=== 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: 1501a5cde42d9e664f251d02093fe2fc81c3abfc ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/01 job id: 47541074 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/02 job id: 47541076 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/03 job id: 47541078 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/04 job id: 47541081 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/05 job id: 47541083 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/06 job id: 47541085 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/07 job id: 47541087 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/08 job id: 47541089 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/09 job id: 47541090 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/10 job id: 47541091 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/11 job id: 47541092 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/12 job id: 47541094 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/13 job id: 47541095 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/14 job id: 47541097 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/15 job id: 47541099 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/16 job id: 47541100 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/17 job id: 47541101 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/18 job id: 47541104 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/19 job id: 47541105 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/20 job id: 47541106 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/21 job id: 47541108 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/22 job id: 47541109 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/23 job id: 47541111 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/24 job id: 47541114 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/25 job id: 47541117 --- 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/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/26 job id: 47541118 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/27 job id: 47541122 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.036 134.533 134.534 farming_run 1 2.0 133.933 133.935 134.503 134.507 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.466986E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.025 115.285 115.286 qs_energies 1 2.0 0.000 0.000 115.019 115.019 mp2_main 1 3.0 0.000 0.000 112.854 112.854 mp2_gpw_main 1 4.0 0.019 0.023 111.857 111.858 mp2_ri_gpw_compute_in 1 5.0 0.172 0.173 92.858 93.205 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.868 55.215 mp2_eri_3c_integrate_gpw 272 7.0 0.154 0.169 41.317 46.420 get_2c_integrals 1 6.0 0.008 0.009 37.179 37.816 integrate_v_rspace 273 8.0 0.436 0.450 24.736 29.659 pw_transfer 6555 10.6 0.373 0.389 27.456 28.012 fft_wrap_pw1pw2 5465 11.4 0.046 0.048 26.088 26.480 grid_integrate_task_list 273 9.0 20.549 25.938 20.549 25.938 fft_wrap_pw1pw2_100 2178 12.4 1.231 1.445 23.607 24.026 compute_2c_integrals 1 7.0 0.002 0.002 19.732 19.733 compute_2c_integrals_loop_lm 1 8.0 0.002 0.003 18.976 19.457 mp2_eri_2c_integrate_gpw 1 9.0 2.371 2.446 18.974 19.454 rpa_ri_compute_en 1 5.0 0.019 0.019 18.893 19.025 cp_fm_cholesky_decompose 12 8.2 17.378 18.001 17.378 18.001 cholesky_decomp 1 7.0 0.000 0.000 16.282 16.912 fft3d_s 5443 13.4 16.144 16.492 16.166 16.514 ao_to_mo_and_store_B_mult_1 272 7.0 10.718 15.293 10.718 15.293 calculate_wavefunction 272 8.0 5.401 5.543 12.512 13.142 rpa_num_int 1 6.0 0.000 0.000 10.634 10.635 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.594 10.627 calc_mat_Q 8 8.0 0.000 0.000 9.467 9.571 contract_S_to_Q 8 9.0 0.000 0.000 8.887 8.990 calc_potential_gpw 544 9.5 0.005 0.005 8.278 8.681 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.275 8.579 parallel_gemm_fm 14 9.1 0.000 0.000 8.483 8.569 parallel_gemm_fm_cosma 14 10.1 8.483 8.569 8.483 8.569 potential_pw2rs 545 10.0 0.108 0.109 7.716 8.348 create_integ_mat 1 6.0 0.006 0.008 7.816 7.816 collocate_single_gaussian 272 10.0 0.039 0.042 7.516 7.777 array2fm 1 7.0 0.000 0.000 6.912 7.245 pw_scatter_s 2720 13.7 4.420 4.567 4.420 4.567 pw_gather_s 2722 13.2 3.882 4.266 3.882 4.266 array2fm_buffer_send 1 8.0 2.988 3.146 2.988 3.146 pw_poisson_solve 545 10.5 1.103 1.146 2.179 2.377 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.857932, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2809.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.039 401.026 401.027 farming_run 1 2.0 399.672 399.685 400.975 400.980 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.227526E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 28402590. