=== 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: 8d96446bea90f37d3e25c2492dcc34fe4f653cd4 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2021.11.002, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 5.2.3, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.0, # SIRIUS 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (15.09.2022) # # \ 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 --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2021.11.002 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 5.2.3 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.0 #USE_QUIP := b4336484fb65b0e73211a8f920ae4361c7c353fd USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MPI_VERSION=3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/01 job id: 42129821 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/02 job id: 42129824 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/03 job id: 42129825 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/04 job id: 42129827 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/05 job id: 42129828 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/06 job id: 42129830 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/07 job id: 42129831 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/08 job id: 42129834 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/09 job id: 42129835 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/10 job id: 42129837 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/11 job id: 42129838 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/12 job id: 42129839 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/13 job id: 42129841 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/14 job id: 42129843 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/15 job id: 42129845 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/16 job id: 42129848 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/17 job id: 42129850 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/18 job id: 42129851 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/19 job id: 42129854 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/20 job id: 42129855 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/21 job id: 42129858 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/22 job id: 42129859 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/23 job id: 42129861 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/24 job id: 42129862 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/25 job id: 42129866 --- 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/26 job id: 42129869 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.036 133.549 133.550 farming_run 1 2.0 133.050 133.051 133.515 133.519 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.474322E+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 228 1113141. MP_Allreduce 489 2263609. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 30 829726. MP_ISendRecv 135 235435. MP_Wait 281 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.048 115.696 115.696 qs_energies 1 2.0 0.000 0.000 115.430 115.433 mp2_main 1 3.0 0.000 0.000 113.360 113.362 mp2_gpw_main 1 4.0 0.020 0.026 112.442 112.444 mp2_ri_gpw_compute_in 1 5.0 0.172 0.181 93.598 94.057 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.471 55.931 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.173 41.762 47.114 get_2c_integrals 1 6.0 0.000 0.000 37.375 37.955 integrate_v_rspace 273 8.0 0.437 0.451 25.147 30.334 pw_transfer 6555 10.6 0.372 0.380 27.400 27.983 fft_wrap_pw1pw2 5465 11.4 0.044 0.046 26.053 26.655 grid_integrate_task_list 273 9.0 20.955 26.626 20.955 26.626 fft_wrap_pw1pw2_100 2178 12.4 1.190 1.419 23.614 24.221 compute_2c_integrals 1 7.0 0.003 0.003 19.720 19.720 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.941 19.433 mp2_eri_2c_integrate_gpw 1 9.0 2.405 2.453 18.938 19.432 rpa_ri_compute_en 1 5.0 0.011 0.012 18.734 18.995 cp_fm_cholesky_decompose 12 8.2 17.647 18.245 17.647 18.245 cholesky_decomp 1 7.0 0.000 0.000 16.501 17.117 fft3d_s 5443 13.4 16.183 16.704 16.205 16.724 ao_to_mo_and_store_B_mult_1 272 7.0 10.868 15.592 10.868 15.592 calculate_wavefunction 272 8.0 5.398 5.556 12.525 13.138 rpa_num_int 1 6.0 0.002 0.026 10.623 10.623 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.521 10.549 calc_mat_Q 8 8.0 0.000 0.000 9.324 9.422 contract_S_to_Q 8 9.0 0.000 0.000 8.747 8.848 calc_potential_gpw 544 9.5 0.006 0.007 8.264 8.722 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.002 0.002 8.235 8.582 parallel_gemm_fm 14 9.1 0.000 0.000 8.321 8.416 parallel_gemm_fm_cosma 14 10.1 8.321 8.416 8.321 8.416 potential_pw2rs 545 10.0 0.107 0.109 7.709 8.350 collocate_single_gaussian 272 10.0 0.041 0.043 7.481 7.743 create_integ_mat 1 6.0 0.021 0.027 7.546 7.546 array2fm 1 7.0 0.000 0.000 6.657 6.981 pw_scatter_s 2720 13.7 4.424 4.610 4.424 4.610 pw_gather_s 2722 13.2 3.883 4.340 3.883 4.340 array2fm_buffer_send 1 8.0 2.937 3.144 2.937 3.144 pw_poisson_solve 545 10.5 1.127 1.175 2.183 2.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.444193, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2745.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.042 394.760 394.761 farming_run 1 2.0 393.992 393.998 394.713 394.717 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223000E+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 703 408373. MP_Allreduce 1825 23678. MP_Sync 38 MP_Alltoall 77 MP_SendRecv 2171 2843495. MP_ISendRecv 1739 144022. MP_Wait 2051 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 0.012 0.044 208.644 208.645 qs_energies 1 2.0 0.000 0.000 208.365 208.372 scf_env_do_scf 1 3.0 0.000 0.000 105.481 105.481 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.579 104.587 rebuild_ks_matrix 4 6.0 0.000 0.000 104.578 104.586 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 104.578 104.586 hfx_ks_matrix 4 8.0 0.001 0.001 104.200 104.203 integrate_four_center 4 9.0 0.144 0.461 104.199 104.202 mp2_main 1 3.0 0.004 0.033 102.594 102.600 mp2_gpw_main 1 4.0 0.032 0.049 101.672 101.684 integrate_four_center_main 4 10.0 0.112 0.508 96.378 98.500 integrate_four_center_bin 264 11.0 96.267 98.448 96.267 98.448 init_scf_loop 1 4.0 0.000 0.000 91.308 91.308 mp2_ri_gpw_compute_in 1 5.0 0.064 0.087 74.900 75.936 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.497 55.532 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.162 42.153 47.300 integrate_v_rspace 95 8.0 0.398 0.566 28.505 33.458 pw_transfer 2240 10.6 0.145 0.162 29.931 30.291 ao_to_mo_and_store_B_mult_1 91 7.0 10.653 29.569 10.653 29.569 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.947 29.358 grid_integrate_task_list 95 9.0 23.789 28.957 23.789 28.957 mp2_ri_gpw_compute_en 1 5.0 0.058 0.066 26.594 28.329 fft_wrap_pw1pw2_100 730 12.4 1.276 1.457 26.692 27.175 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.844 1.906 24.913 24.923 get_2c_integrals 1 6.0 0.002 0.014 20.317 20.338 compute_2c_integrals 1 7.0 0.007 0.012 19.291 19.293 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.883 19.146 mp2_eri_2c_integrate_gpw 1 9.0 1.756 1.884 18.882 19.145 fft3d_s 1823 13.4 18.434 18.951 18.447 18.964 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.171 14.171 calculate_wavefunction 91 8.0 2.026 2.059 9.757 9.973 potential_pw2rs 186 10.0 0.033 0.035 8.658 9.175 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.588 8.742 9.135 local_gemm 172 8.0 8.184 8.570 8.184 8.570 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.239 8.552 mp2_ri_gpw_compute_en_comm 22 7.0 0.499 0.514 7.934 8.385 collocate_single_gaussian 91 10.0 0.017 0.020 7.886 8.196 calc_potential_gpw 182 9.5 0.002 0.002 7.898 8.057 mp2_ri_gpw_compute_en_ener 172 7.0 6.347 6.429 6.347 6.429 mp_sendrecv_dm3 2068 8.0 5.966 6.419 5.966 6.419 pw_gather_s 912 13.2 4.945 5.344 4.945 5.344 mp_sync 38 10.4 2.481 4.956 2.481 4.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.674374, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1510.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 450.134016E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 MP_ISendRecv 45220 5520. MP_Wait 60486 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 0.014 0.029 52.478 52.479 qs_mol_dyn_low 1 2.0 0.003 0.003 52.291 52.298 qs_forces 11 3.9 0.002 0.002 52.228 52.229 qs_energies 11 4.9 0.001 0.002 50.747 50.762 scf_env_do_scf 11 5.9 0.000 0.001 44.863 44.863 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 42.659 42.659 dbcsr_multiply_generic 2286 12.5 0.092 0.097 33.521 33.989 qs_scf_new_mos 108 7.5 0.000 0.000 32.349 32.650 qs_scf_loop_do_ot 108 8.5 0.000 0.001 32.348 32.650 ot_scf_mini 108 9.5 0.002 0.002 30.686 30.887 multiply_cannon 2286 13.5 0.180 0.187 25.772 27.334 multiply_cannon_loop 2286 14.5 1.517 1.588 24.997 26.553 velocity_verlet 10 3.0 0.001 0.002 25.677 25.678 ot_mini 108 10.5 0.001 0.001 19.575 19.837 qs_ot_get_derivative 108 11.5 0.001 0.001 16.552 16.764 mp_waitall_1 267858 16.1 8.773 15.231 8.773 15.231 multiply_cannon_metrocomm3 54864 15.5 0.066 0.071 5.910 13.097 multiply_cannon_multrec 54864 15.5 4.183 6.414 7.456 10.756 rebuild_ks_matrix 119 8.3 0.000 0.000 8.070 8.217 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.069 8.217 multiply_cannon_sync_h2d 54864 15.5 5.849 7.668 5.849 7.668 mp_sum_l 7207 12.9 5.601 7.345 5.601 7.345 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.138 7.276 qs_ot_get_p 119 10.4 0.001 0.001 6.476 6.775 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.918 6.038 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.392 5.826 sum_up_and_integrate 119 10.3 0.012 0.014 4.651 4.660 init_scf_run 11 5.9 0.000 0.001 4.656 4.656 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.655 4.656 integrate_v_rspace 119 11.3 0.002 0.002 4.639 4.649 dbcsr_mm_accdrv_process 76910 16.1 1.199 1.841 3.196 4.439 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.204 4.327 calculate_rho_elec 119 8.7 0.011 0.016 4.204 4.326 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.550 3.611 multiply_cannon_metrocomm1 54864 15.5 0.052 0.059 1.821 3.238 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.886 3.072 apply_single 119 13.6 0.000 0.000 2.886 3.072 calculate_dm_sparse 119 9.5 0.000 0.000 2.888 3.009 rs_pw_transfer 974 11.9 0.012 0.013 2.894 2.968 ot_diis_step 108 11.5 0.006 0.006 2.704 2.705 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.664 2.664 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.442 2.529 jit_kernel_multiply 13 15.8 1.932 2.513 1.932 2.513 calculate_first_density_matrix 1 7.0 0.000 0.000 2.479 2.483 density_rs2pw 119 9.7 0.004 0.004 2.307 2.407 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.405 2.406 cp_fm_redistribute_end 50 14.0 2.182 2.379 2.187 2.381 cp_fm_diag_elpa_base 50 14.0 0.193 2.331 0.194 2.343 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.268 2.270 init_scf_loop 11 6.9 0.000 0.000 2.186 2.186 acc_transpose_blocks 54864 15.5 0.226 0.243 1.750 2.163 grid_integrate_task_list 119 12.3 2.023 2.121 2.023 2.121 wfi_extrapolate 11 7.9 0.001 0.001 2.114 2.114 mp_sum_d 4129 12.0 1.359 2.026 1.359 2.026 potential_pw2rs 119 12.3 0.004 0.004 1.997 2.021 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.903 1.949 pw_transfer 1439 11.6 0.051 0.054 1.837 1.914 make_m2s 4572 13.5 0.053 0.055 1.834 1.890 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.763 1.842 make_images 4572 14.5 0.133 0.138 1.752 1.808 fft3d_ps 1201 14.6 0.360 0.462 1.542 1.619 mp_alltoall_d11v 2130 13.8 1.386 1.545 1.386 1.545 mp_waitany 12084 13.8 1.258 1.443 1.258 1.443 fft_wrap_pw1pw2_140 487 13.2 0.079 0.093 1.347 1.425 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.389 1.408 grid_collocate_task_list 119 9.7 1.283 1.342 1.283 1.342 dbcsr_dot_sd 1205 11.9 0.047 0.058 0.770 1.171 prepare_preconditioner 11 7.9 0.000 0.000 1.069 1.098 make_preconditioner 11 8.9 0.000 0.000 1.069 1.098 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.017 1.054 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.479000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=429.