 222 9.7 0.003 0.004 13.109 15.342 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.400 12.372 multiply_cannon_sync_h2d 444 9.7 8.790 8.837 8.790 8.837 arnoldi_extremal 4 6.8 0.000 0.000 7.487 7.495 arnoldi_normal_ev 4 7.8 0.014 0.021 7.487 7.495 build_subspace 16 8.4 0.026 0.036 6.911 6.921 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.527 5.678 dbcsr_matrix_vector_mult_local 304 10.0 5.086 5.405 5.088 5.408 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.093 5.189 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.849 5.119 dbcsr_mm_accdrv_process 1814 10.4 0.320 0.373 4.521 4.656 compute_matrix_preconditioner 1 6.0 0.002 0.002 4.343 4.350 dbcsr_mm_accdrv_process_sort 1814 11.4 4.177 4.309 4.177 4.309 mp_allgather_i34 111 8.7 1.221 3.915 1.221 3.915 ls_scf_post 1 4.0 0.000 0.000 3.782 3.785 mp_sum_l 887 5.1 2.830 3.708 2.830 3.708 make_images_sizes 222 9.7 0.000 0.000 1.475 3.625 mp_alltoall_i44 222 10.7 1.475 3.624 1.475 3.624 ls_scf_store_result 1 5.0 0.000 0.000 3.497 3.533 calculate_norms 792 9.8 3.230 3.277 3.230 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.090 3.165 qs_energies_init_hamiltonians 1 3.0 0.011 0.024 2.948 2.948 dbcsr_merge_all 275 8.9 0.886 0.917 2.859 2.943 dbcsr_complete_redistribute 5 7.6 1.427 1.466 2.735 2.860 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.223 2.712 matrix_ls_to_qs 2 6.0 0.000 0.000 2.384 2.530 dbcsr_sort_data 325 11.1 2.439 2.497 2.439 2.497 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.411 2.414 rebuild_ks_matrix 3 7.3 0.000 0.000 2.345 2.348 qs_ks_build_kohn_sham_matrix 3 8.3 0.009 0.011 2.345 2.348 dbcsr_data_new 6591 9.6 1.836 2.345 1.836 2.345 dbcsr_new_transposed 4 7.5 0.248 0.265 2.299 2.312 dbcsr_frobenius_norm 74 6.6 2.055 2.127 2.206 2.236 dbcsr_data_release 12724 10.6 1.999 2.216 1.999 2.216 dbcsr_add_d 103 6.2 0.000 0.000 2.145 2.202 dbcsr_add_anytype 103 7.2 0.859 0.890 2.145 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=109.497000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7007.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/7e52d8fe64174c63d4297b1f87b53e7f026adc7a_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 588.619776E+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 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4806047. 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 0.030 0.094 307.942 307.943 qs_mol_dyn_low 1 2.0 0.003 0.027 306.603 306.641 qs_forces 5 3.8 0.016 0.023 306.514 306.516 qs_energies 5 4.8 0.020 0.178 302.886 302.926 scf_env_do_scf 5 5.8 0.002 0.042 285.489 285.492 scf_env_do_scf_inner_loop 105 6.6 0.004 0.074 248.165 248.168 qs_scf_new_mos 105 7.6 0.000 0.001 195.321 195.580 qs_scf_loop_do_ot 105 8.6 0.001 0.001 195.320 195.579 ot_scf_mini 105 9.6 0.003 0.005 184.905 185.107 dbcsr_multiply_generic 1445 12.2 0.139 0.146 141.235 142.270 velocity_verlet 4 3.0 0.002 0.012 125.261 125.262 multiply_cannon 1445 13.2 0.279 0.289 118.266 120.659 multiply_cannon_loop 1445 14.2 2.871 3.042 115.307 117.928 qs_ot_get_p 112 10.4 0.001 0.001 88.902 89.288 qs_ot_p2m_diag 40 11.0 0.020 0.038 75.482 75.572 cp_dbcsr_syevd 40 12.0 0.002 0.003 71.804 71.809 cp_fm_syevd 40 13.0 0.002 0.010 64.939 65.101 ot_mini 105 10.6 0.001 0.001 64.590 