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.031 209.809 209.810 qs_energies 1 2.0 0.000 0.000 209.551 209.573 scf_env_do_scf 1 3.0 0.000 0.000 106.505 106.505 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.464 105.473 rebuild_ks_matrix 4 6.0 0.000 0.000 105.463 105.472 qs_ks_build_kohn_sham_matrix 4 7.0 0.059 0.068 105.463 105.472 hfx_ks_matrix 4 8.0 0.001 0.001 105.014 105.018 integrate_four_center 4 9.0 0.144 0.461 105.013 105.018 mp2_main 1 3.0 0.000 0.000 102.717 102.739 mp2_gpw_main 1 4.0 0.031 0.044 101.681 101.704 integrate_four_center_main 4 10.0 0.095 0.512 96.574 99.164 integrate_four_center_bin 265 11.0 96.480 99.159 96.480 99.159 init_scf_loop 1 4.0 0.000 0.000 92.304 92.305 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.575 75.706 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.099 55.230 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.159 41.652 46.809 integrate_v_rspace 95 8.0 0.398 0.570 28.087 33.089 pw_transfer 2240 10.6 0.144 0.173 29.982 30.443 fft_wrap_pw1pw2 1868 11.4 0.018 0.022 28.985 29.386 ao_to_mo_and_store_B_mult_1 91 7.0 10.773 28.995 10.773 28.995 mp2_ri_gpw_compute_en 1 5.0 0.057 0.075 26.958 28.797 grid_integrate_task_list 95 9.0 23.313 28.506 23.313 28.506 fft_wrap_pw1pw2_100 730 12.4 1.289 1.474 26.672 27.038 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.834 1.907 25.184 25.193 get_2c_integrals 1 6.0 0.000 0.000 20.389 20.413 compute_2c_integrals 1 7.0 0.002 0.003 19.377 19.379 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.917 19.216 mp2_eri_2c_integrate_gpw 1 9.0 1.738 1.893 18.916 19.215 fft3d_s 1823 13.4 18.426 18.930 18.439 18.943 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.199 14.199 calculate_wavefunction 91 8.0 2.013 2.042 9.745 10.002 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.594 8.823 9.463 potential_pw2rs 186 10.0 0.033 0.035 8.711 9.325 local_gemm 172 8.0 8.265 8.890 8.265 8.890 mp2_ri_gpw_compute_en_comm 22 7.0 0.496 0.516 8.133 8.838 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.274 8.631 calc_potential_gpw 182 9.5 0.002 0.002 7.931 8.155 collocate_single_gaussian 91 10.0 0.017 0.022 7.902 8.102 mp_sendrecv_dm3 2068 8.0 6.173 6.850 6.173 6.850 mp2_ri_gpw_compute_en_ener 172 7.0 6.349 6.433 6.349 6.433 pw_gather_s 912 13.2 4.928 5.568 4.928 5.568 mp_sync 37 10.5 3.073 5.158 3.073 5.158 pw_scatter_s 910 13.7 3.935 4.287 3.935 4.287 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.676428, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1514.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 451.518464E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.040 57.673 57.674 qs_mol_dyn_low 1 2.0 0.004 0.012 56.895 57.188 qs_forces 11 3.9 0.003 0.004 56.463 56.464 qs_energies 11 4.9 0.001 0.002 54.856 54.871 scf_env_do_scf 11 5.9 0.001 0.004 48.083 48.083 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 45.685 45.685 qs_scf_new_mos 108 7.5 0.000 0.001 34.504 34.812 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.503 34.812 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.156 34.564 ot_scf_mini 108 9.5 0.002 0.002 32.774 32.964 velocity_verlet 10 3.0 0.003 0.010 28.133 28.143 multiply_cannon 2286 13.5 0.186 0.195 26.271 28.021 multiply_cannon_loop 2286 14.5 1.480 1.567 25.372 27.104 ot_mini 108 10.5 0.001 0.001 19.535 19.781 qs_ot_get_derivative 108 11.5 0.001 0.001 16.516 16.692 mp_waitall_1 245248 16.5 8.780 14.952 8.780 14.952 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.984 12.962 multiply_cannon_multrec 54864 15.5 4.238 6.537 7.764 11.216 qs_ot_get_p 119 10.4 0.001 0.001 8.539 8.817 rebuild_ks_matrix 119 8.3 0.000 0.000 8.497 8.620 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.032 8.497 8.620 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.468 7.584 mp_sum_l 7287 12.8 5.345 7.157 5.345 7.157 multiply_cannon_sync_h2d 54864 15.5 5.869 6.905 5.869 6.905 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.832 6.261 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.625 5.662 init_scf_run 11 5.9 0.000 0.001 5.409 5.410 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.409 5.410 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.316 5.402 sum_up_and_integrate 119 10.3 0.012 0.014 4.976 4.996 integrate_v_rspace 119 11.3 0.002 0.003 4.964 4.985 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.866 4.981 calculate_rho_elec 119 8.7 0.012 0.017 4.866 4.980 dbcsr_mm_accdrv_process 76910 16.1 1.129 1.788 3.447 4.806 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.790 4.790 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.501 4.503 cp_fm_redistribute_end 50 14.0 2.292 4.464 2.304 4.470 cp_fm_diag_elpa_base 50 14.0 2.155 4.334 2.162 4.345 rs_pw_transfer 974 11.9 0.012 0.013 3.276 3.404 calculate_dm_sparse 119 9.5 0.000 0.001 3.114 3.276 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.935 3.140 apply_single 119 13.6 0.000 0.000 2.934 3.140 calculate_first_density_matrix 1 7.0 0.000 0.001 2.967 2.974 jit_kernel_multiply 13 15.8 2.259 2.931 2.259 2.931 density_rs2pw 119 9.7 0.004 0.004 2.754 2.861 multiply_cannon_metrocomm1 54864 15.5 0.053 0.058 1.777 2.790 ot_diis_step 108 11.5 0.006 0.006 2.742 2.743 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.638 2.640 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.469 2.536 init_scf_loop 11 6.9 0.000 0.000 2.369 2.370 wfi_extrapolate 11 7.9 0.001 0.001 2.360 2.361 pw_transfer 1439 11.6 0.052 0.057 2.241 2.347 potential_pw2rs 119 12.3 0.004 0.004 2.276 2.325 make_m2s 4572 13.5 0.053 0.055 2.234 2.321 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.262 