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.174144E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 MP_ISendRecv 33558 37093. MP_Wait 40318 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 0.020 0.052 41.591 41.591 qs_mol_dyn_low 1 2.0 0.003 0.003 41.322 41.330 qs_forces 11 3.9 0.002 0.002 41.259 41.261 qs_energies 11 4.9 0.002 0.005 39.507 39.511 scf_env_do_scf 11 5.9 0.001 0.001 34.107 34.108 scf_env_do_scf_inner_loop 108 6.5 0.002 0.010 31.185 31.185 dbcsr_multiply_generic 2286 12.5 0.101 0.104 22.298 22.725 qs_scf_new_mos 108 7.5 0.001 0.001 21.275 21.513 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.274 21.512 ot_scf_mini 108 9.5 0.002 0.003 20.322 20.491 velocity_verlet 10 3.0 0.002 0.002 19.944 19.945 multiply_cannon 2286 13.5 0.209 0.219 16.566 18.369 multiply_cannon_loop 2286 14.5 0.913 0.981 15.282 16.925 ot_mini 108 10.5 0.001 0.001 12.586 12.828 mp_waitall_1 217478 16.2 7.011 12.060 7.011 12.060 qs_ot_get_derivative 108 11.5 0.001 0.001 10.059 10.236 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.309 9.707 multiply_cannon_multrec 27432 15.5 1.938 4.261 5.836 8.630 rebuild_ks_matrix 119 8.3 0.000 0.000 7.671 7.815 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.671 7.815 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.802 6.928 dbcsr_mm_accdrv_process 47894 16.0 3.047 5.153 3.830 5.684 qs_ot_get_p 119 10.4 0.001 0.001 4.721 4.951 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.838 4.679 sum_up_and_integrate 119 10.3 0.024 0.026 4.645 4.652 integrate_v_rspace 119 11.3 0.002 0.002 4.621 4.627 mp_sum_l 7207 12.9 2.302 4.224 2.302 4.224 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.168 4.202 calculate_rho_elec 119 8.7 0.021 0.024 4.168 4.202 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.080 4.168 apply_single 119 13.6 0.000 0.000 3.080 4.168 init_scf_run 11 5.9 0.000 0.001 4.084 4.084 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.084 4.084 rs_pw_transfer 974 11.9 0.010 0.011 3.270 3.391 make_m2s 4572 13.5 0.052 0.054 2.984 3.243 make_images 4572 14.5 0.201 0.238 2.895 3.152 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.992 3.015 init_scf_loop 11 6.9 0.000 0.002 2.884 2.885 multiply_cannon_sync_h2d 27432 15.5 2.135 2.747 2.135 2.747 density_rs2pw 119 9.7 0.004 0.004 2.504 2.591 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.524 2.525 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.431 2.523 ot_diis_step 108 11.5 0.011 0.011 2.463 2.464 potential_pw2rs 119 12.3 0.006 0.006 2.360 2.385 calculate_first_density_matrix 1 7.0 0.000 0.002 2.362 2.364 calculate_dm_sparse 119 9.5 0.000 0.001 2.115 2.184 pw_transfer 1439 11.6 0.063 0.066 2.099 2.154 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.073 2.074 fft_wrap_pw1pw2 1201 12.6 0.007 0.008 2.010 2.066 cp_fm_redistribute_end 50 14.0 1.703 2.037 1.708 2.039 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.999 2.001 cp_fm_diag_elpa_base 50 14.0 0.321 1.958 0.329 1.987 prepare_preconditioner 11 7.9 0.000 0.000 1.897 1.924 make_preconditioner 11 8.9 0.000 0.002 1.897 1.924 jit_kernel_multiply 10 16.2 0.730 1.906 0.730 1.906 grid_integrate_task_list 119 12.3 1.813 1.901 1.813 1.901 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.791 1.850 make_images_data 4572 15.5 0.045 0.051 1.283 1.786 fft3d_ps 1201 14.6 0.503 0.556 1.718 1.768 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.720 1.758 wfi_extrapolate 11 7.9 0.001 0.001 1.651 1.651 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.132 1.608 mp_alltoall_d11v 2130 13.8 1.441 1.599 1.441 1.599 mp_allgather_i34 2286 14.5 0.725 1.526 0.725 1.526 acc_transpose_blocks 27432 15.5 0.107 0.113 1.203 1.514 fft_wrap_pw1pw2_140 487 13.2 0.075 0.081 1.421 1.476 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.260 1.307 grid_collocate_task_list 119 9.7 1.242 1.304 1.242 1.304 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.288 1.297 make_images_sizes 4572 15.5 0.005 0.005 0.884 1.196 mp_alltoall_i44 4572 16.5 0.879 1.191 0.879 1.191 mp_sum_d 4129 12.0 0.670 1.134 0.670 1.134 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.989 0.992 mp_alltoall_z22v 1201 16.6 0.903 0.979 0.903 0.979 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.935 0.949 rs_pw_transfer_PW2RS_50 119 14.3 0.585 0.602 0.886 0.948 cp_fm_cholesky_invert 11 10.9 0.895 0.900 0.895 0.900 acc_transpose_blocks_kernels 27432 16.5 0.184 0.271 0.666 0.892 dbcsr_complete_redistribute 329 12.2 0.083 0.118 0.762 0.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.591000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=463.454545, yerr=1.233151 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 520.794112E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_ISendRecv 22134 57667. MP_Wait 33054 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.037 32.877 32.878 qs_mol_dyn_low 1 2.0 0.003 0.003 32.623 32.631 qs_forces 11 3.9 0.005 0.006 32.561 32.564 qs_energies 11 4.9 0.002 0.005 30.969 30.971 scf_env_do_scf 11 5.9 0.001 0.001 26.183 26.183 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 23.455 23.456 dbcsr_multiply_generic 2286 12.5 0.093 0.095 16.728 16.814 velocity_verlet 10 3.0 0.002 0.002 15.737 15.739 qs_scf_new_mos 108 7.5 0.001 0.001 15.333 15.350 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.333 15.350 ot_scf_mini 108 9.5 0.002 0.003 14.589 14.604 multiply_cannon 2286 13.5 0.197 0.206 13.146 13.898 multiply_cannon_loop 2286 14.5 0.644 0.673 12.287 13.026 ot_mini 108 10.5 0.001 0.001 9.038 9.053 qs_ot_get_derivative 108 11.5 0.001 0.001 7.505 7.516 multiply_cannon_multrec 18288 15.5 1.939 2.886 6.727 6.999 rebuild_ks_matrix 119 8.3 0.000 0.000 6.438 6.456 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.438 6.455 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.687 5.703 dbcsr_mm_accdrv_process 38222 16.0 4.062 5.475 4.704 5.534 sum_up_and_integrate 119 10.3 0.030 0.031 4.058 4.064 mp_waitall_1 169478 16.3 3.209 4.050 3.209 4.050 integrate_v_rspace 119 11.3 0.002 0.003 4.027 4.038 init_scf_run 11 5.9 0.000 0.001 3.544 3.544 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.544 3.544 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.797 3.425 qs_ot_get_p 119 10.4 0.001 0.002 3.347 3.370 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.329 3.337 calculate_rho_elec 119 8.7 0.030 0.031 3.329 3.337 init_scf_loop 11 6.9 0.001 0.005 2.707 2.708 rs_pw_transfer 974 11.9 0.009 0.010 2.490 2.596 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.063 2.359 apply_single 119 13.6 0.000 0.000 2.062 2.359 calculate_first_density_matrix 1 7.0 0.001 0.003 2.261 2.262 multiply_cannon_metrocomm3 18288 15.5 0.045 0.046 1.505 2.209 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.153 2.160 make_m2s 4572 13.5 0.044 0.045 1.952 2.108 make_images 4572 14.5 0.191 0.202 1.867 2.023 density_rs2pw 119 9.7 0.004 0.004 1.925 2.001 grid_integrate_task_list 119 12.3 1.802 1.897 1.802 1.897 prepare_preconditioner 11 7.9 0.000 0.000 1.874 1.877 make_preconditioner 11 8.9 0.000 0.002 1.874 1.877 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.872 1.872 potential_pw2rs 119 12.3 0.007 0.008 1.845 1.855 calculate_dm_sparse 119 9.5 0.000 0.000 1.832 1.847 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.726 1.809 pw_transfer 1439 11.6 0.063 0.065 1.747 1.758 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.743 1.745 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.657 1.671 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.653 1.662 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.606 1.606 mp_sum_l 7207 12.9 1.247 1.583 1.247 1.583 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.572 1.581 cp_fm_redistribute_end 50 14.0 1.188 1.578 1.190 1.578 cp_fm_diag_elpa_base 50 14.0 0.371 1.510 0.386 1.549 multiply_cannon_sync_h2d 18288 15.5 1.360 1.537 1.360 1.537 ot_diis_step 108 11.5 0.011 0.011 1.501 1.502 fft3d_ps 1201 14.6 0.503 0.519 1.345 1.355 acc_transpose_blocks 18288 15.5 0.075 0.076 1.271 1.298 jit_kernel_multiply 7 16.3 0.589 1.277 0.589 1.277 fft_wrap_pw1pw2_140 487 13.2 0.085 0.089 1.256 1.270 grid_collocate_task_list 119 9.7 1.204 1.270 1.204 1.270 wfi_extrapolate 11 7.9 0.001 0.001 1.230 1.230 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.184 1.187 make_images_data 4572 15.5 0.045 0.049 0.849 1.024 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.993 0.994 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.926 0.948 hybrid_alltoall_any 4725 16.4 0.055 0.115 0.727 0.915 acc_transpose_blocks_kernels 18288 16.5 0.211 0.221 0.827 0.845 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.827 cp_fm_cholesky_invert 11 10.9 0.816 0.821 0.816 0.821 mp_alltoall_d11v 2130 13.8 0.730 0.820 0.730 0.820 mp_alltoall_z22v 1201 16.6 0.678 0.735 0.678 0.735 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.649 0.731 mp_allgather_i34 2286 14.5 0.318 0.689 0.318 0.689 mp_waitany 9880 13.7 0.583 0.669 0.583 0.669 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.878000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.181818, yerr=1.695887 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 556.064768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_ISendRecv 16422 74133. MP_Wait 24482 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.047 35.365 35.365 qs_mol_dyn_low 1 2.0 0.003 0.004 35.136 35.144 qs_forces 11 3.9 0.002 0.003 35.077 35.078 qs_energies 11 4.9 0.001 0.004 33.355 33.362 scf_env_do_scf 11 5.9 0.001 0.001 28.302 28.304 scf_env_do_scf_inner_loop 108 6.5 0.002 0.009 24.666 24.666 dbcsr_multiply_generic 2286 12.5 0.100 0.102 18.222 18.319 velocity_verlet 10 3.0 0.002 0.002 18.086 18.088 qs_scf_new_mos 108 7.5 0.001 0.001 16.441 16.492 qs_scf_loop_do_ot 108 8.5 0.001 0.002 16.441 16.491 ot_scf_mini 108 9.5 0.002 0.003 15.493 15.547 multiply_cannon 2286 13.5 0.254 0.295 14.378 14.852 multiply_cannon_loop 2286 14.5 0.943 0.970 13.345 13.742 ot_mini 108 10.5 0.001 0.001 9.394 9.464 multiply_cannon_multrec 27432 15.5 2.329 3.017 8.571 8.950 qs_ot_get_derivative 108 11.5 0.001 0.001 7.565 7.619 dbcsr_mm_accdrv_process 47916 15.9 5.421 7.073 6.148 7.328 rebuild_ks_matrix 119 8.3 0.000 0.000 6.525 6.580 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.525 6.580 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.797 5.844 sum_up_and_integrate 119 10.3 0.035 0.038 3.815 3.821 integrate_v_rspace 119 11.3 0.002 0.002 3.780 3.787 init_scf_run 11 5.9 0.000 0.001 3.669 3.670 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.669 3.670 init_scf_loop 11 6.9 0.000 0.001 3.613 3.614 qs_ot_get_p 119 10.4 0.001 0.001 3.467 3.544 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.337 3.369 calculate_rho_elec 119 8.7 0.040 0.046 3.337 3.368 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.747 3.231 mp_waitall_1 145218 16.4 2.168 2.858 2.168 2.858 prepare_preconditioner 11 7.9 0.000 0.000 2.754 2.762 make_preconditioner 11 8.9 0.000 0.001 2.753 2.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.347 2.682 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.145 2.601 apply_single 119 13.6 0.000 0.000 2.145 2.601 make_m2s 4572 13.5 0.054 0.055 2.324 2.453 make_images 4572 14.5 0.274 0.336 2.217 2.344 rs_pw_transfer 974 11.9 0.009 0.009 2.138 2.271 calculate_first_density_matrix 1 7.0 0.000 0.002 2.263 2.266 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.116 2.125 calculate_dm_sparse 119 9.5 0.000 0.000 2.041 2.103 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.976 2.004 density_rs2pw 119 9.7 0.003 0.004 1.829 1.940 grid_integrate_task_list 119 12.3 1.824 1.902 1.824 1.902 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.837 1.838 jit_kernel_multiply 10 15.9 0.664 1.811 0.664 1.811 pw_transfer 1439 11.6 0.063 0.066 1.762 1.791 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.788 1.788 ot_diis_step 108 11.5 0.012 0.012 1.788 1.788 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.673 1.704 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.643 1.656 potential_pw2rs 119 12.3 0.008 0.009 1.613 1.617 mp_sum_l 7207 12.9 1.099 1.602 1.099 1.602 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.511 1.513 cp_fm_redistribute_end 50 14.0 0.997 1.482 0.998 1.482 acc_transpose_blocks 27432 15.5 0.111 0.115 1.459 1.478 cp_fm_diag_elpa_base 50 14.0 0.460 1.410 0.482 1.448 fft3d_ps 1201 14.6 0.531 0.584 1.354 1.376 wfi_extrapolate 11 7.9 0.001 0.001 1.354 1.354 fft_wrap_pw1pw2_140 487 13.2 0.084 0.092 1.319 1.353 grid_collocate_task_list 119 9.7 1.219 1.314 1.219 1.314 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.710 1.273 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.217 1.230 cp_fm_upper_to_full 72 13.5 0.828 1.148 0.828 1.148 dbcsr_complete_redistribute 329 12.2 0.121 0.152 0.857 1.146 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.125 1.142 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.113 1.114 multiply_cannon_sync_h2d 27432 15.5 0.982 1.035 0.982 1.035 make_images_data 4572 15.5 0.045 0.049 0.885 1.032 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.768 0.959 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.617 0.896 mp_alltoall_d11v 2130 13.8 0.789 0.892 0.789 0.892 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.792 0.870 cp_fm_cholesky_invert 11 10.9 0.845 0.849 0.845 0.849 acc_transpose_blocks_kernels 27432 16.5 0.270 0.278 0.827 0.839 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.821 0.826 mp_alltoall_i22 627 13.8 0.453 0.758 0.453 0.758 mp_alltoall_z22v 1201 16.6 0.707 0.733 0.707 0.733 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.365000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.454545, yerr=4.142104 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 597.176320E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_ISendRecv 10710 94533. MP_Wait 16690 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.032 29.703 29.704 qs_mol_dyn_low 1 2.0 0.003 0.007 29.476 29.483 qs_forces 11 3.9 0.005 0.017 29.402 29.407 qs_energies 11 4.9 0.001 0.001 27.672 27.683 scf_env_do_scf 11 5.9 0.000 0.001 22.664 22.665 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 19.946 19.947 velocity_verlet 10 3.0 0.003 0.003 15.185 15.188 dbcsr_multiply_generic 2286 12.5 0.092 0.095 13.093 13.202 qs_scf_new_mos 108 7.5 0.001 0.001 11.945 11.973 