64.714 cp_fm_redistribute_end 40 14.0 29.611 59.076 29.623 59.083 cp_fm_syevd_base 40 14.0 29.438 58.908 29.438 58.908 mp_waitall_1 488190 16.1 42.398 52.377 42.398 52.377 multiply_cannon_multrec 69360 15.2 29.183 34.508 39.099 44.168 qs_ot_get_derivative 55 11.6 0.001 0.001 42.447 42.592 multiply_cannon_metrocomm3 69360 15.2 0.200 0.212 28.137 39.331 init_scf_loop 7 6.6 0.002 0.092 37.254 37.257 rebuild_ks_matrix 110 8.4 0.000 0.000 34.709 34.919 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.080 34.708 34.918 multiply_cannon_sync_h2d 69360 15.2 28.405 32.185 28.405 32.185 qs_ks_update_qs_env 112 7.6 0.001 0.001 31.923 32.113 prepare_preconditioner 7 7.6 0.000 0.001 31.345 31.375 make_preconditioner 7 8.6 0.002 0.109 31.345 31.375 qs_rho_update_rho_low 110 7.6 0.001 0.006 25.506 25.917 calculate_rho_elec 110 8.6 0.029 0.033 25.506 25.916 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.350 23.583 apply_single 62 13.6 0.000 0.000 23.350 23.583 make_full_inverse_cholesky 7 9.6 0.001 0.033 22.227 22.290 ot_new_cg_direction 55 11.6 0.001 0.001 21.351 21.353 density_rs2pw 110 9.6 0.005 0.007 18.905 19.470 qs_ot_get_orbitals 105 10.6 0.001 0.001 15.972 16.295 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 15.258 15.376 transfer_rs2pw 445 10.6 0.007 0.008 14.675 15.228 cp_fm_cholesky_invert 7 10.6 14.851 14.869 14.851 14.869 mp_sum_l 4764 12.2 13.291 14.448 13.291 14.448 init_scf_run 5 5.8 0.000 0.001 14.267 14.268 scf_env_initial_rho_setup 5 6.8 0.001 0.053 14.267 14.268 pw_transfer 1645 12.4 0.082 0.114 13.675 14.095 fft_wrap_pw1pw2 1425 13.5 0.013 0.017 13.534 13.959 sum_up_and_integrate 60 10.3 0.001 0.002 13.233 13.245 integrate_v_rspace 60 11.3 0.001 0.002 13.216 13.228 calculate_dm_sparse 110 9.5 0.000 0.001 12.134 12.379 check_diag 80 13.5 8.597 8.898 11.663 11.846 fft_wrap_pw1pw2_240 915 15.0 1.173 1.285 11.442 11.838 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 11.506 11.595 qs_vxc_create 110 10.4 0.003 0.036 11.272 11.337 dbcsr_mm_accdrv_process 154766 15.8 6.161 6.397 9.788 10.752 fft3d_pb 915 16.0 2.399 2.700 9.584 10.169 multiply_cannon_metrocomm1 69360 15.2 0.096 0.102 5.515 9.951 make_m2s 2890 13.2 0.078 0.085 9.218 9.889 make_images 2890 14.2 0.240 0.261 9.111 9.780 potential_pw2rs 60 12.3 0.003 0.003 8.915 8.983 mp_sendrecv_dv 168740 12.6 8.683 8.953 8.683 8.953 make_full_single_inverse 7 9.6 0.001 0.002 8.718 8.749 acc_transpose_blocks 69360 15.2 0.351 0.369 8.096 8.734 transfer_rs2pw_30 110 11.6 1.247 1.301 7.859 8.492 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.002 8.471 8.488 calculate_first_density_matrix 1 7.0 0.000 0.003 8.370 8.383 xc_rho_set_and_dset_create 110 12.4 0.076 0.098 7.664 7.918 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.735 7.806 xc_vxc_pw_create 60 11.3 0.038 0.050 7.673 7.739 transfer_pw2rs 245 13.2 0.003 0.004 7.590 7.658 mp_alltoall_z22v 2340 17.7 6.729 7.522 6.729 7.522 xc_pw_derive 510 13.4 0.006 0.007 6.892 6.965 cp_fm_cholesky_decompose 14 10.2 6.773 6.782 6.773 6.782 mp_alltoall_d11v 1300 13.8 5.357 6.457 5.357 6.457 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=307.943000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 7e52d8fe64174c63d4297b1f87b53e7f026adc7a Summary: empty Status: OK