2.313 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 2.166 2.273 make_images 4572 14.5 0.133 0.138 2.152 2.238 acc_transpose_blocks 54864 15.5 0.228 0.250 1.749 2.207 grid_integrate_task_list 119 12.3 2.016 2.138 2.016 2.138 mp_sum_d 4135 12.0 1.393 1.976 1.393 1.976 fft3d_ps 1201 14.6 0.373 0.484 1.808 1.919 fft_wrap_pw1pw2_140 487 13.2 0.185 0.203 1.646 1.753 mp_alltoall_d11v 2130 13.8 1.556 1.689 1.556 1.689 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.567 1.591 grid_collocate_task_list 119 9.7 1.351 1.440 1.351 1.440 mp_waitany 12084 13.8 1.298 1.426 1.298 1.426 prepare_preconditioner 11 7.9 0.000 0.000 1.199 1.241 make_preconditioner 11 8.9 0.000 0.000 1.199 1.241 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.139 1.189 dbcsr_dot_sd 1205 11.9 0.049 0.060 0.786 1.188 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=57.674000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=430.545455, yerr=0.655555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 487.514112E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 913371. MP_SendRecv 16779 37093. MP_ISendRecv 16779 37093. MP_Wait 23539 MP_comm_split 50 MP_ISend 5720 128509. MP_IRecv 5720 128509. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.060 44.341 44.343 qs_mol_dyn_low 1 2.0 0.003 0.004 43.600 43.642 qs_forces 11 3.9 0.002 0.003 43.299 43.299 qs_energies 11 4.9 0.009 0.059 41.507 41.517 scf_env_do_scf 11 5.9 0.001 0.002 35.409 35.412 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 32.547 32.547 dbcsr_multiply_generic 2286 12.5 0.102 0.105 23.529 24.125 qs_scf_new_mos 108 7.5 0.001 0.001 22.219 22.477 qs_scf_loop_do_ot 108 8.5 0.001 0.001 22.218 22.477 ot_scf_mini 108 9.5 0.003 0.004 21.265 21.440 velocity_verlet 10 3.0 0.001 0.001 20.365 20.371 multiply_cannon 2286 13.5 0.212 0.227 17.434 19.359 multiply_cannon_loop 2286 14.5 0.901 0.975 16.112 18.245 ot_mini 108 10.5 0.001 0.001 13.079 13.319 mp_waitall_1 200699 16.5 6.793 12.381 6.793 12.381 qs_ot_get_derivative 108 11.5 0.001 0.001 10.501 10.675 multiply_cannon_metrocomm3 27432 15.5 0.067 0.069 4.616 10.490 multiply_cannon_multrec 27432 15.5 1.980 4.440 6.287 9.367 rebuild_ks_matrix 119 8.3 0.000 0.000 7.883 8.064 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.883 8.063 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.973 7.137 dbcsr_mm_accdrv_process 47894 16.0 3.165 5.504 4.237 6.319 qs_ot_get_p 119 10.4 0.001 0.001 5.126 5.373 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.934 5.118 sum_up_and_integrate 119 10.3 0.024 0.028 4.816 4.823 mp_sum_l 7287 12.8 2.732 4.815 2.732 4.815 integrate_v_rspace 119 11.3 0.002 0.003 4.791 4.800 init_scf_run 11 5.9 0.000 0.001 4.781 4.781 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.781 4.781 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.448 4.654 apply_single 119 13.6 0.000 0.000 3.448 4.654 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.428 4.481 calculate_rho_elec 119 8.7 0.021 0.024 4.428 4.481 rs_pw_transfer 974 11.9 0.010 0.011 3.558 4.065 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.464 3.492 make_m2s 4572 13.5 0.052 0.054 2.949 3.277 density_rs2pw 119 9.7 0.004 0.004 2.689 3.225 make_images 4572 14.5 0.201 0.238 2.860 3.186 calculate_first_density_matrix 1 7.0 0.004 0.032 3.103 3.105 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.981 2.982 init_scf_loop 11 6.9 0.001 0.004 2.836 2.839 multiply_cannon_sync_h2d 27432 15.5 2.145 2.729 2.145 2.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.587 2.589 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.564 2.565 jit_kernel_multiply 11 16.0 1.021 2.555 1.021 2.555 cp_fm_redistribute_end 50 14.0 1.297 2.525 1.302 2.528 potential_pw2rs 119 12.3 0.006 0.006 2.492 2.518 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.408 2.498 ot_diis_step 108 11.5 0.011 0.011 2.496 2.497 cp_fm_diag_elpa_base 50 14.0 1.187 2.411 1.218 2.450 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.310 2.358 calculate_dm_sparse 119 9.5 0.000 0.001 2.156 2.240 pw_transfer 1439 11.6 0.066 0.069 2.159 2.195 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.067 2.105 acc_transpose_blocks 27432 15.5 0.110 0.114 1.242 1.929 make_images_data 4572 15.5 0.045 0.051 1.314 1.921 grid_integrate_task_list 119 12.3 1.840 1.917 1.840 1.917 prepare_preconditioner 11 7.9 0.000 0.000 1.830 1.858 make_preconditioner 11 8.9 0.000 0.001 1.830 1.858 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.714 1.770 hybrid_alltoall_any 4725 16.4 0.051 0.111 1.137 1.687 fft3d_ps 1201 14.6 0.520 0.580 1.617 1.653 mp_allgather_i34 2286 14.5 0.756 1.621 0.756 1.621 wfi_extrapolate 11 7.9 0.001 0.001 1.614 1.614 fft_wrap_pw1pw2_140 487 13.2 0.203 0.214 1.507 1.547 mp_alltoall_d11v 2130 13.8 1.387 1.515 1.387 1.515 grid_collocate_task_list 119 9.7 1.270 1.396 1.270 1.396 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.295 1.340 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.318 1.332 acc_transpose_blocks_kernels 27432 16.5 0.182 0.269 0.694 1.294 mp_sum_d 4135 12.0 0.709 1.143 0.709 1.143 rs_pw_transfer_RS2PW_140 130 11.5 0.169 0.190 0.596 1.125 rs_pw_transfer_PW2RS_50 119 14.3 0.583 0.602 0.970 1.118 make_images_sizes 4572 15.5 0.005 0.005 0.802 1.112 mp_alltoall_i44 4572 16.5 0.797 1.107 0.797 1.107 mp_waitany 5720 13.7 0.537 1.065 0.537 1.065 jit_kernel_transpose 5 15.5 0.512 1.025 0.512 1.025 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.994 0.994 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.969 0.985 mp_alltoall_z22v 1201 16.6 0.824 0.907 0.824 0.907 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=44.343000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.454545, yerr=1.157084 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 521.035776E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607886. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.121 0.456 37.795 37.796 qs_mol_dyn_low 1 2.0 0.004 0.037 36.645 36.658 qs_forces 11 3.9 0.006 0.034 36.548 36.550 qs_energies 11 4.9 0.003 0.015 34.846 34.850 scf_env_do_scf 11 5.9 0.001 0.002 28.958 28.958 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.950 25.950 dbcsr_multiply_generic 2286 12.5 0.097 0.106 17.921 18.014 velocity_verlet 10 3.0 0.006 0.010 17.083 17.086 qs_scf_new_mos 108 7.5 0.001 0.001 16.501 16.523 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.500 16.522 ot_scf_mini 108 9.5 0.003 0.004 15.736 15.747 multiply_cannon 2286 13.5 0.196 0.202 13.860 14.600 multiply_cannon_loop 2286 14.5 0.636 0.665 12.963 13.747 ot_mini 108 10.5 0.001 0.001 9.377 9.397 qs_ot_get_derivative 108 11.5 0.001 0.001 7.828 7.838 multiply_cannon_multrec 18288 15.5 1.957 2.904 7.333 7.642 rebuild_ks_matrix 119 8.3 0.000 0.000 7.075 7.091 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.075 7.091 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.256 6.272 dbcsr_mm_accdrv_process 38222 16.0 4.131 6.086 5.291 6.143 mp_waitall_1 158411 16.6 3.402 4.630 3.402 4.630 sum_up_and_integrate 119 10.3 0.032 0.032 4.601 4.605 integrate_v_rspace 119 11.3 0.003 0.003 4.569 4.577 init_scf_run 11 5.9 0.000 0.001 4.475 4.475 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.475 4.475 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.224 4.233 calculate_rho_elec 119 8.7 0.031 0.031 4.224 4.232 qs_ot_get_p 119 10.4 0.001 0.001 4.078 4.096 rs_pw_transfer 974 11.9 0.009 0.010 3.381 3.619 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.936 3.591 calculate_first_density_matrix 1 7.0 0.003 0.016 3.052 3.053 init_scf_loop 11 6.9 0.007 0.031 2.985 2.986 density_rs2pw 119 9.7 0.004 0.004 2.674 2.927 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.807 2.812 make_m2s 4572 13.5 0.044 0.045 2.389 2.556 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.507 2.507 make_images 4572 14.5 0.192 0.203 2.305 2.470 jit_kernel_multiply 10 15.8 1.109 2.458 1.109 2.458 multiply_cannon_metrocomm3 18288 15.5 0.044 0.045 1.434 2.452 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.109 2.446 apply_single 119 13.6 0.000 0.000 2.109 2.446 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.350 2.354 potential_pw2rs 119 12.3 0.007 0.008 2.314 2.331 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.178 2.188 cp_fm_diag_elpa_base 50 14.0 2.148 2.164 2.175 2.185 pw_transfer 1439 11.6 0.066 0.070 2.153 2.167 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.109 2.121 prepare_preconditioner 11 7.9 0.000 0.000 2.088 2.090 make_preconditioner 11 8.9 0.000 0.001 2.088 2.090 mp_sum_l 7287 12.8 1.605 2.084 1.605 2.084 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.060 2.074 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.923 2.016 calculate_dm_sparse 119 9.5 0.000 0.001 1.981 2.003 grid_integrate_task_list 119 12.3 1.801 1.895 1.801 1.895 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.764 1.770 multiply_cannon_sync_h2d 18288 15.5 1.394 1.595 1.394 1.595 fft3d_ps 1201 14.6 0.528 0.544 1.548 1.571 ot_diis_step 108 11.5 0.011 0.011 1.522 1.522 fft_wrap_pw1pw2_140 487 13.2 0.253 0.261 1.505 1.519 grid_collocate_task_list 119 9.7 1.242 1.385 1.242 1.385 wfi_extrapolate 11 7.9 0.001 0.001 1.365 1.365 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.327 1.332 acc_transpose_blocks 18288 15.5 0.076 0.077 1.276 1.292 make_images_data 4572 15.5 0.045 0.049 1.043 1.286 hybrid_alltoall_any 4725 16.4 0.055 0.115 0.886 1.081 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.026 1.028 mp_alltoall_d11v 2130 13.8 0.865 1.006 0.865 1.006 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.946 0.971 cp_fm_cholesky_invert 11 10.9 0.912 0.916 0.912 0.916 mp_waitany 9880 13.7 0.631 0.905 0.631 0.905 mp_alltoall_z22v 1201 16.6 0.844 0.905 0.844 0.