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.945 11.972 ot_scf_mini 108 9.5 0.002 0.002 11.232 11.257 multiply_cannon 2286 13.5 0.239 0.250 10.126 10.581 multiply_cannon_loop 2286 14.5 0.334 0.346 9.117 9.400 multiply_cannon_multrec 9144 15.5 1.700 1.949 6.095 6.340 ot_mini 108 10.5 0.001 0.001 6.295 6.324 rebuild_ks_matrix 119 8.3 0.000 0.000 6.059 6.086 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.058 6.085 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.409 5.434 qs_ot_get_derivative 108 11.5 0.001 0.001 4.918 4.943 dbcsr_mm_accdrv_process 12550 15.8 3.159 4.288 4.294 4.381 sum_up_and_integrate 119 10.3 0.037 0.041 3.751 3.755 integrate_v_rspace 119 11.3 0.003 0.003 3.713 3.718 init_scf_run 11 5.9 0.000 0.001 3.531 3.531 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.530 3.530 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.454 3.463 calculate_rho_elec 119 8.7 0.059 0.061 3.453 3.463 qs_ot_get_p 119 10.4 0.001 0.001 3.045 3.078 init_scf_loop 11 6.9 0.000 0.000 2.689 2.690 mp_waitall_1 121218 16.5 2.147 2.627 2.147 2.627 calculate_first_density_matrix 1 7.0 0.000 0.000 2.291 2.292 make_m2s 4572 13.5 0.034 0.035 2.030 2.167 make_images 4572 14.5 0.268 0.300 1.941 2.076 rs_pw_transfer 974 11.9 0.008 0.008 1.919 2.013 jit_kernel_multiply 10 15.4 1.096 2.005 1.096 2.005 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.969 1.972 density_rs2pw 119 9.7 0.003 0.004 1.841 1.938 prepare_preconditioner 11 7.9 0.000 0.000 1.929 1.933 make_preconditioner 11 8.9 0.000 0.000 1.929 1.933 grid_integrate_task_list 119 12.3 1.852 1.912 1.852 1.912 pw_transfer 1439 11.6 0.063 0.065 1.889 1.900 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.813 1.841 calculate_dm_sparse 119 9.5 0.000 0.000 1.793 1.811 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.799 1.810 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.779 1.803 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.754 1.755 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.749 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.551 1.561 potential_pw2rs 119 12.3 0.010 0.010 1.510 1.518 fft3d_ps 1201 14.6 0.541 0.553 1.467 1.474 fft_wrap_pw1pw2_140 487 13.2 0.082 0.086 1.403 1.416 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.414 1.416 cp_fm_redistribute_end 50 14.0 0.704 1.385 0.705 1.385 cp_fm_diag_elpa_base 50 14.0 0.635 1.306 0.678 1.360 grid_collocate_task_list 119 9.7 1.272 1.356 1.272 1.356 ot_diis_step 108 11.5 0.013 0.013 1.355 1.355 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.271 1.283 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.215 1.277 apply_single 119 13.6 0.000 0.000 1.215 1.277 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.235 1.244 wfi_extrapolate 11 7.9 0.001 0.001 1.193 1.193 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.187 1.191 make_images_data 4572 15.5 0.038 0.042 0.954 1.188 hybrid_alltoall_any 4725 16.4 0.062 0.175 0.917 1.142 cp_fm_cholesky_invert 11 10.9 1.004 1.007 1.004 1.007 acc_transpose_blocks 9144 15.5 0.039 0.040 0.985 0.994 mp_alltoall_d11v 2130 13.8 0.917 0.992 0.917 0.992 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.867 0.916 mp_alltoall_z22v 1201 16.6 0.799 0.835 0.799 0.835 mp_allgather_i34 2286 14.5 0.330 0.822 0.330 0.822 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.794 0.800 multiply_cannon_sync_h2d 9144 15.5 0.708 0.780 0.708 0.780 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.776 0.778 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.354 0.759 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.741 0.745 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.691 0.745 mp_sum_l 7207 12.9 0.506 0.660 0.506 0.660 dbcsr_complete_redistribute 329 12.2 0.166 0.172 0.602 0.630 jit_kernel_transpose 5 15.6 0.624 0.628 0.624 0.628 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.704000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=566.545455, yerr=3.962635 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 753.647616E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_ISendRecv 4998 189067. MP_Wait 8898 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 0.020 0.043 42.502 42.504 qs_mol_dyn_low 1 2.0 0.003 0.004 42.258 42.265 qs_forces 11 3.9 0.001 0.002 42.193 42.194 qs_energies 11 4.9 0.001 0.001 40.200 40.203 scf_env_do_scf 11 5.9 0.000 0.001 34.430 34.430 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 26.179 26.180 velocity_verlet 10 3.0 0.002 0.005 23.964 23.969 dbcsr_multiply_generic 2286 12.5 0.102 0.103 18.049 18.298 qs_scf_new_mos 108 7.5 0.001 0.001 16.868 16.957 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.867 16.956 ot_scf_mini 108 9.5 0.002 0.002 15.745 15.839 multiply_cannon 2286 13.5 0.303 0.311 13.786 14.708 multiply_cannon_loop 2286 14.5 0.346 0.353 12.497 13.369 ot_mini 108 10.5 0.001 0.001 9.204 9.318 multiply_cannon_multrec 9144 15.5 3.329 4.673 8.607 8.738 init_scf_loop 11 6.9 0.000 0.000 8.217 8.220 prepare_preconditioner 11 7.9 0.000 0.000 7.268 7.281 make_preconditioner 11 8.9 0.000 0.000 7.268 7.281 rebuild_ks_matrix 119 8.3 0.000 0.000 7.070 7.218 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.070 7.217 qs_ot_get_derivative 108 11.5 0.001 0.001 7.120 7.213 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.792 7.152 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.393 6.527 dbcsr_mm_accdrv_process 12550 15.8 4.152 5.659 5.155 6.473 cp_fm_upper_to_full 72 14.2 3.268 4.681 3.268 4.681 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.914 3.918 calculate_rho_elec 119 8.7 0.118 0.121 3.913 3.917 sum_up_and_integrate 119 10.3 0.064 0.066 3.872 3.879 qs_ot_get_p 119 10.4 0.001 0.001 3.720 3.866 integrate_v_rspace 119 11.3 0.003 0.003 3.808 3.817 mp_waitall_1 97218 16.6 2.983 3.742 2.983 3.742 init_scf_run 11 5.9 0.000 0.001 3.711 3.711 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.711 3.711 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.618 3.015 dbcsr_complete_redistribute 329 12.2 0.285 0.292 2.098 2.933 make_m2s 4572 13.5 0.038 0.038 2.563 2.723 make_images 4572 14.5 0.355 0.385 2.442 2.602 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.739 2.574 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.193 2.503 apply_single 119 13.6 0.000 0.000 2.193 2.503 mp_alltoall_i22 627 13.8 1.453 2.332 1.453 2.332 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.300 2.302 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.470 2.297 calculate_dm_sparse 119 9.5 0.000 0.000 2.235 2.250 calculate_first_density_matrix 1 7.0 0.000 0.000 2.211 2.211 pw_transfer 1439 11.6 0.066 0.067 2.204 2.208 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.414 2.188 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.109 2.112 grid_integrate_task_list 119 12.3 2.056 2.067 2.056 2.067 ot_diis_step 108 11.5 0.014 0.014 2.050 2.051 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.003 2.004 density_rs2pw 119 9.7 0.003 0.004 1.955 1.965 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.939 1.940 mp_sum_l 7207 12.9 1.163 1.910 1.163 1.910 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.815 1.863 jit_kernel_multiply 10 15.5 0.975 1.833 0.975 1.833 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.754 1.755 fft3d_ps 1201 14.6 0.567 0.577 1.743 1.746 fft_wrap_pw1pw2_140 487 13.2 0.088 0.090 1.722 1.727 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.660 1.706 cp_fm_cholesky_invert 11 10.9 1.688 1.692 1.688 1.692 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.658 1.662 cp_fm_diag_elpa_base 50 14.0 1.508 1.565 1.656 1.660 rs_pw_transfer 974 11.9 0.009 0.009 1.554 1.577 grid_collocate_task_list 119 9.7 1.487 1.499 1.487 1.499 hybrid_alltoall_any 4725 16.4 0.087 0.149 1.274 1.494 make_images_data 4572 15.5 0.042 0.046 1.246 1.484 wfi_extrapolate 11 7.9 0.001 0.001 1.437 1.437 potential_pw2rs 119 12.3 0.014 0.015 1.350 1.355 mp_alltoall_d11v 2130 13.8 1.258 1.282 1.258 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.267 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.198 1.219 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.093 1.106 mp_alltoall_z22v 1201 16.6 1.040 1.058 1.040 1.058 multiply_cannon_sync_h2d 9144 15.5 1.039 1.044 1.039 1.044 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.993 1.040 qs_create_task_list 11 7.9 0.000 0.000 0.944 0.956 generate_qs_task_list 11 8.9 0.372 0.391 0.944 0.955 acc_transpose_blocks 9144 15.5 0.039 0.039 0.919 0.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.891 0.905 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.504000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=705.909091, yerr=15.364981 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 499.929088E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 2465369. MP_ISendRecv 41800 9096. MP_Wait 58168 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.074 82.316 82.321 qs_mol_dyn_low 1 2.0 0.003 0.004 81.923 81.933 qs_forces 11 3.9 0.002 0.003 81.855 81.856 qs_energies 11 4.9 0.023 0.033 79.058 79.072 scf_env_do_scf 11 5.9 0.001 0.001 70.133 70.136 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 64.546 64.546 dbcsr_multiply_generic 2055 12.4 0.105 0.107 52.201 52.500 qs_scf_new_mos 99 7.5 0.000 0.001 47.816 47.950 qs_scf_loop_do_ot 99 8.5 0.000 0.001 47.816 47.950 ot_scf_mini 99 9.5 0.002 0.002 45.455 45.556 multiply_cannon 2055 13.4 0.177 0.186 42.713 43.739 multiply_cannon_loop 2055 14.4 1.571 1.613 41.793 42.839 velocity_verlet 10 3.0 0.001 0.002 41.472 41.473 ot_mini 99 10.5 0.001 0.001 27.590 27.701 qs_ot_get_derivative 99 11.5 0.001 0.002 20.786 20.896 multiply_cannon_multrec 49320 15.4 12.130 12.916 17.267 18.028 rebuild_ks_matrix 110 8.3 0.000 0.001 14.300 14.472 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.299 14.472 mp_waitall_1 241148 16.1 11.733 12.911 11.733 12.911 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.571 12.722 multiply_cannon_sync_h2d 49320 15.4 9.992 10.727 9.992 10.727 qs_ot_get_p 110 10.4 0.001 0.001 9.318 9.439 multiply_cannon_metrocomm3 49320 15.4 0.077 0.081 6.932 8.240 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.177 7.740 apply_single 110 13.6 0.000 0.001 7.177 7.740 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.010 7.566 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.893 6.966 sum_up_and_integrate 110 10.3 0.036 0.043 6.930 6.946 integrate_v_rspace 110 11.3 0.002 0.003 6.893 6.918 init_scf_run 11 5.9 0.000 0.001 6.859 6.860 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.859 6.859 ot_diis_step 99 11.5 0.006 0.006 6.541 6.541 qs_rho_update_rho_low 110 7.6 0.000 0.001 6.117 6.259 calculate_rho_elec 110 8.6 0.020 0.024 6.117 6.258 qs_ot_p2m_diag 48 11.0 0.012 0.018 6.084 6.119 mp_sum_l 6514 12.8 5.137 5.992 5.137 5.992 init_scf_loop 11 6.9 0.000 0.000 5.560 5.560 dbcsr_mm_accdrv_process 87628 16.1 2.082 2.189 5.017 5.347 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.034 5.034 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.569 4.571 cp_fm_redistribute_end 48 14.0 3.981 4.546 3.985 4.546 cp_fm_diag_elpa_base 48 14.0 0.555 4.450 0.559 4.477 make_m2s 4110 13.4 0.060 0.065 3.903 4.000 calculate_dm_sparse 110 9.5 0.000 0.001 3.881 3.970 wfi_extrapolate 11 7.9 0.001 0.001 3.968 3.968 multiply_cannon_metrocomm1 49320 15.4 0.061 0.064 2.752 3.913 make_images 4110 14.4 0.178 0.191 3.808 3.908 rs_pw_transfer 902 11.9 0.012 0.013 3.673 3.793 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.565 3.568 prepare_preconditioner 11 7.9 0.000 0.000 3.413 3.437 make_preconditioner 11 8.9 0.000 0.000 3.413 3.437 grid_integrate_task_list 110 12.3 3.257 3.385 3.257 3.385 density_rs2pw 110 9.6 0.004 0.004 3.211 3.354 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.287 3.331 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.234 3.284 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.184 3.224 pw_transfer 1331 11.6 0.054 0.068 2.794 2.856 calculate_first_density_matrix 1 7.0 0.000 0.001 2.802 2.811 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 2.707 2.772 jit_kernel_multiply 13 15.9 2.648 2.757 2.648 2.757 mp_alltoall_d11v 2046 13.8 2.113 2.638 2.113 2.638 potential_pw2rs 110 12.3 0.006 0.007 2.521 2.543 fft_wrap_pw1pw2_140 451 13.1 0.169 0.191 2.280 2.345 fft3d_ps 1111 14.6 0.750 0.836 2.245 2.301 acc_transpose_blocks 49320 15.4 0.213 0.224 2.203 2.275 grid_collocate_task_list 110 9.6 2.088 2.176 2.088 2.176 mp_waitany 14300 13.8 1.831 2.001 1.831 2.001 mp_sum_d 3883 11.9 1.419 1.938 1.419 1.938 make_images_data 4110 15.4 0.043 0.046 1.789 1.928 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.868 1.897 cp_fm_cholesky_invert 11 10.9 1.830 1.834 1.830 1.834 hybrid_alltoall_any 4261 16.3 0.082 0.476 1.575 1.831 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.794 1.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.321000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.363636, yerr=1.772144 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 586.600448E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 3203306. MP_ISendRecv 20680 26400. MP_Wait 32692 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 68.531 68.532 qs_mol_dyn_low 1 2.0 0.003 0.003 68.317 68.327 qs_forces 11 3.9 0.002 0.003 68.230 68.231 qs_energies 11 4.9 0.001 0.002 64.930 64.933 scf_env_do_scf 11 5.9 0.000 0.001 56.486 56.489 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.598 48.598 dbcsr_multiply_generic 2055 12.4 0.113 0.117 37.922 38.142 velocity_verlet 10 3.0 0.008 0.015 36.374 36.375 qs_scf_new_mos 99 7.5 0.001 0.001 32.793 32.954 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.793 32.954 multiply_cannon 2055 13.4 0.222 0.247 31.128 32.409 ot_scf_mini 99 9.5 0.003 0.003 31.116 31.261 multiply_cannon_loop 2055 14.4 0.938 0.957 29.751 30.821 ot_mini 99 10.5 0.001 0.001 18.403 18.555 multiply_cannon_multrec 24660 15.4 7.592 9.432 13.784 15.572 rebuild_ks_matrix 110 8.3 0.000 0.000 13.629 13.733 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.629 13.733 qs_ot_get_derivative 99 11.5 0.001 0.001 12.515 12.665 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.021 12.110 mp_waitall_1 186928 16.3 8.014 10.775 8.014 10.775 multiply_cannon_sync_h2d 24660 15.4 7.046 8.204 7.046 8.204 init_scf_loop 11 6.9 0.000 0.000 7.850 7.850 multiply_cannon_metrocomm3 24660 15.4 0.070 0.072 5.234 7.798 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.585 7.419 apply_single 110 13.6 0.000 0.001 6.585 7.419 sum_up_and_integrate 110 10.3 0.052 0.060 6.514 6.527 integrate_v_rspace 110 11.3 0.002 0.002 6.462 6.476 dbcsr_mm_accdrv_process 52282 16.1 4.713 5.466 6.034 6.360 qs_ot_get_p 110 10.4 0.001 0.001 6.023 6.199 init_scf_run 11 5.9 0.000 0.001 6.098 6.098 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.097 6.098 prepare_preconditioner 11 7.9 0.000 0.000 5.827 5.846 make_preconditioner 11 8.9 0.000 0.000 5.827 5.846 ot_diis_step 99 11.5 0.010 0.010 5.834 5.834 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.714 5.722 calculate_rho_elec 110 8.6 0.039 0.047 5.714 5.722 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.388 5.543 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.691 5.440 make_m2s 4110 13.4 0.056 0.059 4.321 4.851 make_images 4110 14.4 0.401 0.445 4.213 4.739 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.100 4.123 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.660 3.660 wfi_extrapolate 11 7.9 0.001 0.001 3.523 3.523 pw_transfer 1331 11.6 0.065 0.073 3.289 3.432 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.182 3.327 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.234 3.316 grid_integrate_task_list 110 12.3 3.149 3.295 3.149 3.295 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.282 3.284 density_rs2pw 110 9.6 0.004 0.004 3.040 3.203 rs_pw_transfer 902 11.9 0.012 0.014 2.939 3.137 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.103 3.104 cp_fm_redistribute_end 48 14.0 2.313 3.075 2.315 3.075 cp_fm_diag_elpa_base 48 14.0 0.727 2.951 0.758 3.036 calculate_dm_sparse 110 9.5 0.001 0.001 2.976 3.006 make_images_data 4110 15.4 0.047 0.051 2.443 2.987 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.930 2.971 hybrid_alltoall_any 4261 16.3 0.102 0.441 2.124 2.920 fft_wrap_pw1pw2_140 451 13.1 0.200 0.217 2.638 2.778 cp_fm_cholesky_invert 11 10.9 2.750 2.758 2.750 2.758 fft3d_ps 1111 14.6 1.074 1.265 2.539 2.669 calculate_first_density_matrix 1 7.0 0.000 0.000 2.486 2.488 potential_pw2rs 110 12.3 0.008 0.009 2.405 2.421 mp_sum_l 6514 12.8 1.846 2.392 1.846 2.392 grid_collocate_task_list 110 9.6 2.119 2.262 2.119 2.262 mp_alltoall_d11v 2046 13.8 1.764 2.167 1.764 2.167 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.978 1.995 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.834 1.835 mp_allgather_i34 2055 14.4 0.704 1.799 0.704 1.799 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.775 1.785 multiply_cannon_metrocomm4 22605 15.4 0.073 0.077 0.779 1.689 jit_kernel_multiply 8 16.3 0.965 1.597 0.965 1.597 acc_transpose_blocks 24660 15.4 0.107 0.111 1.544 1.585 mp_irecv_dv 57340 16.2 0.655 1.569 0.655 1.569 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.547 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.366 1.477 dbcsr_complete_redistribute 325 12.2 0.245 0.303 1.193 1.465 cp_fm_cholesky_decompose 22 10.9 1.405 1.412 1.405 1.412 mp_sum_d 3883 11.9 0.931 1.386 0.931 1.386 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.532000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=554.090909, yerr=7.229416 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 658.280448E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_ISendRecv 13640 27424. MP_Wait 32318 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.101 0.291 61.366 61.369 qs_mol_dyn_low 1 2.0 0.003 0.005 60.727 60.745 qs_forces 11 3.9 0.007 0.043 60.599 60.616 qs_energies 11 4.9 0.003 0.010 57.392 57.397 scf_env_do_scf 11 5.9 0.001 0.001 49.388 49.388 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 40.394 40.395 velocity_verlet 10 3.0 0.003 0.006 33.061 33.068 dbcsr_multiply_generic 2055 12.4 0.108 0.111 28.760 29.093 qs_scf_new_mos 99 7.5 0.001 0.001 25.752 25.862 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.752 25.862 ot_scf_mini 99 9.5 0.003 0.004 24.504 24.622 multiply_cannon 2055 13.4 0.212 0.225 22.208 23.589 multiply_cannon_loop 2055 14.4 0.623 0.639 21.003 22.195 ot_mini 99 10.5 0.001 0.002 13.888 14.011 rebuild_ks_matrix 110 8.3 0.000 0.000 12.261 12.414 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.017 12.260 12.414 mp_waitall_1 146766 16.3 7.583 11.128 7.583 11.128 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.828 10.965 multiply_cannon_multrec 16440 15.4 3.796 5.053 9.575 10.699 qs_ot_get_derivative 99 11.5 0.001 0.002 9.369 9.487 init_scf_loop 11 6.9 0.001 0.002 8.956 8.959 multiply_cannon_metrocomm3 16440 15.4 0.042 0.044 4.569 7.952 prepare_preconditioner 11 7.9 0.000 0.000 7.161 7.178 make_preconditioner 11 8.9 0.000 0.001 7.161 7.178 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.531 6.860 sum_up_and_integrate 110 10.3 0.060 0.062 6.491 6.504 integrate_v_rspace 110 11.3 0.002 0.003 6.431 6.445 qs_ot_get_p 110 10.4 0.001 0.002 5.782 5.952 dbcsr_mm_accdrv_process 34862 16.1 4.550 5.215 5.635 5.808 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.601 5.616 calculate_rho_elec 110 8.6 0.058 0.058 5.600 5.615 init_scf_run 11 5.9 0.000 0.001 5.536 5.536 scf_env_initial_rho_setup 11 6.9 0.001 0.004 5.536 5.536 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.021 5.481 apply_single 110 13.6 0.000 0.000 5.021 5.481 make_m2s 4110 13.4 0.050 0.052 4.306 4.653 make_images 4110 14.4 0.395 0.513 4.191 4.539 ot_diis_step 99 11.5 0.011 0.013 4.482 4.482 multiply_cannon_sync_h2d 16440 15.4 3.611 4.287 3.611 4.287 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.057 4.063 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.167 3.810 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.702 3.703 grid_integrate_task_list 110 12.3 3.175 3.454 3.175 3.454 pw_transfer 1331 11.6 0.065 0.070 3.245 3.256 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.154 3.156 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.138 3.152 cp_fm_redistribute_end 48 14.0 1.959 3.123 1.961 3.123 cp_fm_diag_elpa_base 48 14.0 1.098 2.982 1.157 3.089 density_rs2pw 110 9.6 0.004 0.004 2.869 3.057 wfi_extrapolate 11 7.9 0.001 0.007 2.993 2.994 cp_fm_cholesky_invert 11 10.9 2.985 2.993 2.985 2.993 rs_pw_transfer 902 11.9 0.011 0.012 2.763 2.966 make_images_data 4110 15.4 0.044 0.048 2.514 2.956 hybrid_alltoall_any 4261 16.3 0.105 0.374 2.204 2.835 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.772 2.773 fft_wrap_pw1pw2_140 451 13.1 0.209 0.213 2.650 2.663 calculate_dm_sparse 110 9.5 0.001 0.001 2.522 2.549 multiply_cannon_metrocomm4 14385 15.4 0.045 0.050 0.903 2.543 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.440 2.510 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.435 2.484 fft3d_ps 1111 14.6 1.065 1.075 2.451 2.465 calculate_first_density_matrix 1 7.0 0.001 0.003 2.457 2.458 mp_sum_l 6514 12.8 1.703 2.420 1.703 2.420 mp_irecv_dv 48980 15.7 0.832 2.415 0.832 2.415 grid_collocate_task_list 110 9.6 2.175 2.372 2.175 2.372 potential_pw2rs 110 12.3 0.010 0.010 2.307 2.316 mp_alltoall_d11v 2046 13.8 1.817 2.058 1.817 2.058 qs_energies_init_hamiltonians 11 5.9 0.004 0.021 1.992 1.994 dbcsr_complete_redistribute 325 12.2 0.320 0.374 1.453 1.899 cp_fm_upper_to_full 70 13.6 1.351 1.783 1.351 1.783 cp_fm_cholesky_decompose 22 10.9 1.734 1.756 1.734 1.756 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.647 1.671 mp_allgather_i34 2055 14.4 0.526 1.628 0.526 1.628 jit_kernel_multiply 8 16.4 0.691 1.498 0.691 1.498 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.365 1.496 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.462 1.474 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.945 1.371 mp_waitany 17072 13.8 1.158 1.359 1.158 1.359 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.299 1.308 acc_transpose_blocks 16440 15.4 0.072 0.074 1.223 1.235 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=61.369000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=621.909091, yerr=9.394917 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 732.987392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_ISendRecv 10120 43184. MP_Wait 25102 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.034 68.195 68.197 qs_mol_dyn_low 1 2.0 0.003 0.010 67.915 67.924 qs_forces 11 3.9 0.002 0.002 67.843 67.844 qs_energies 11 4.9 0.002 0.004 64.390 64.394 scf_env_do_scf 11 5.9 0.001 0.001 55.783 55.786 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 42.976 42.977 velocity_verlet 10 3.0 0.004 0.005 38.522 38.525 dbcsr_multiply_generic 2055 12.4 0.116 0.119 30.407 30.675 qs_scf_new_mos 99 7.5 0.001 0.001 28.144 28.270 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.143 28.270 ot_scf_mini 99 9.5 0.003 0.003 26.489 26.592 multiply_cannon 2055 13.4 0.244 0.262 22.990 23.789 multiply_cannon_loop 2055 14.4 0.887 0.913 21.386 22.009 ot_mini 99 10.5 0.001 0.001 14.839 14.959 multiply_cannon_multrec 24660 15.4 4.219 6.853 12.816 14.031 init_scf_loop 11 6.9 0.001 0.003 12.751 12.753 rebuild_ks_matrix 110 8.3 0.000 0.000 12.216 12.309 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.216 12.309 prepare_preconditioner 11 7.9 0.000 0.000 10.947 10.960 make_preconditioner 11 8.9 0.000 0.001 10.947 10.960 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.792 10.877 qs_ot_get_derivative 99 11.5 0.001 0.001 10.637 10.744 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.104 10.619 dbcsr_mm_accdrv_process 52304 16.0 6.894 8.576 8.451 9.421 mp_waitall_1 126806 16.4 5.005 6.890 5.005 6.890 qs_ot_get_p 110 10.4 0.001 0.001 6.557 6.683 sum_up_and_integrate 110 10.3 0.068 0.071 6.573 6.588 integrate_v_rspace 110 11.3 0.003 0.003 6.505 6.524 make_m2s 4110 13.4 0.059 0.061 5.677 6.030 make_images 4110 14.4 0.579 0.703 5.536 5.889 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.875 5.884 calculate_rho_elec 110 8.6 0.077 0.081 5.874 5.883 init_scf_run 11 5.9 0.000 0.001 5.833 5.834 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.833 5.834 cp_fm_upper_to_full 70 13.8 3.356 4.795 3.356 4.795 qs_ot_p2m_diag 48 11.0 0.055 0.064 4.549 4.565 ot_diis_step 99 11.5 0.011 0.012 4.160 4.160 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.051 4.130 apply_single 110 13.6 0.000 0.000 4.050 4.130 dbcsr_complete_redistribute 325 12.2 0.406 0.453 2.882 4.052 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.048 4.049 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.609 3.516 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.445 3.501 pw_transfer 1331 11.6 0.064 0.073 3.461 3.495 grid_integrate_task_list 110 12.3 3.274 3.438 3.274 3.438 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.260 3.417 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.356 3.393 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.389 3.392 cp_fm_cholesky_invert 11 10.9 3.341 3.352 3.341 3.352 cp_fm_redistribute_end 48 14.0 1.684 3.339 1.686 3.341 multiply_cannon_sync_h2d 24660 15.4 3.153 3.306 3.153 3.306 cp_fm_diag_elpa_base 48 14.0 1.568 3.194 1.651 3.305 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.194 3.255 make_images_data 4110 15.4 0.046 0.049 2.840 3.203 hybrid_alltoall_any 4261 16.3 0.120 0.458 2.489 3.169 density_rs2pw 110 9.6 0.004 0.004 3.000 3.133 calculate_dm_sparse 110 9.5 0.001 0.001 3.084 3.114 wfi_extrapolate 11 7.9 0.001 0.001 3.080 3.081 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.833 2.977 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.949 2.950 mp_alltoall_i22 605 13.7 1.721 2.933 1.721 2.933 rs_pw_transfer 902 11.9 0.010 0.011 2.680 2.830 fft_wrap_pw1pw2_140 451 13.1 0.201 0.212 2.781 2.822 fft3d_ps 1111 14.6 1.065 1.097 2.658 2.682 calculate_first_density_matrix 1 7.0 0.000 0.001 2.654 2.