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.896 rs_pw_transfer_RS2PW_140 130 11.5 0.121 0.125 0.615 0.868 make_images_sizes 4572 15.5 0.005 0.005 0.608 0.859 mp_alltoall_i44 4572 16.5 0.603 0.854 0.603 0.854 acc_transpose_blocks_kernels 18288 16.5 0.210 0.217 0.825 0.835 mp_sendrecv_dv 11067 12.7 0.791 0.806 0.791 0.806 dbcsr_complete_redistribute 329 12.2 0.105 0.169 0.656 0.787 rs_pw_transfer_PW2RS_50 119 14.3 0.399 0.410 0.720 0.768 mp_allgather_i34 2286 14.5 0.356 0.760 0.356 0.760 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=37.796000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.636364, yerr=1.871933 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 562.028544E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.027 38.893 38.894 qs_mol_dyn_low 1 2.0 0.003 0.004 38.645 38.653 qs_forces 11 3.9 0.002 0.003 38.587 38.587 qs_energies 11 4.9 0.001 0.001 36.776 36.782 scf_env_do_scf 11 5.9 0.000 0.001 30.974 30.976 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 27.176 27.177 dbcsr_multiply_generic 2286 12.5 0.100 0.104 19.438 19.584 velocity_verlet 10 3.0 0.002 0.002 19.374 19.376 qs_scf_new_mos 108 7.5 0.001 0.001 17.688 17.747 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.687 17.746 ot_scf_mini 108 9.5 0.002 0.003 16.723 16.774 multiply_cannon 2286 13.5 0.222 0.255 15.263 15.756 multiply_cannon_loop 2286 14.5 0.942 0.975 14.234 14.599 ot_mini 108 10.5 0.001 0.001 9.921 9.985 multiply_cannon_multrec 27432 15.5 2.475 3.203 9.075 9.538 qs_ot_get_derivative 108 11.5 0.001 0.001 8.057 8.110 dbcsr_mm_accdrv_process 47916 15.9 5.675 7.727 6.510 7.781 rebuild_ks_matrix 119 8.3 0.000 0.000 7.220 7.287 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.220 7.286 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.418 6.476 init_scf_run 11 5.9 0.000 0.001 4.421 4.421 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.421 4.421 sum_up_and_integrate 119 10.3 0.036 0.039 4.349 4.358 integrate_v_rspace 119 11.3 0.003 0.003 4.313 4.322 qs_ot_get_p 119 10.4 0.001 0.001 4.081 4.156 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.109 4.137 calculate_rho_elec 119 8.7 0.040 0.046 4.108 4.137 init_scf_loop 11 6.9 0.000 0.000 3.776 3.776 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.920 3.316 mp_waitall_1 137007 16.6 2.497 3.048 2.497 3.048 rs_pw_transfer 974 11.9 0.009 0.010 2.813 3.011 prepare_preconditioner 11 7.9 0.000 0.000 2.863 2.872 make_preconditioner 11 8.9 0.000 0.000 2.863 2.872 calculate_first_density_matrix 1 7.0 0.000 0.000 2.868 2.870 make_m2s 4572 13.5 0.054 0.055 2.667 2.802 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.460 2.792 make_images 4572 14.5 0.270 0.333 2.561 2.696 density_rs2pw 119 9.7 0.004 0.004 2.466 2.659 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.594 2.603 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.317 2.318 pw_transfer 1439 11.6 0.066 0.070 2.235 2.276 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.252 2.252 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.070 2.204 apply_single 119 13.6 0.000 0.000 2.070 2.203 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.143 2.187 calculate_dm_sparse 119 9.5 0.000 0.000 2.065 2.140 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.118 2.136 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.097 2.124 potential_pw2rs 119 12.3 0.008 0.009 2.053 2.065 jit_kernel_multiply 9 16.1 0.775 1.971 0.775 1.971 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.923 1.932 grid_integrate_task_list 119 12.3 1.833 1.930 1.833 1.930 cp_fm_diag_elpa_base 50 14.0 1.879 1.896 1.920 1.929 ot_diis_step 108 11.5 0.012 0.012 1.821 1.822 fft_wrap_pw1pw2_140 487 13.2 0.288 0.302 1.696 1.742 acc_transpose_blocks 27432 15.5 0.111 0.113 1.640 1.735 fft3d_ps 1201 14.6 0.558 0.605 1.581 1.610 wfi_extrapolate 11 7.9 0.001 0.001 1.485 1.485 mp_sum_l 7287 12.8 1.078 1.397 1.078 1.397 grid_collocate_task_list 119 9.7 1.248 1.382 1.248 1.382 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.804 1.285 make_images_data 4572 15.5 0.044 0.048 1.073 1.273 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.235 1.246 cp_fm_upper_to_full 72 14.2 0.830 1.184 0.830 1.184 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.157 1.173 hybrid_alltoall_any 4725 16.4 0.062 0.151 0.883 1.157 dbcsr_complete_redistribute 329 12.2 0.120 0.145 0.877 1.154 acc_transpose_blocks_kernels 27432 16.5 0.269 0.277 1.000 1.102 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.091 1.092 multiply_cannon_sync_h2d 27432 15.5 0.991 1.042 0.991 1.042 mp_alltoall_d11v 2130 13.8 0.889 1.016 0.889 1.016 cp_fm_cholesky_invert 11 10.9 0.967 0.971 0.967 0.971 mp_alltoall_z22v 1201 16.6 0.907 0.944 0.907 0.944 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.899 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.628 0.899 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.874 jit_kernel_transpose 5 15.6 0.732 0.839 0.732 0.839 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=38.894000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=533.363636, yerr=4.333227 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 608.817152E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 31.542 31.542 qs_mol_dyn_low 1 2.0 0.003 0.003 31.321 31.329 qs_forces 11 3.9 0.002 0.002 31.263 31.264 qs_energies 11 4.9 0.001 0.001 29.483 29.486 scf_env_do_scf 11 5.9 0.000 0.001 24.121 24.121 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 21.282 21.283 velocity_verlet 10 3.0 0.002 0.002 15.883 15.886 dbcsr_multiply_generic 2286 12.5 0.100 0.105 13.595 13.691 qs_scf_new_mos 108 7.5 0.001 0.001 12.440 12.473 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.439 12.472 ot_scf_mini 108 9.5 0.002 0.002 11.705 11.731 