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.543 2.578 jit_kernel_multiply 12 15.8 1.222 2.371 1.222 2.371 grid_collocate_task_list 110 9.6 2.226 2.358 2.226 2.358 potential_pw2rs 110 12.3 0.013 0.013 2.309 2.328 mp_alltoall_d11v 2046 13.8 2.006 2.312 2.006 2.312 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.240 2.241 cp_fm_cholesky_decompose 22 10.9 1.952 2.003 1.952 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.780 1.810 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.597 1.699 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.002 1.679 1.694 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.673 1.683 mp_sum_l 6514 12.8 1.060 1.636 1.060 1.636 acc_transpose_blocks 24660 15.4 0.104 0.106 1.595 1.617 mp_alltoall_z22v 1111 16.6 1.376 1.445 1.376 1.445 multiply_cannon_metrocomm4 20550 15.4 0.056 0.059 0.841 1.424 mp_allgather_i34 2055 14.4 0.702 1.379 0.702 1.379 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=68.197000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.818182, yerr=10.088861 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 831.344640E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_ISendRecv 6600 54848. MP_Wait 17226 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.069 0.082 56.338 56.341 qs_mol_dyn_low 1 2.0 0.003 0.003 56.056 56.064 qs_forces 11 3.9 0.002 0.002 55.990 55.990 qs_energies 11 4.9 0.001 0.002 52.364 52.371 scf_env_do_scf 11 5.9 0.000 0.001 44.072 44.072 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 35.937 35.938 velocity_verlet 10 3.0 0.002 0.002 31.309 31.312 dbcsr_multiply_generic 2055 12.4 0.104 0.106 23.260 23.418 qs_scf_new_mos 99 7.5 0.001 0.001 21.605 21.646 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.604 21.645 ot_scf_mini 99 9.5 0.002 0.002 20.341 20.367 multiply_cannon 2055 13.4 0.250 0.260 17.541 18.868 multiply_cannon_loop 2055 14.4 0.325 0.337 16.145 16.494 rebuild_ks_matrix 110 8.3 0.000 0.000 11.564 11.597 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 11.564 11.596 ot_mini 99 10.5 0.001 0.001 10.822 10.841 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.293 10.322 multiply_cannon_multrec 8220 15.4 3.236 4.517 7.614 8.659 init_scf_loop 11 6.9 0.000 0.000 8.083 8.084 mp_waitall_1 106626 16.5 6.323 7.946 6.323 7.946 qs_ot_get_derivative 99 11.5 0.001 0.001 6.997 7.021 prepare_preconditioner 11 7.9 0.000 0.000 6.455 6.459 make_preconditioner 11 8.9 0.000 0.000 6.454 6.459 sum_up_and_integrate 110 10.3 0.079 0.081 6.274 6.286 integrate_v_rspace 110 11.3 0.003 0.003 6.194 6.207 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.025 6.104 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.759 5.773 calculate_rho_elec 110 8.6 0.115 0.115 5.759 5.772 qs_ot_get_p 110 10.4 0.001 0.001 5.540 5.566 dbcsr_mm_accdrv_process 17442 15.9 2.957 4.209 4.248 5.264 init_scf_run 11 5.9 0.000 0.001 5.167 5.168 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.167 5.167 make_m2s 4110 13.4 0.038 0.040 4.348 4.645 make_images 4110 14.4 0.639 0.698 4.218 4.513 multiply_cannon_metrocomm3 8220 15.4 0.017 0.017 2.964 4.385 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.085 4.090 ot_diis_step 99 11.5 0.012 0.012 3.796 3.796 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.765 3.766 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.710 3.764 apply_single 110 13.6 0.000 0.000 3.710 3.764 grid_integrate_task_list 110 12.3 3.370 3.509 3.370 3.509 pw_transfer 1331 11.6 0.065 0.069 3.364 3.379 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.257 3.274 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.193 3.193 cp_fm_cholesky_invert 11 10.9 3.186 3.190 3.186 3.190 cp_fm_redistribute_end 48 14.0 0.811 3.156 0.818 3.157 cp_fm_diag_elpa_base 48 14.0 2.152 2.950 2.331 3.114 multiply_cannon_sync_h2d 8220 15.4 2.913 3.045 2.913 3.045 density_rs2pw 110 9.6 0.004 0.004 2.777 2.948 make_images_data 4110 15.4 0.038 0.043 2.466 2.927 hybrid_alltoall_any 4261 16.3 0.199 0.865 2.393 2.815 fft_wrap_pw1pw2_140 451 13.1 0.211 0.215 2.756 2.777 wfi_extrapolate 11 7.9 0.001 0.001 2.767 2.767 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.656 2.656 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.548 2.550 calculate_dm_sparse 110 9.5 0.001 0.001 2.491 2.538 fft3d_ps 1111 14.6 1.117 1.152 2.515 2.534 grid_collocate_task_list 110 9.6 2.332 2.499 2.332 2.499 rs_pw_transfer 902 11.9 0.010 0.010 2.264 2.436 calculate_first_density_matrix 1 7.0 0.000 0.000 2.309 2.310 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.174 2.183 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.146 2.167 potential_pw2rs 110 12.3 0.015 0.016 1.996 2.001 mp_alltoall_d11v 2046 13.8 1.757 1.988 1.757 1.988 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.768 1.978 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.923 1.938 cp_fm_cholesky_decompose 22 10.9 1.785 1.806 1.785 1.806 mp_allgather_i34 2055 14.4 0.574 1.690 0.574 1.690 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.501 1.611 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.605 1.610 dbcsr_complete_redistribute 325 12.2 0.558 0.581 1.515 1.605 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.433 1.444 jit_kernel_multiply 7 15.7 0.977 1.372 0.977 1.372 qs_create_task_list 11 7.9 0.000 0.001 1.231 1.330 generate_qs_task_list 11 8.9 0.378 0.446 1.230 1.329 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.836 1.293 mp_waitany 9240 13.8 1.054 1.247 1.054 1.247 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.210 1.224 mp_alltoall_z22v 1111 16.6 1.164 1.189 1.164 1.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.341000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=782.636364, yerr=14.735688 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 1.316012E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_ISendRecv 4620 360267. MP_Wait 7524 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 0.021 0.039 86.248 86.249 qs_mol_dyn_low 1 2.0 0.003 0.004 85.947 85.955 qs_forces 11 3.9 0.002 0.002 85.880 85.880 qs_energies 11 4.9 0.001 0.002 81.796 81.797 scf_env_do_scf 11 5.9 0.001 0.001 71.921 71.921 velocity_verlet 10 3.0 0.002 0.002 55.409 55.415 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.052 43.054 dbcsr_multiply_generic 2055 12.4 0.119 0.122 28.778 28.908 init_scf_loop 11 6.9 0.000 0.000 28.795 28.799 prepare_preconditioner 11 7.9 0.000 0.000 26.846 26.860 make_preconditioner 11 8.9 0.000 0.000 26.846 26.860 qs_scf_new_mos 99 7.5 0.001 0.001 26.603 26.690 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.602 26.689 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.004 26.309 ot_scf_mini 99 9.5 0.002 0.002 24.827 24.908 multiply_cannon 2055 13.4 0.346 0.363 21.596 22.288 multiply_cannon_loop 2055 14.4 0.343 0.345 19.772 20.075 cp_fm_upper_to_full 70 14.2 12.906 18.463 12.906 18.463 ot_mini 99 10.5 0.001 0.001 13.800 13.874 rebuild_ks_matrix 110 8.3 0.000 0.001 13.199 13.284 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.199 13.283 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.987 12.064 dbcsr_complete_redistribute 325 12.2 1.021 1.046 7.449 10.705 multiply_cannon_multrec 8220 15.4 4.367 4.574 9.575 9.688 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.395 9.649 qs_ot_get_derivative 99 11.5 0.001 0.001 9.138 9.219 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.826 9.042 mp_waitall_1 87304 16.6 7.918 8.969 7.918 8.969 mp_alltoall_i22 605 13.7 5.456 8.737 5.456 8.737 sum_up_and_integrate 110 10.3 0.150 0.152 6.663 6.678 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.497 6.534 calculate_rho_elec 110 8.6 0.227 0.227 6.497 6.534 integrate_v_rspace 110 11.3 0.004 0.004 6.512 6.527 qs_ot_get_p 110 10.4 0.001 0.001 5.803 5.890 make_m2s 4110 13.4 0.043 0.043 5.289 5.812 init_scf_run 11 5.9 0.000 0.001 5.634 5.635 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.634 5.634 make_images 4110 14.4 0.880 0.930 5.101 5.624 cp_fm_cholesky_invert 11 10.9 5.498 5.503 5.498 5.503 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.789 5.324 apply_single 110 13.6 0.000 0.000 4.789 5.324 dbcsr_mm_accdrv_process 11614 15.7 3.367 3.701 5.065 5.321 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 4.850 5.152 ot_diis_step 99 11.5 0.016 0.016 4.636 4.637 qs_ot_p2m_diag 48 11.0 0.151 0.156 4.208 4.216 multiply_cannon_sync_h2d 8220 15.4 3.948 3.951 3.948 3.951 pw_transfer 1331 11.6 0.073 0.074 3.799 3.805 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.766 3.766 grid_integrate_task_list 110 12.3 3.658 3.716 3.658 3.716 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.683 3.690 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 3.640 3.641 hybrid_alltoall_any 4261 16.3 0.256 0.554 2.884 3.600 make_images_data 4110 15.4 0.041 0.045 2.897 3.528 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.903 3.422 wfi_extrapolate 11 7.9 0.001 0.001 3.240 3.240 fft_wrap_pw1pw2_140 451 13.1 0.216 0.217 3.187 3.191 calculate_dm_sparse 110 9.5 0.001 0.001 3.144 3.182 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.154 3.155 cp_fm_diag_elpa_base 48 14.0 2.610 2.814 3.152 3.153 density_rs2pw 110 9.6 0.004 0.004 2.994 3.011 fft3d_ps 1111 14.6 1.267 1.280 2.888 2.898 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.863 2.867 grid_collocate_task_list 110 9.6 2.626 2.655 2.626 2.655 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.422 2.464 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.345 2.372 calculate_first_density_matrix 1 7.0 0.000 0.000 2.290 2.290 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.228 rs_pw_transfer 902 11.9 0.010 0.011 2.151 2.190 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.098 2.177 mp_alltoall_d11v 2046 13.8 2.074 2.107 2.074 2.107 potential_pw2rs 110 12.3 0.021 0.022 2.090 2.098 cp_fm_cholesky_decompose 22 10.9 2.062 2.078 2.062 2.078 qs_create_task_list 11 7.9 0.001 0.001 1.902 1.948 generate_qs_task_list 11 8.9 0.739 0.795 1.902 1.947 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.807 1.850 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.814 1.822 jit_kernel_multiply 10 15.2 1.494 1.783 1.494 1.783 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.249000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1190.909091, yerr=59.830753 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 626.610176E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2090093. MP_ISendRecv 48640 18752. MP_Wait 66796 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 205.230 205.231 qs_mol_dyn_low 1 2.0 0.003 0.003 204.840 204.853 qs_forces 11 3.9 0.003 0.004 204.754 204.755 qs_energies 11 4.9 0.001 0.002 199.211 199.222 scf_env_do_scf 11 5.9 0.001 0.001 182.717 182.721 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.791 161.792 dbcsr_multiply_generic 2507 12.6 0.174 0.177 124.247 125.081 velocity_verlet 10 3.0 0.001 0.002 123.573 123.574 qs_scf_new_mos 117 7.6 0.001 0.001 122.384 122.583 qs_scf_loop_do_ot 117 8.6 0.001 0.001 122.383 122.583 ot_scf_mini 117 9.6 0.003 0.003 115.759 115.961 multiply_cannon 2507 13.6 0.239 0.249 101.076 102.999 multiply_cannon_loop 2507 14.6 2.117 2.205 98.805 100.339 ot_mini 117 10.6 0.001 0.001 65.715 65.902 multiply_cannon_multrec 60168 15.6 33.367 35.406 41.732 44.088 qs_ot_get_derivative 117 11.6 0.001 0.004 40.566 40.766 rebuild_ks_matrix 128 8.3 0.001 0.001 33.645 33.880 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.645 33.879 mp_waitall_1 291448 16.2 29.100 32.123 29.100 32.123 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.290 30.507 multiply_cannon_sync_h2d 60168 15.6 27.369 29.615 27.369 29.615 qs_ot_get_p 128 10.4 0.001 0.001 28.083 28.358 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.541 25.625 apply_single 128 13.6 0.001 0.001 24.541 25.624 ot_diis_step 117 11.6 0.007 0.008 24.793 24.795 qs_ot_p2m_diag 83 11.4 0.077 0.092 21.318 21.396 init_scf_loop 11 6.9 0.000 0.000 20.842 20.844 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 18.685 18.841 cp_dbcsr_syevd 83 12.4 0.004 0.005 18.750 18.751 multiply_cannon_metrocomm3 60168 15.6 0.111 0.117 15.591 17.973 prepare_preconditioner 11 7.9 0.000 0.000 16.217 16.246 make_preconditioner 11 8.9 0.000 0.000 16.217 16.246 cp_fm_diag_elpa 83 13.4 0.000 0.001 15.839 15.846 cp_fm_redistribute_end 83 14.4 12.490 15.744 12.504 15.749 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.421 15.577 cp_fm_diag_elpa_base 83 14.4 3.196 15.438 3.231 15.551 make_m2s 5014 13.6 0.102 0.111 13.990 14.391 sum_up_and_integrate 128 10.3 0.090 0.108 14.224 14.239 make_images 5014 14.6 0.404 0.421 13.810 14.221 integrate_v_rspace 128 11.3 0.003 0.004 14.134 14.151 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.013 13.134 calculate_rho_elec 128 8.7 0.045 0.063 13.012 13.133 init_scf_run 11 5.9 0.000 0.001 12.354 12.354 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.354 12.354 cp_fm_cholesky_invert 11 10.9 9.278 9.286 9.278 9.286 wfi_extrapolate 11 7.9 0.001 0.001 9.077 9.077 mp_sum_l 7870 13.0 7.923 8.911 7.923 8.911 calculate_dm_sparse 128 9.5 0.001 0.001 8.432 8.576 dbcsr_mm_accdrv_process 124484 16.2 3.290 3.470 7.930 8.510 multiply_cannon_metrocomm1 60168 15.6 0.089 0.093 6.100 8.147 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.996 8.132 make_images_data 5014 15.6 0.069 0.074 6.899 7.954 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.824 7.915 grid_integrate_task_list 128 12.3 7.018 7.521 7.018 7.521 density_rs2pw 128 9.7 0.006 0.007 6.786 7.401 pw_transfer 1547 11.6 0.075 0.106 7.051 7.364 hybrid_alltoall_any 5200 16.5 0.291 2.278 6.023 7.334 fft_wrap_pw1pw2 1291 12.7 0.010 0.014 6.848 7.137 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.716 6.724 rs_pw_transfer 1046 11.9 0.017 0.019 5.999 6.656 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.849 6.021 fft_wrap_pw1pw2_140 523 13.2 0.442 0.508 5.801 5.979 fft3d_ps 1291 14.7 2.092 2.568 5.644 5.877 mp_alltoall_d11v 2415 14.1 4.430 5.731 4.430 5.731 grid_collocate_task_list 128 9.7 4.732 5.089 4.732 5.089 mp_sum_d 4459 12.1 3.647 4.745 3.647 4.745 cp_fm_cholesky_decompose 22 10.9 4.716 4.729 4.716 4.729 potential_pw2rs 128 12.3 0.010 0.011 4.670 4.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.231000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=593.545455, yerr=6.827267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 826.376192E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100943E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834845264 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4085 57203. MP_Allreduce 11252 946. MP_Sync 170 MP_Alltoall 1969 6422356. MP_ISendRecv 24064 47072. MP_Wait 37948 MP_comm_split 83 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 0.033 0.054 191.504 191.518 qs_mol_dyn_low 1 2.0 0.003 0.003 191.105 191.118 qs_forces 11 3.9 0.003 0.003 191.024 191.025 qs_energies 11 4.9 0.001 0.003 184.311 184.321 scf_env_do_scf 11 5.9 0.001 0.002 168.082 168.092 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 134.502 134.504 velocity_verlet 10 3.0 0.002 0.002 120.275 120.276 dbcsr_multiply_generic 2507 12.6 0.184 0.190 97.362 98.545 qs_scf_new_mos 117 7.6 0.001 0.001 96.090 96.519 qs_scf_loop_do_ot 117 8.6 0.001 0.001 96.089 96.518 ot_scf_mini 117 9.6 0.004 0.004 91.275 91.784 multiply_cannon 2507 13.6 0.478 0.533 77.399 81.503 multiply_cannon_loop 2507 14.6 1.265 1.302 74.107 76.926 ot_mini 117 10.6 0.001 0.001 49.938 50.447 mp_waitall_1 226760 16.4 24.709 37.953 24.709 37.953 multiply_cannon_multrec 30084 15.6 21.972 26.636 31.612 36.491 init_scf_loop 11 6.9 0.000 0.000 33.491 33.493 rebuild_ks_matrix 128 8.3 0.001 0.001 32.608 33.063 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 32.607 33.063 qs_ks_update_qs_env 128 7.6 0.001 0.001 29.388 29.808 prepare_preconditioner 11 7.9 0.000 0.000 29.131 29.195 make_preconditioner 11 8.9 0.000 0.000 29.131 29.195 qs_ot_get_derivative 117 11.6 0.001 0.002 28.031 28.536 make_full_inverse_cholesky 11 9.9 0.000 0.000 27.824 28.376 multiply_cannon_metrocomm3 30084 15.6 0.096 0.100 15.688 27.901 qs_ot_get_p 128 10.4 0.001 0.001 23.106 23.608 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.963 23.081 apply_single 128 13.6 0.001 0.001 21.963 23.080 ot_diis_step 117 11.6 0.014 0.015 21.725 21.728 multiply_cannon_sync_h2d 30084 15.6 19.315 21.445 19.315 21.445 qs_ot_p2m_diag 83 11.4 0.188 0.216 18.242 18.280 cp_fm_cholesky_invert 11 10.9 17.104 17.117 17.104 17.117 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.065 17.066 make_m2s 5014 13.6 0.089 0.094 14.075 15.593 make_images 5014 14.6 1.165 1.349 13.868 15.387 sum_up_and_integrate 128 10.3 0.116 0.131 14.323 14.353 integrate_v_rspace 128 11.3 0.003 0.003 14.206 14.239 cp_fm_diag_elpa 83 13.4 0.000 0.001 13.866 13.876 cp_fm_redistribute_end 83 14.4 8.119 13.791 8.131 13.792 cp_fm_diag_elpa_base 83 14.4 5.419 13.251 5.641 13.675 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.857 12.894 calculate_rho_elec 128 8.7 0.087 0.105 12.857 12.893 init_scf_run 11 5.9 0.000 0.001 11.490 11.492 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.490 11.491 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.089 11.467 multiply_cannon_metrocomm4 27577 15.6 0.096 0.112 3.748 10.439 make_images_data 5014 15.6 0.066 0.074 8.388 10.161 mp_irecv_dv 69486 16.3 3.554 10.060 3.554 10.060 dbcsr_mm_accdrv_process 62242 16.2 4.558 5.340 9.107 9.724 hybrid_alltoall_any 5200 16.5 0.342 1.505 7.172 9.338 wfi_extrapolate 11 7.9 0.001 0.001 8.384 8.384 pw_transfer 1547 11.6 0.085 0.102 7.623 7.680 grid_integrate_task_list 128 12.3 7.162 7.505 7.162 7.505 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.399 7.449 cp_fm_cholesky_decompose 22 10.9 7.205 7.283 7.205 7.283 density_rs2pw 128 9.7 0.006 0.006 6.696 7.047 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.219 7.006 calculate_dm_sparse 128 9.5 0.001 0.001 6.469 6.612 fft_wrap_pw1pw2_140 523 13.2 0.468 0.522 6.415 6.466 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.141 6.149 rs_pw_transfer 1046 11.9 0.015 0.017 5.674 5.983 fft3d_ps 1291 14.7 2.771 2.921 5.772 5.813 mp_sum_l 7870 13.0 4.110 5.727 4.110 5.727 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.367 5.435 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.289 5.417 grid_collocate_task_list 128 9.7 4.907 5.294 4.907 5.294 potential_pw2rs 128 12.3 0.014 0.016 4.833 4.853 mp_alltoall_d11v 2415 14.1 4.038 4.757 4.038 4.757 mp_allgather_i34 2507 14.6 1.855 4.694 1.855 4.694 dbcsr_complete_redistribute 395 12.7 0.766 0.848 3.128 3.996 mp_sum_d 4464 12.1 2.629 3.974 2.629 3.974 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.518000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=787.272727, yerr=3.569846 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 936.562688E+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 4083 57222. MP_Allreduce 11246 986. MP_Sync 168 MP_Alltoall 1700 9383497. MP_ISendRecv 15748 75008. MP_Wait 29528 MP_comm_split 82 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 0.034 0.053 175.370 175.371 qs_mol_dyn_low 1 2.0 0.003 0.003 174.922 174.936 qs_forces 11 3.9 0.003 0.003 174.777 174.781 qs_energies 11 4.9 0.008 0.027 168.281 168.290 scf_env_do_scf 11 5.9 0.001 0.001 152.723 152.725 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 116.868 116.869 velocity_verlet 10 3.0 0.001 0.002 111.962 111.965 qs_scf_new_mos 116 7.6 0.001 0.001 81.273 81.563 qs_scf_loop_do_ot 116 8.6 0.001 0.002 81.272 81.562 dbcsr_multiply_generic 2485 12.5 0.180 0.186 80.098 81.093 ot_scf_mini 116 9.6 0.004 0.004 77.189 77.508 multiply_cannon 2485 13.5 0.493 0.514 60.874 65.040 multiply_cannon_loop 2485 14.5 0.860 0.887 57.742 60.063 ot_mini 116 10.6 0.001 0.001 41.806 42.128 init_scf_loop 11 6.9 0.001 0.005 35.755 35.756 mp_waitall_1 176908 16.5 24.101 33.616 24.101 33.616 prepare_preconditioner 11 7.9 0.000 0.000 31.784 31.831 make_preconditioner 11 8.9 0.000 0.001 31.784 31.831 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.401 30.796 rebuild_ks_matrix 127 8.3 0.001 0.001 29.468 29.892 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 29.467 29.891 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.519 26.906 multiply_cannon_multrec 19880 15.5 13.562 16.836 22.263 25.627 multiply_cannon_metrocomm3 19880 15.5 0.058 0.061 14.489 23.565 qs_ot_get_derivative 116 11.6 0.001 0.002 22.343 22.658 qs_ot_get_p 127 10.4 0.001 0.004 21.068 21.498 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.607 20.566 apply_single 127 13.6 0.001 0.001 19.607 20.565 ot_diis_step 116 11.6 0.018 0.018 19.362 19.363 qs_ot_p2m_diag 82 11.4 0.262 0.269 16.876 16.888 cp_dbcsr_syevd 82 12.4 0.005 0.005 15.857 15.859 multiply_cannon_sync_h2d 19880 15.5 14.209 15.655 14.209 15.655 make_m2s 4970 13.5 0.080 0.086 14.285 15.181 cp_fm_cholesky_invert 11 10.9 14.945 14.954 14.945 14.954 make_images 4970 14.5 1.182 1.265 14.055 14.944 sum_up_and_integrate 127 10.3 0.133 0.145 13.959 13.994 integrate_v_rspace 127 11.3 0.004 0.013 13.826 13.855 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.686 12.722 calculate_rho_elec 127 8.7 0.130 0.145 12.686 12.721 cp_fm_diag_elpa 82 13.4 0.000 0.001 12.593 12.594 cp_fm_redistribute_end 82 14.4 4.765 12.514 4.782 12.518 cp_fm_diag_elpa_base 82 14.4 7.307 11.875 7.710 12.396 init_scf_run 11 5.9 0.000 0.001 10.485 10.486 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.485 10.486 make_images_data 4970 15.5 0.060 0.067 8.626 9.917 hybrid_alltoall_any 5155 16.4 0.428 1.950 7.429 9.372 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.680 8.904 multiply_cannon_metrocomm4 17395 15.5 0.062 0.070 3.364 8.833 mp_irecv_dv 49801 16.2 3.241 8.588 3.241 8.588 dbcsr_mm_accdrv_process 41158 16.2 4.344 5.008 8.155 8.367 grid_integrate_task_list 127 12.3 7.243 7.761 7.243 7.761 cp_fm_cholesky_decompose 22 10.9 7.676 7.702 7.676 7.702 pw_transfer 1535 11.6 0.084 0.103 7.355 7.471 cp_fm_upper_to_full 104 14.5 5.743 7.398 5.743 7.398 wfi_extrapolate 11 7.9 0.001 0.001 7.351 7.351 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 7.134 7.256 density_rs2pw 127 9.7 0.006 0.006 6.267 6.751 dbcsr_complete_redistribute 393 12.7 1.177 1.205 4.636 6.422 fft_wrap_pw1pw2_140 519 13.2 0.473 0.521 6.198 6.322 calculate_dm_sparse 127 9.5 0.001 0.001 5.731 5.819 rs_pw_transfer 1038 11.9 0.014 0.014 5.071 5.558 fft3d_ps 1281 14.7 2.673 2.875 5.413 5.483 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.464 5.470 grid_collocate_task_list 127 9.7 5.030 5.445 5.030 5.445 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.437 5.219 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.513 5.213 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.568 4.702 mp_alltoall_d11v 2401 14.1 4.120 4.651 4.120 4.651 potential_pw2rs 127 12.3 0.020 0.023 4.467 4.484 mp_sum_l 7804 13.0 3.143 4.465 3.143 4.465 mp_allgather_i34 2485 14.5 1.661 4.326 1.661 4.326 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.363 4.080 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.007 4.035 mp_alltoall_i22 712 14.1 1.963 3.845 1.963 3.845 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.713 3.714 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.371000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.272727, yerr=12.563229 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 4.387242E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 6026880 0.0% 0.0% 100.0% average stack size 0.0 0.0 1129.3 marketing flops 145.651870E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.141641E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1153224 MPI messages size (bytes): total size 2.039489E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.768511E+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 322096 36390305792 131072 < size <= 4194304 721976 792118951936 4194304 < size <= 16777216 70800 669922227920 16777216 < size 30960 541065216000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4127 56626. MP_Allreduce 11356 1063. MP_Sync 172 MP_Alltoall 1724 12509627. MP_ISendRecv 11868 75008. MP_Wait 28546 MP_comm_split 84 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 0.065 0.176 191.671 191.673 qs_mol_dyn_low 1 2.0 0.003 0.003 191.095 191.108 qs_forces 11 3.9 0.003 0.007 190.978 190.986 qs_energies 11 4.9 0.003 0.008 183.923 183.932 scf_env_do_scf 11 5.9 0.001 0.002 167.059 167.070 velocity_verlet 10 3.0 0.005 0.005 127.374 127.377 scf_env_do_scf_inner_loop 118 6.6 0.005 0.013 119.106 119.108 qs_scf_new_mos 118 7.6 0.001 0.001 82.770 83.079 qs_scf_loop_do_ot 118 8.6 0.001 0.001 82.769 83.079 dbcsr_multiply_generic 2529 12.6 0.192 0.198 80.666 81.286 ot_scf_mini 118 9.6 0.003 0.004 78.136 78.434 multiply_cannon 2529 13.6 0.559 0.586 55.319 58.359 multiply_cannon_loop 2529 14.6 1.194 1.225 51.415 52.948 init_scf_loop 11 6.9 0.001 0.005 47.818 47.820 prepare_preconditioner 11 7.9 0.000 0.000 43.697 43.724 make_preconditioner 11 8.9 0.000 0.002 43.697 43.724 ot_mini 118 10.6 0.001 0.001 43.151 43.417 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.255 42.332 multiply_cannon_multrec 30348 15.6 14.185 19.566 26.193 30.926 rebuild_ks_matrix 129 8.3 0.001 0.001 29.486 29.774 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.020 29.486 29.773 mp_waitall_1 155106 16.5 18.261 27.786 18.261 27.786 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.678 26.924 qs_ot_get_derivative 118 11.6 0.002 0.002 23.370 23.661 make_m2s 5058 13.6 0.096 0.101 20.862 21.966 make_images 5058 14.6 1.981 2.296 20.553 21.663 qs_ot_get_p 129 10.4 0.001 0.002 21.005 21.318 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 19.201 19.665 apply_single 129 13.6 0.001 0.001 19.201 19.665 ot_diis_step 118 11.6 0.018 0.019 19.649 19.651 cp_fm_cholesky_invert 11 10.9 17.017 17.028 17.017 17.028 qs_ot_p2m_diag 84 11.4 0.347 0.395 16.775 16.829 cp_fm_upper_to_full 106 14.7 11.236 16.609 11.236 16.609 multiply_cannon_metrocomm3 30348 15.6 0.047 0.049 6.690 15.701 cp_dbcsr_syevd 84 12.4 0.005 0.005 15.428 15.430 sum_up_and_integrate 129 10.3 0.142 0.155 14.409 14.437 integrate_v_rspace 129 11.3 0.003 0.004 14.267 14.299 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.526 13.559 calculate_rho_elec 129 8.7 0.176 0.192 13.526 13.558 dbcsr_complete_redistribute 397 12.7 1.520 1.632 9.205 13.053 make_images_data 5058 15.6 0.064 0.069 11.189 12.800 multiply_cannon_sync_h2d 30348 15.6 11.780 12.736 11.780 12.736 dbcsr_mm_accdrv_process 62780 16.2 7.551 8.497 11.582 12.148 cp_fm_diag_elpa 84 13.4 0.000 0.001 12.126 12.131 cp_fm_redistribute_end 84 14.4 2.093 12.032 2.112 12.037 cp_fm_diag_elpa_base 84 14.4 9.277 11.392 9.883 11.872 hybrid_alltoall_any 5245 16.5 0.528 2.199 10.065 11.821 copy_fm_to_dbcsr 210 11.7 0.001 0.002 7.729 11.541 init_scf_run 11 5.9 0.000 0.001 10.867 10.868 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.867 10.868 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.418 10.117 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.830 10.032 mp_alltoall_i22 720 14.1 5.651 9.474 5.651 9.474 pw_transfer 1559 11.6 