multiply_cannon 2286 13.5 0.231 0.240 10.419 11.146 multiply_cannon_loop 2286 14.5 0.330 0.341 9.429 9.641 rebuild_ks_matrix 119 8.3 0.000 0.000 6.472 6.497 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.472 6.497 ot_mini 108 10.5 0.001 0.001 6.443 6.474 multiply_cannon_multrec 9144 15.5 1.604 1.858 6.086 6.358 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.778 5.801 qs_ot_get_derivative 108 11.5 0.001 0.001 5.072 5.098 dbcsr_mm_accdrv_process 12550 15.8 3.161 4.093 4.381 4.457 sum_up_and_integrate 119 10.3 0.037 0.042 4.048 4.053 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.019 4.026 calculate_rho_elec 119 8.7 0.060 0.061 4.018 4.025 integrate_v_rspace 119 11.3 0.003 0.003 4.011 4.016 init_scf_run 11 5.9 0.000 0.001 3.849 3.849 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.849 3.849 qs_ot_get_p 119 10.4 0.001 0.001 3.353 3.394 init_scf_loop 11 6.9 0.000 0.000 2.815 2.817 mp_waitall_1 115863 16.7 2.242 2.803 2.242 2.803 calculate_first_density_matrix 1 7.0 0.000 0.000 2.543 2.544 density_rs2pw 119 9.7 0.004 0.004 2.290 2.439 rs_pw_transfer 974 11.9 0.008 0.009 2.272 2.420 make_m2s 4572 13.5 0.034 0.035 2.198 2.383 make_images 4572 14.5 0.268 0.302 2.109 2.294 pw_transfer 1439 11.6 0.066 0.069 2.270 2.288 qs_ot_p2m_diag 50 11.0 0.022 0.023 2.244 2.248 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.176 2.194 jit_kernel_multiply 10 15.6 1.183 2.142 1.183 2.142 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.013 2.014 prepare_preconditioner 11 7.9 0.000 0.000 1.995 2.001 make_preconditioner 11 8.9 0.000 0.000 1.995 2.001 grid_integrate_task_list 119 12.3 1.862 1.940 1.862 1.940 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.911 1.913 calculate_dm_sparse 119 9.5 0.000 0.000 1.885 1.906 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.865 1.900 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.780 1.795 fft_wrap_pw1pw2_140 487 13.2 0.366 0.377 1.729 1.748 potential_pw2rs 119 12.3 0.010 0.011 1.728 1.737 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.721 1.731 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.657 1.664 cp_fm_diag_elpa_base 50 14.0 1.625 1.645 1.655 1.662 fft3d_ps 1201 14.6 0.563 0.575 1.508 1.528 grid_collocate_task_list 119 9.7 1.305 1.428 1.305 1.428 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.362 1.374 ot_diis_step 108 11.5 0.012 0.013 1.345 1.346 hybrid_alltoall_any 4725 16.4 0.063 0.177 0.964 1.325 make_images_data 4572 15.5 0.039 0.042 1.015 1.319 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.253 1.261 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.250 1.250 wfi_extrapolate 11 7.9 0.001 0.001 1.243 1.243 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.210 1.239 apply_single 119 13.6 0.000 0.000 1.210 1.238 mp_alltoall_d11v 2130 13.8 0.926 1.059 0.926 1.059 acc_transpose_blocks 9144 15.5 0.038 0.039 1.037 1.051 cp_fm_cholesky_invert 11 10.9 1.025 1.028 1.025 1.028 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.872 0.926 mp_alltoall_z22v 1201 16.6 0.819 0.858 0.819 0.858 mp_allgather_i34 2286 14.5 0.323 0.836 0.323 0.836 multiply_cannon_metrocomm1 9144 15.5 0.022 0.023 0.608 0.825 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.819 0.821 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.796 0.805 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.792 0.802 multiply_cannon_sync_h2d 9144 15.5 0.710 0.780 0.710 0.780 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.694 0.749 make_images_sizes 4572 15.5 0.005 0.005 0.498 0.725 mp_alltoall_i44 4572 16.5 0.494 0.720 0.494 0.720 mp_sum_l 7287 12.8 0.539 0.705 0.539 0.705 jit_kernel_transpose 5 15.6 0.675 0.685 0.675 0.685 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.372 0.676 dbcsr_complete_redistribute 329 12.2 0.170 0.183 0.626 0.656 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=31.542000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=574.909091, yerr=7.513074 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 750.735360E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.052 47.507 47.508 qs_mol_dyn_low 1 2.0 0.003 0.005 46.451 46.460 qs_forces 11 3.9 0.020 0.022 46.391 46.392 qs_energies 11 4.9 0.001 0.001 44.308 44.313 scf_env_do_scf 11 5.9 0.001 0.001 36.046 36.047 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 27.860 27.861 velocity_verlet 10 3.0 0.002 0.002 24.773 24.778 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.818 21.284 multiply_cannon 2286 13.5 0.300 0.308 16.150 17.571 qs_scf_new_mos 108 7.5 0.001 0.001 17.410 17.500 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.409 17.499 multiply_cannon_loop 2286 14.5 0.344 0.350 14.790 16.392 ot_scf_mini 108 9.5 0.002 0.002 16.264 16.367 multiply_cannon_multrec 9144 15.5 3.509 5.197 9.439 9.979 ot_mini 108 10.5 0.001 0.001 9.747 9.887 init_scf_loop 11 6.9 0.000 0.000 8.158 8.159 rebuild_ks_matrix 119 8.3 0.000 0.000 7.729 7.908 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.728 7.907 qs_ot_get_derivative 108 11.5 0.001 0.001 7.651 7.758 dbcsr_mm_accdrv_process 12550 15.8 4.775 6.505 5.806 7.611 prepare_preconditioner 11 7.9 0.000 0.000 7.145 7.160 make_preconditioner 