0.085 0.100 8.119 8.188 cp_fm_cholesky_decompose 22 10.9 7.916 8.004 7.916 8.004 grid_integrate_task_list 129 12.3 7.561 7.972 7.561 7.972 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.892 7.966 wfi_extrapolate 11 7.9 0.001 0.001 7.722 7.722 multiply_cannon_metrocomm4 25290 15.6 0.075 0.086 2.787 7.294 mp_irecv_dv 76751 16.2 2.642 7.032 2.642 7.032 fft_wrap_pw1pw2_140 527 13.2 0.483 0.496 6.904 6.985 density_rs2pw 129 9.7 0.006 0.006 6.514 6.920 calculate_dm_sparse 129 9.5 0.001 0.001 6.310 6.390 fft3d_ps 1301 14.7 2.814 2.893 6.110 6.183 mp_alltoall_d11v 2429 14.1 5.200 5.770 5.200 5.770 grid_collocate_task_list 129 9.7 5.234 5.728 5.234 5.728 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.565 5.596 rs_pw_transfer 1054 12.0 0.013 0.015 4.890 5.304 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.440 4.534 potential_pw2rs 129 12.3 0.022 0.023 4.495 4.523 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.451 4.518 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 4.468 4.469 qs_ot_get_orbitals 118 10.6 0.001 0.001 4.315 4.381 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=191.673000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.363636, yerr=15.387555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239167967232 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239167967232 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 94.233007E+12 0.0% 0.0% 100.0% flops max/rank 5.910121E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383584 0.0% 0.0% 100.0% number of processed stacks 1976928 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.9 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 1.507230E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 242784 MPI messages size (bytes): total size 1.341806E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.526748E+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 115008 60297314304 4194304 < size <= 16777216 105840 554906419200 16777216 < size 20352 726592338480 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 9010 51. MP_Alltoall 9724 794507. MP_ISend 40420 2100460. MP_IRecv 40420 2099564. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 57630. MP_Allreduce 11104 1171. MP_Sync 88 MP_Alltoall 1724 18848066. MP_ISendRecv 7740 122880. MP_Wait 20114 MP_ISend 10760 423501. MP_IRecv 10760 423501. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.144 178.863 178.897 qs_mol_dyn_low 1 2.0 0.003 0.003 178.306 178.319 qs_forces 11 3.9 0.003 0.003 178.200 178.206 qs_energies 11 4.9 0.002 0.002 170.849 170.859 scf_env_do_scf 11 5.9 0.001 0.001 153.589 153.603 velocity_verlet 10 3.0 0.004 0.005 117.105 117.109 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 116.250 116.251 qs_scf_new_mos 118 7.6 0.001 0.001 79.765 79.854 qs_scf_loop_do_ot 118 8.6 0.001 0.001 79.764 79.853 dbcsr_multiply_generic 2529 12.6 0.182 0.187 76.231 76.890 ot_scf_mini 118 9.6 0.004 0.004 75.177 75.257 multiply_cannon 2529 13.6 0.587 0.629 55.831 59.864 multiply_cannon_loop 2529 14.6 0.454 0.464 50.992 52.247 ot_mini 118 10.6 0.001 0.001 40.983 41.082 init_scf_loop 11 6.9 0.000 0.000 37.182 37.183 mp_waitall_1 130746 16.6 27.610 34.916 27.610 34.916 prepare_preconditioner 11 7.9 0.000 0.000 33.286 33.312 make_preconditioner 11 8.9 0.000 0.000 33.286 33.311 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.999 31.264 rebuild_ks_matrix 129 8.3 0.001 0.001 29.128 29.298 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.017 29.128 29.297 qs_ks_update_qs_env 129 7.6 0.001 0.001 26.623 26.783 multiply_cannon_multrec 10116 15.6 10.452 16.419 17.827 22.816 qs_ot_get_p 129 10.4 0.001 0.001 20.931 21.069 qs_ot_get_derivative 118 11.6 0.002 0.002 20.656 20.733 multiply_cannon_metrocomm3 10116 15.6 0.022 0.024 13.025 20.692 ot_diis_step 118 11.6 0.020 0.023 20.247 20.248 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.737 20.122 apply_single 129 13.6 0.001 0.001 19.737 20.122 cp_fm_cholesky_invert 11 10.9 19.114 19.122 19.114 19.122 make_m2s 5058 13.6 0.067 0.071 16.553 18.888 make_images 5058 14.6 2.343 2.922 16.245 18.570 qs_ot_p2m_diag 84 11.4 0.502 0.508 16.887 16.904 cp_dbcsr_syevd 84 12.4 0.005 0.005 15.687 15.688 sum_up_and_integrate 129 10.3 0.181 0.191 14.407 14.457 integrate_v_rspace 129 11.3 0.004 0.004 14.225 14.286 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.504 13.528 calculate_rho_elec 129 8.7 0.260 0.271 13.504 13.527 make_images_data 5058 15.6 0.052 0.061 9.966 12.697 cp_fm_diag_elpa 84 13.4 0.000 0.000 12.465 12.473 cp_fm_diag_elpa_base 84 14.4 12.203 12.297 12.457 12.464 hybrid_alltoall_any 5245 16.5 0.785 3.533 9.757 12.400 multiply_cannon_sync_h2d 10116 15.6 11.546 12.265 11.546 12.265 init_scf_run 11 5.9 0.000 0.001 10.397 10.397 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.397 10.397 multiply_cannon_metrocomm1 10116 15.6 0.029 0.030 4.919 8.990 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 8.407 8.463 cp_fm_cholesky_decompose 22 10.9 8.311 8.407 8.311 8.407 grid_integrate_task_list 129 12.3 7.870 8.209 7.870 8.209 pw_transfer 1559 11.6 0.084 0.093 8.080 8.093 fft_wrap_pw1pw2 1301 12.7 0.010 0.010 7.857 7.876 dbcsr_mm_accdrv_process 20934 16.1 2.644 3.315 7.006 7.693 wfi_extrapolate 11 7.9 0.001 0.001 7.437 7.437 mp_allgather_i34 2529 14.6 3.019 7.256 3.019 7.256 fft_wrap_pw1pw2_140 527 13.2 0.507 0.530 6.844 6.869 density_rs2pw 129 9.7 0.005 0.006 6.236 6.572 calculate_dm_sparse 129 9.5 0.001 0.001 6.258 6.362 fft3d_ps 1301 14.7 2.745 2.810 6.001 6.041 dbcsr_complete_redistribute 397 12.7 2.140 2.232 5.509 5.883 grid_collocate_task_list 129 9.7 5.522 5.811 5.522 5.811 multiply_cannon_metrocomm4 7587 15.6 0.024 0.027 1.893 5.798 mp_irecv_dv 29102 15.9 1.856 5.708 1.856 5.708 mp_alltoall_d11v 2429 14.1 5.109 5.682 5.109 5.682 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.259 5.270 qs_energies_init_hamiltonians 11 5.9 0.002 0.002 5.248 5.248 rs_pw_transfer 1054 12.0 0.013 0.013 4.319 4.655 potential_pw2rs 129 12.3 0.027 0.028 4.282 4.300 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.157 4.206 copy_fm_to_dbcsr 210 11.7 0.002 0.002 3.618 3.969 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.868 3.920 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.564 3.871 copy_dbcsr_to_fm 187 11.8 0.004 0.004 3.690 3.746 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.570 3.594 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=178.897000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1413.909091, yerr=54.366525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 3.003470E+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 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_ISendRecv 3612 218624. MP_Wait 11682 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 0.064 0.111 287.123 287.142 qs_mol_dyn_low 1 2.0 0.003 0.004 286.552 286.564 qs_forces 11 3.9 0.003 0.003 286.463 286.466 qs_energies 11 4.9 0.002 0.002 277.852 277.867 scf_env_do_scf 11 5.9 0.001 0.001 256.309 256.328 velocity_verlet 10 3.0 0.002 0.002 208.125 208.132 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 132.562 132.564 init_scf_loop 11 6.9 0.000 0.000 123.491 123.493 prepare_preconditioner 11 7.9 0.000 0.000 118.776 118.806 make_preconditioner 11 8.9 0.000 0.000 118.776 118.806 make_full_inverse_cholesky 11 9.9 0.000 0.000 95.092 115.913 qs_scf_new_mos 118 7.6 0.001 0.001 90.358 90.465 qs_scf_loop_do_ot 118 8.6 0.001 0.001 90.358 90.464 ot_scf_mini 118 9.6 0.004 0.004 85.549 85.595 dbcsr_multiply_generic 2535 12.6 0.214 0.222 83.153 83.702 cp_fm_upper_to_full 106 14.8 52.626 75.282 52.626 75.282 multiply_cannon 2535 13.6 0.707 0.766 58.731 59.336 multiply_cannon_loop 2535 14.6 0.478 0.482 55.133 56.433 ot_mini 118 10.6 0.001 0.001 44.272 44.311 dbcsr_complete_redistribute 397 12.7 4.031 4.081 29.290 41.871 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.927 38.510 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.637 36.098 mp_alltoall_i22 720 14.1 21.452 34.177 21.452 34.177 rebuild_ks_matrix 129 8.3 0.001 0.001 33.709 33.752 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 33.708 33.751 cp_fm_cholesky_invert 11 10.9 33.304 33.311 33.304 33.311 mp_waitall_1 106626 16.7 28.062 31.975 28.062 31.975 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.500 31.547 qs_ot_get_p 129 10.4 0.001 0.001 25.725 25.732 qs_ot_get_derivative 118 11.6 0.002 0.002 24.191 24.234 qs_ot_p2m_diag 84 11.4 0.889 0.895 21.694 21.725 make_m2s 5070 13.6 0.077 0.080 19.834 20.710 make_images 5070 14.6 3.810 4.008 19.353 20.231 ot_diis_step 118 11.6 0.022 0.023 20.052 20.052 cp_dbcsr_syevd 84 12.4 0.006 0.006 19.965 19.966 multiply_cannon_metrocomm3 10140 15.6 0.023 0.023 18.685 19.882 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.245 19.438 apply_single 129 13.6 0.001 0.001 19.244 19.438 multiply_cannon_multrec 10140 15.6 10.628 12.369 18.095 18.171 cp_fm_diag_elpa 84 13.4 0.000 0.000 16.746 16.747 cp_fm_diag_elpa_base 84 14.4 12.328 13.937 16.742 16.743 sum_up_and_integrate 129 10.3 0.324 0.325 15.762 15.857 multiply_cannon_sync_h2d 10140 15.6 15.780 15.789 15.780 15.789 integrate_v_rspace 129 11.3 0.004 0.004 15.438 15.533 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.008 15.027 calculate_rho_elec 129 8.7 0.487 0.488 15.007 15.026 hybrid_alltoall_any 5257 16.5 1.308 3.057 10.662 12.490 make_images_data 5070 15.6 0.061 0.065 10.526 12.235 init_scf_run 11 5.9 0.000 0.001 11.654 11.654 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.654 11.654 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.503 9.541 dbcsr_mm_accdrv_process 20958 16.1 3.891 5.734 7.229 9.058 cp_fm_cholesky_decompose 22 10.9 8.910 8.948 8.910 8.948 grid_integrate_task_list 129 12.3 8.602 8.770 8.602 8.770 wfi_extrapolate 11 7.9 0.001 0.001 8.574 8.574 pw_transfer 1559 11.6 0.091 0.092 8.313 8.322 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 8.077 8.087 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 7.884 7.885 fft_wrap_pw1pw2_140 527 13.2 0.544 0.545 7.172 7.186 mp_alltoall_d11v 2429 14.1 6.933 7.073 6.933 7.073 calculate_dm_sparse 129 9.5 0.001 0.001 6.643 6.723 grid_collocate_task_list 129 9.7 6.386 6.418 6.386 6.418 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.272 6.383 fft3d_ps 1301 14.7 2.780 2.789 6.107 6.116 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.041 6.108 density_rs2pw 129 9.7 0.005 0.005 6.001 6.032 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=287.142000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2661.818182, yerr=179.053480 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 1.260274E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 MP_ISendRecv 570 19200. MP_Wait 1302 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 0.018 0.082 85.693 85.694 qs_energies 1 2.0 0.000 0.000 85.196 85.201 ls_scf 1 3.0 0.000 0.000 84.285 84.291 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.383 73.602 multiply_cannon 111 7.7 0.018 0.021 56.061 57.223 multiply_cannon_loop 111 8.7 0.205 0.224 52.455 53.829 ls_scf_main 1 4.0 0.000 0.000 52.728 52.732 density_matrix_trs4 2 5.0 0.002 0.003 47.315 47.390 ls_scf_init_scf 1 4.0 0.000 0.000 28.530 28.535 ls_scf_init_matrix_S 1 5.0 0.000 0.001 27.449 27.504 mp_waitall_1 11316 10.9 22.618 25.439 22.618 25.439 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.327 25.341 multiply_cannon_multrec 2664 9.7 8.218 8.905 15.590 17.332 multiply_cannon_sync_h2d 2664 9.7 13.833 15.346 13.833 15.346 make_m2s 222 7.7 0.008 0.010 13.703 14.251 make_images 222 8.7 0.098 0.109 13.682 14.233 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.389 12.236 make_images_data 222 9.7 0.004 0.005 8.146 9.073 hybrid_alltoall_any 227 10.6 0.214 1.831 6.862 9.002 dbcsr_mm_accdrv_process 4760 10.4 0.509 0.660 6.991 8.017 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.404 7.970 dbcsr_mm_accdrv_process_sort 4760 11.4 6.282 7.233 6.282 7.233 calculate_norms 4752 9.8 5.575 6.241 5.575 6.241 apply_matrix_preconditioner 6 5.3 0.000 0.002 4.945 5.141 mp_sum_l 807 5.4 3.106 4.799 3.106 4.799 make_images_sizes 222 9.7 0.000 0.000 1.103 3.924 mp_alltoall_i44 222 10.7 1.103 3.924 1.103 3.924 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.276 3.537 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.053 3.388 mp_irecv_dv 6231 10.9 2.036 3.358 2.036 3.358 arnoldi_extremal 4 6.8 0.000 0.000 3.208 3.233 arnoldi_normal_ev 4 7.8 0.001 0.005 3.208 3.233 build_subspace 16 8.4 0.009 0.012 3.116 3.118 ls_scf_post 1 4.0 0.000 0.000 3.027 3.032 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.882 dbcsr_special_finalize 555 9.7 0.004 0.006 2.103 2.736 dbcsr_merge_single_wm 555 10.7 0.467 0.609 2.096 2.727 make_images_pack 222 9.7 2.208 2.641 2.210 2.643 dbcsr_matrix_vector_mult 304 9.0 0.004 0.010 2.312 2.550 dbcsr_sort_data 658 11.4 1.888 2.455 1.888 2.455 dbcsr_matrix_vector_mult_local 304 10.0 2.063 2.453 2.065 2.455 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.234 2.339 buffer_matrices_ensure_size 222 8.7 1.815 2.254 1.815 2.254 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.751 1.757 rebuild_ks_matrix 3 7.3 0.000 0.000 1.741 1.747 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.009 1.741 