11 8.9 0.000 0.000 7.145 7.160 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.955 7.119 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.725 7.028 init_scf_run 11 5.9 0.000 0.001 6.182 6.182 scf_env_initial_rho_setup 11 6.9 0.000 0.000 6.182 6.182 mp_waitall_1 94719 16.7 3.963 6.027 3.963 6.027 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.635 4.770 calculate_rho_elec 119 8.7 0.118 0.121 4.634 4.769 calculate_first_density_matrix 1 7.0 0.000 0.000 4.566 4.586 cp_fm_upper_to_full 72 14.2 3.172 4.527 3.172 4.527 sum_up_and_integrate 119 10.3 0.064 0.066 4.371 4.377 integrate_v_rspace 119 11.3 0.003 0.003 4.306 4.313 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 2.361 4.015 qs_ot_get_p 119 10.4 0.001 0.001 3.620 3.775 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.463 3.465 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.763 3.342 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.004 3.156 calculate_dm_sparse 119 9.5 0.000 0.000 3.056 3.109 pw_transfer 1439 11.6 0.069 0.069 2.972 2.982 dbcsr_complete_redistribute 329 12.2 0.358 0.367 2.103 2.908 make_m2s 4572 13.5 0.037 0.038 2.705 2.891 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.874 2.883 make_images 4572 14.5 0.353 0.385 2.585 2.772 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.376 2.683 apply_single 119 13.6 0.000 0.000 2.376 2.682 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.763 2.560 density_rs2pw 119 9.7 0.004 0.004 2.479 2.501 mp_sum_l 7287 12.8 1.434 2.392 1.434 2.392 fft_wrap_pw1pw2_140 487 13.2 0.616 0.622 2.366 2.379 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.223 2.225 mp_alltoall_i22 627 13.8 1.507 2.213 1.507 2.213 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.414 2.209 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.074 2.103 grid_integrate_task_list 119 12.3 2.069 2.085 2.069 2.085 ot_diis_step 108 11.5 0.014 0.014 2.048 2.049 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.926 1.927 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.864 1.914 fft3d_ps 1201 14.6 0.596 0.606 1.891 1.898 rs_pw_transfer 974 11.9 0.009 0.009 1.862 1.885 jit_kernel_multiply 6 15.7 1.004 1.799 1.004 1.799 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.788 1.796 potential_pw2rs 119 12.3 0.014 0.015 1.709 1.713 cp_fm_cholesky_invert 11 10.9 1.694 1.697 1.694 1.697 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.602 1.602 cp_fm_diag_elpa_base 50 14.0 1.449 1.502 1.600 1.600 copy_dbcsr_to_fm 153 11.3 0.002 0.002 0.835 1.564 wfi_extrapolate 11 7.9 0.001 0.001 1.527 1.527 grid_collocate_task_list 119 9.7 1.515 1.523 1.515 1.523 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.253 1.508 make_images_data 4572 15.5 0.043 0.045 1.217 1.490 mp_alltoall_d11v 2130 13.8 1.262 1.391 1.262 1.391 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.274 1.311 acc_transpose_blocks 9144 15.5 0.038 0.039 1.211 1.226 mp_alltoall_z22v 1201 16.6 1.159 1.185 1.159 1.185 dbcsr_desymmetrize_deep 153 12.3 0.044 0.047 0.438 1.171 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.116 1.133 multiply_cannon_sync_h2d 9144 15.5 1.044 1.053 1.044 1.053 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.986 1.035 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.986 1.001 acc_transpose_blocks_kernels 9144 16.5 0.117 0.119 0.955 0.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=47.508000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=703.636364, yerr=14.398691 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.653504E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1702427. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.028 88.821 88.821 qs_mol_dyn_low 1 2.0 0.003 0.003 88.481 88.490 qs_forces 11 3.9 0.003 0.003 88.095 88.096 qs_energies 11 4.9 0.002 0.002 84.708 84.733 scf_env_do_scf 11 5.9 0.001 0.001 74.626 74.628 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 68.391 68.391 dbcsr_multiply_generic 2055 12.4 0.107 0.110 52.438 52.797 qs_scf_new_mos 99 7.5 0.000 0.001 49.560 49.648 qs_scf_loop_do_ot 99 8.5 0.001 0.001 49.559 49.648 ot_scf_mini 99 9.5 0.002 0.002 47.111 47.192 velocity_verlet 10 3.0 0.030 0.219 46.217 46.246 multiply_cannon 2055 13.4 0.182 0.188 43.055 44.080 multiply_cannon_loop 2055 14.4 1.524 1.559 41.878 42.744 ot_mini 99 10.5 0.001 0.001 27.402 27.490 qs_ot_get_derivative 99 11.5 0.001 0.001 20.585 20.685 multiply_cannon_multrec 49320 15.4 12.393 13.013 17.392 18.279 rebuild_ks_matrix 110 8.3 0.000 0.000 15.778 15.900 qs_ks_build_kohn_sham_matrix 110 9.3 0.055 0.363 15.778 15.899 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.485 13.596 mp_waitall_1 220248 16.4 11.772 12.909 11.772 12.909 qs_ot_get_p 110 10.4 0.001 0.002 10.949 11.023 multiply_cannon_sync_h2d 49320 15.4 10.242 10.769 10.242 10.769 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 8.042 8.552 sum_up_and_integrate 110 10.3 0.036 0.043 7.894 8.159 integrate_v_rspace 110 11.3 0.003 0.003 7.857 8.129 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.230 7.879 apply_single 110 13.6 0.000 0.000 7.230 7.878 multiply_cannon_metrocomm3 49320 15.4 0.079 0.082 6.600 7.690 