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.694000, 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/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 2.091782E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_ISendRecv 282 57600. MP_Wait 828 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 0.039 0.046 91.085 91.086 qs_energies 1 2.0 0.000 0.000 90.641 90.650 ls_scf 1 3.0 0.000 0.000 89.329 89.339 dbcsr_multiply_generic 111 6.7 0.015 0.017 75.402 75.702 multiply_cannon 111 7.7 0.030 0.039 53.070 57.212 ls_scf_main 1 4.0 0.000 0.000 54.981 54.991 multiply_cannon_loop 111 8.7 0.117 0.123 49.829 53.302 density_matrix_trs4 2 5.0 0.002 0.003 49.358 49.543 ls_scf_init_scf 1 4.0 0.000 0.000 30.841 30.842 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.703 29.798 mp_waitall_1 9246 10.9 20.988 29.605 20.988 29.605 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.294 27.311 multiply_cannon_multrec 1332 9.7 13.199 16.793 22.479 27.240 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 11.469 20.276 make_m2s 222 7.7 0.006 0.007 15.483 16.129 make_images 222 8.7 1.580 1.976 15.453 16.101 dbcsr_mm_accdrv_process 4041 10.4 0.288 0.455 8.872 10.469 dbcsr_mm_accdrv_process_sort 4041 11.4 8.465 10.013 8.465 10.013 make_images_data 222 9.7 0.004 0.004 8.935 9.870 hybrid_alltoall_any 227 10.6 0.523 2.488 8.309 9.547 mp_sum_l 807 5.4 5.554 8.544 5.554 8.544 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.249 7.706 mp_irecv_dv 3311 11.0 3.229 7.631 3.229 7.631 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.321 6.832 calculate_norms 2376 9.8 6.010 6.699 6.010 6.699 multiply_cannon_sync_h2d 1332 9.7 4.831 5.974 4.831 5.974 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.015 5.210 arnoldi_extremal 4 6.8 0.000 0.000 4.741 4.768 arnoldi_normal_ev 4 7.8 0.001 0.004 4.741 4.768 build_subspace 16 8.4 0.014 0.021 4.482 4.486 ls_scf_post 1 4.0 0.000 0.000 3.507 3.516 dbcsr_matrix_vector_mult 304 9.0 0.005 0.017 3.179 3.400 ls_scf_store_result 1 5.0 0.000 0.000 3.219 3.319 dbcsr_matrix_vector_mult_local 304 10.0 2.780 3.256 2.782 3.257 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.209 3.111 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.524 2.618 mp_allgather_i34 111 8.7 0.940 2.450 0.940 2.450 make_images_pack 222 9.7 2.024 2.413 2.027 2.415 dbcsr_sort_data 436 11.2 1.824 2.048 1.824 2.048 dbcsr_data_new 4174 10.1 1.608 1.840 1.608 1.840 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.086000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1697.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 2.760040E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_ISendRecv 186 57600. MP_Wait 732 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 0.066 0.182 95.071 95.073 qs_energies 1 2.0 0.000 0.000 94.200 94.222 ls_scf 1 3.0 0.000 0.001 92.805 92.827 dbcsr_multiply_generic 111 6.7 0.016 0.017 77.417 77.686 ls_scf_main 1 4.0 0.001 0.012 58.126 58.130 multiply_cannon 111 7.7 0.043 0.114 52.811 56.659 multiply_cannon_loop 111 8.7 0.100 0.106 49.218 53.737 density_matrix_trs4 2 5.0 0.002 0.003 52.260 52.444 mp_waitall_1 7374 11.0 23.970 34.524 23.970 34.524 ls_scf_init_scf 1 4.0 0.000 0.003 31.056 31.059 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.807 29.888 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.332 27.348 multiply_cannon_multrec 888 9.7 12.591 15.066 21.154 24.263 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.834 23.781 make_m2s 222 7.7 0.006 0.007 17.410 18.703 make_images 222 8.7 1.976 2.287 17.372 18.664 hybrid_alltoall_any 227 10.6 0.621 2.861 9.423 11.232 make_images_data 222 9.7 0.003 0.004 9.904 11.005 dbcsr_mm_accdrv_process 3754 10.4 0.254 0.447 8.110 9.271 mp_sum_l 807 5.4 5.748 9.239 5.748 9.239 dbcsr_mm_accdrv_process_sort 3754 11.4 7.734 8.858 7.734 8.858 multiply_cannon_sync_h2d 888 9.7 6.012 7.439 6.012 7.439 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.471 7.157 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.395 7.140 mp_irecv_dv 2335 11.1 2.456 7.103 2.456 7.103 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.848 7.067 arnoldi_extremal 4 6.8 0.000 0.000 5.211 5.233 arnoldi_normal_ev 4 7.8 0.001 0.005 5.211 5.233 apply_matrix_preconditioner 6 5.3 0.002 0.002 4.986 5.215 build_subspace 16 8.4 0.014 0.020 4.900 4.906 calculate_norms 1584 9.8 4.272 4.631 4.272 4.631 mp_allgather_i34 111 8.7 1.430 3.992 1.430 3.992 dbcsr_matrix_vector_mult 304 9.0 0.005 0.016 3.518 3.813 ls_scf_post 1 4.0 0.003 0.023 3.623 3.647 dbcsr_matrix_vector_mult_local 304 10.0 3.065 3.604 3.067 3.606 ls_scf_store_result 1 5.0 0.000 0.000 3.360 3.444 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.734 2.830 make_images_sizes 222 9.7 0.000 0.000 1.215 2.316 mp_alltoall_i44 222 10.7 1.214 2.316 1.214 2.316 dbcsr_sort_data 325 11.1 1.850 2.139 1.850 2.139 make_images_pack 222 9.7 1.811 2.123 1.814 2.125 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.945 1.947 rebuild_ks_matrix 3 7.3 0.000 0.000 1.927 1.929 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.010 1.927 1.929 dbcsr_data_release 9322 10.9 1.290 1.918 1.290 1.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=95.073000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2189.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 3.368215E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_ISendRecv 138 86400. MP_Wait 600 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 0.079 0.165 97.172 97.174 qs_energies 1 2.0 0.000 0.000 96.249 96.262 ls_scf 1 3.0 0.000 0.001 94.613 94.627 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.358 78.628 ls_scf_main 1 4.0 0.000 0.001 58.766 58.767 multiply_cannon 111 7.7 0.047 0.109 51.666 55.801 density_matrix_trs4 2 5.0 0.002 0.003 52.749 52.895 multiply_cannon_loop 111 8.7 0.115 0.126 46.540 50.170 ls_scf_init_scf 1 4.0 0.000 0.003 32.591 32.593 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.320 31.395 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.863 28.875 mp_waitall_1 6438 11.0 22.626 28.660 22.626 28.660 multiply_cannon_multrec 1332 9.7 14.078 17.492 21.963 24.315 make_m2s 222 7.7 0.007 0.008 20.852 22.341 make_images 222 8.7 3.139 3.602 20.802 22.293 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.264 17.201 make_images_data 222 9.7 0.004 0.005 11.616 13.245 hybrid_alltoall_any 227 10.6 0.798 3.827 10.884 12.917 dbcsr_mm_accdrv_process 3641 10.4 0.187 0.375 7.530 9.088 dbcsr_mm_accdrv_process_sort 3641 11.4 7.159 8.664 7.159 8.664 mp_sum_l 807 5.4 4.319 8.250 4.319 8.250 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.311 6.548 multiply_cannon_sync_h2d 1332 9.7 5.523 6.044 5.523 6.044 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.090 5.947 mp_irecv_dv 3229 10.9 2.068 5.872 2.068 5.872 arnoldi_extremal 4 6.8 0.000 0.000 5.272 5.287 arnoldi_normal_ev 4 7.8 0.001 0.005 5.272 5.287 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.566 5.035 build_subspace 16 8.4 0.015 0.021 4.936 4.946 apply_matrix_preconditioner 6 5.3 0.001 0.001 4.581 4.802 mp_allgather_i34 111 8.7 2.240 4.710 2.240 4.710 calculate_norms 2376 9.8 4.195 4.563 4.195 4.563 dbcsr_matrix_vector_mult 304 9.0 0.006 0.017 3.622 3.913 dbcsr_matrix_vector_mult_local 304 10.0 3.244 3.724 3.246 3.726 dbcsr_sort_data 658 11.4 3.033 3.400 3.033 3.400 ls_scf_post 1 4.0 0.000 0.001 3.256 3.269 dbcsr_special_finalize 555 9.7 0.006 0.007 2.797 3.250 dbcsr_merge_single_wm 555 10.7 0.536 0.670 2.788 3.242 ls_scf_store_result 1 5.0 0.000 0.000 2.986 3.061 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.874 2.933 dbcsr_data_release 10477 10.7 1.572 2.408 1.572 2.408 dbcsr_finalize 304 7.8 0.049 0.061 1.787 1.979 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.174000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2744.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 4.617716E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_ISendRecv 90 115200. MP_Wait 573 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 0.060 0.068 98.588 98.589 qs_energies 1 2.0 0.000 0.000 97.745 97.750 ls_scf 1 3.0 0.000 0.000 95.833 95.838 dbcsr_multiply_generic 111 6.7 0.017 0.018 77.421 77.617 ls_scf_main 1 4.0 0.000 0.000 61.777 61.778 multiply_cannon 111 7.7 0.102 0.212 55.103 60.429 density_matrix_trs4 2 5.0 0.002 0.003 54.843 54.955 multiply_cannon_loop 111 8.7 0.070 0.080 50.480 52.268 mp_waitall_1 5481 11.0 25.708 30.656 25.708 30.656 ls_scf_init_scf 1 4.0 0.000 0.000 30.450 30.455 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.281 29.320 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.116 27.124 multiply_cannon_multrec 444 9.7 14.075 16.678 21.148 23.952 make_m2s 222 7.7 0.004 0.005 17.353 19.947 make_images 222 8.7 3.722 4.446 17.292 19.887 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.022 16.068 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 5.913 14.895 make_images_data 222 9.7 0.003 0.004 9.569 12.120 hybrid_alltoall_any 227 10.6 0.791 3.763 9.312 11.834 multiply_cannon_sync_h2d 444 9.7 6.547 8.107 6.547 8.107 dbcsr_mm_accdrv_process 3003 10.4 0.175 0.342 6.777 7.924 dbcsr_mm_accdrv_process_sort 3003 11.4 6.463 7.575 6.463 7.575 mp_allgather_i34 111 8.7 2.853 7.035 2.853 7.035 arnoldi_extremal 4 6.8 0.000 0.000 5.979 5.989 arnoldi_normal_ev 4 7.8 0.002 0.005 5.979 5.989 build_subspace 16 8.4 0.015 0.020 5.582 5.593 mp_sum_l 807 5.4 3.016 5.006 3.016 5.006 apply_matrix_preconditioner 6 5.3 0.001 0.002 4.560 4.697 dbcsr_matrix_vector_mult 304 9.0 0.007 0.016 4.239 4.449 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.653 4.320 mp_irecv_dv 1241 11.2 1.634 4.297 1.634 4.297 dbcsr_matrix_vector_mult_local 304 10.0 3.730 4.174 3.732 4.176 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.102 3.885 calculate_norms 792 9.8 3.555 3.726 3.555 3.726 ls_scf_post 1 4.0 0.000 0.000 3.606 3.611 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.366 3.459 ls_scf_store_result 1 5.0 0.000 0.000 3.379 3.433 make_images_sizes 222 9.7 0.000 0.000 1.141 3.199 mp_alltoall_i44 222 10.7 1.141 3.199 1.141 3.199 dbcsr_finalize 304 7.8 0.062 0.077 2.195 2.333 dbcsr_merge_all 275 8.9 0.473 0.536 2.044 2.162 dbcsr_data_release 10123 10.8 1.326 2.013 1.326 2.013 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=98.589000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3667.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/8d96446bea90f37d3e25c2492dcc34fe4f653cd4_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 8.732586E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_ISendRecv 84 732600. MP_Wait 309 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 0.085 0.102 108.020 108.021 qs_energies 1 2.0 0.000 0.000 106.605 106.616 ls_scf 1 3.0 0.000 0.000 103.723 103.734 dbcsr_multiply_generic 111 6.7 0.023 0.026 77.847 77.966 ls_scf_main 1 4.0 0.000 0.000 65.924 65.925 density_matrix_trs4 2 5.0 0.002 0.003 57.075 57.141 multiply_cannon 111 7.7 0.130 0.246 50.315 52.119 multiply_cannon_loop 111 8.7 0.067 0.070 46.903 47.834 ls_scf_init_scf 1 4.0 0.000 0.000 34.117 34.118 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.787 32.799 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.950 29.958 mp_waitall_1 4569 11.1 22.646 26.525 22.646 26.525 make_m2s 222 7.7 0.005 0.005 23.956 24.880 make_images 222 8.7 4.590 4.976 23.850 24.771 multiply_cannon_multrec 444 9.7 17.920 18.564 22.565 23.203 hybrid_alltoall_any 227 10.6 1.659 3.622 12.910 15.822 make_images_data 222 9.7 0.003 0.003 13.172 15.578 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.877 11.434 multiply_cannon_sync_h2d 444 9.7 8.850 8.896 8.850 8.896 arnoldi_extremal 4 6.8 0.000 0.000 7.424 7.433 arnoldi_normal_ev 4 7.8 0.003 0.009 7.424 7.433 build_subspace 16 8.4 0.026 0.036 6.860 6.868 dbcsr_matrix_vector_mult 304 9.0 0.009 0.025 5.525 5.657 dbcsr_matrix_vector_mult_local 304 10.0 5.107 5.372 5.109 5.374 apply_matrix_preconditioner 6 5.3 0.001 0.001 5.033 5.295 ls_scf_dm_to_ks 2 5.0 0.000 0.000 4.794 4.884 dbcsr_mm_accdrv_process 1814 10.4 0.189 0.345 4.475 4.598 dbcsr_mm_accdrv_process_sort 1814 11.4 4.158 4.291 4.158 4.291 ls_scf_post 1 4.0 0.000 0.000 3.683 3.693 make_images_sizes 222 9.7 0.000 0.000 1.443 3.576 mp_alltoall_i44 222 10.7 1.443 3.576 1.443 3.576 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.439 mp_allgather_i34 111 8.7 1.050 3.314 1.050 3.314 calculate_norms 792 9.8 3.236 3.277 3.236 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.083 3.118 dbcsr_complete_redistribute 5 7.6 1.446 1.505 2.784 2.902 dbcsr_merge_all 275 8.9 0.892 0.917 2.869 2.898 dbcsr_data_release 12724 10.6 2.344 2.864 2.344 2.864 qs_energies_init_hamiltonians 1 3.0 0.002 0.002 2.851 2.851 matrix_ls_to_qs 2 6.0 0.000 0.000 2.426 2.535 dbcsr_sort_data 325 11.1 2.441 2.493 2.441 2.493 dbcsr_new_transposed 4 7.5 0.244 0.254 2.325 2.335 dbcsr_frobenius_norm 74 6.6 2.057 2.133 2.182 2.224 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.201 2.203 dbcsr_add_d 103 6.2 0.000 0.000 2.132 2.203 dbcsr_add_anytype 103 7.2 0.859 0.892 2.132 2.202 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.021000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6777.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 8d96446bea90f37d3e25c2492dcc34fe4f653cd4 Summary: empty Status: OK