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.558 7.680 calculate_rho_elec 110 8.6 0.021 0.026 7.557 7.679 init_scf_run 11 5.9 0.000 0.001 7.557 7.557 scf_env_initial_rho_setup 11 6.9 0.000 0.001 7.556 7.557 qs_ot_p2m_diag 48 11.0 0.012 0.018 7.508 7.544 ot_diis_step 99 11.5 0.005 0.006 6.595 6.596 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.568 6.568 init_scf_loop 11 6.9 0.000 0.000 6.192 6.193 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.883 5.911 cp_fm_diag_elpa_base 48 14.0 5.864 5.893 5.881 5.910 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.621 5.670 rs_pw_transfer 902 11.9 0.012 0.013 4.862 5.352 mp_sum_l 6594 12.7 4.510 5.231 4.510 5.231 dbcsr_mm_accdrv_process 87628 16.1 1.927 2.005 4.874 5.196 density_rs2pw 110 9.6 0.004 0.005 4.244 4.865 make_m2s 4110 13.4 0.060 0.064 4.412 4.535 make_images 4110 14.4 0.178 0.191 4.316 4.441 wfi_extrapolate 11 7.9 0.001 0.001 4.330 4.330 pw_transfer 1331 11.6 0.056 0.067 3.868 4.197 calculate_dm_sparse 110 9.5 0.001 0.001 4.034 4.128 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.779 4.109 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.740 4.106 prepare_preconditioner 11 7.9 0.000 0.000 3.933 3.949 make_preconditioner 11 8.9 0.000 0.000 3.933 3.949 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.793 3.798 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.685 3.741 potential_pw2rs 110 12.3 0.006 0.007 3.356 3.673 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.438 3.502 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.406 3.464 fft_wrap_pw1pw2_140 451 13.1 0.452 0.501 3.080 3.420 grid_integrate_task_list 110 12.3 3.232 3.401 3.232 3.401 fft3d_ps 1111 14.6 0.783 0.872 2.971 3.272 calculate_first_density_matrix 1 7.0 0.000 0.001 3.096 3.099 mp_alltoall_d11v 2046 13.8 2.337 2.814 2.337 2.814 jit_kernel_multiply 13 15.9 2.673 2.772 2.673 2.772 mp_waitany 14300 13.8 2.076 2.750 2.076 2.750 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.426 2.450 grid_collocate_task_list 110 9.6 2.155 2.448 2.155 2.448 acc_transpose_blocks 49320 15.4 0.212 0.219 2.247 2.403 cp_fm_cholesky_invert 11 10.9 2.223 2.227 2.223 2.227 mp_alltoall_z22v 1111 16.6 1.814 2.214 1.814 2.214 make_images_data 4110 15.4 0.043 0.046 1.918 2.124 hybrid_alltoall_any 4261 16.3 0.083 0.481 1.651 1.979 mp_sum_d 3889 11.9 1.410 1.960 1.410 1.960 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.894 1.929 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=88.821000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.181818, yerr=2.979267 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.923456E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 3608167. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.031 73.747 73.748 qs_mol_dyn_low 1 2.0 0.003 0.004 73.353 73.503 qs_forces 11 3.9 0.003 0.003 73.271 73.272 qs_energies 11 4.9 0.001 0.002 69.844 69.847 scf_env_do_scf 11 5.9 0.000 0.001 60.728 60.731 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 52.367 52.368 dbcsr_multiply_generic 2055 12.4 0.117 0.120 39.257 39.414 velocity_verlet 10 3.0 0.001 0.002 38.106 38.108 qs_scf_new_mos 99 7.5 0.001 0.001 35.105 35.285 qs_scf_loop_do_ot 99 8.5 0.001 0.001 35.104 35.285 ot_scf_mini 99 9.5 0.003 0.003 33.410 33.601 multiply_cannon 2055 13.4 0.221 0.244 32.076 33.248 multiply_cannon_loop 2055 14.4 0.920 0.941 30.653 31.536 ot_mini 99 10.5 0.001 0.001 19.373 19.552 multiply_cannon_multrec 24660 15.4 7.676 9.360 14.478 16.329 rebuild_ks_matrix 110 8.3 0.000 0.000 14.334 14.488 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 14.334 14.487 qs_ot_get_derivative 99 11.5 0.001 0.001 13.488 13.675 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.618 12.759 mp_waitall_1 176588 16.5 8.356 10.930 8.356 10.930 init_scf_loop 11 6.9 0.000 0.000 8.319 8.320 multiply_cannon_sync_h2d 24660 15.4 7.042 8.296 7.042 8.296 multiply_cannon_metrocomm3 24660 15.4 0.067 0.070 5.406 8.244 qs_ot_get_p 110 10.4 0.001 0.001 7.295 7.525 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.620 7.242 apply_single 110 13.6 0.000 0.001 6.620 7.242 sum_up_and_integrate 110 10.3 0.052 0.058 7.087 7.098 integrate_v_rspace 110 11.3 0.002 0.003 7.035 7.048 dbcsr_mm_accdrv_process 52282 16.1 4.735 5.910 6.640 6.952 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.748 6.760 calculate_rho_elec 110 8.6 0.040 0.048 6.747 6.760 init_scf_run 11 5.9 0.000 0.001 6.669 6.669 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.668 6.669 prepare_preconditioner 11 7.9 0.000 0.000 6.227 6.245 make_preconditioner 11 8.9 0.000 0.000 6.227 6.245 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.793 5.961 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.087 5.898 ot_diis_step 99 11.5 0.010 0.010 5.834 5.834 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.243 5.263 make_m2s 4110 13.4 0.057 0.060 4.678 5.190 make_images 4110 14.4 0.401 0.448 4.569 5.076 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.784 4.784 density_rs2pw 110 9.6 0.004 0.005 3.798 4.334 rs_pw_transfer 902 11.9 0.013 0.014 3.606 4.221 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.105 4.119 pw_transfer 1331 11.6 0.067 0.075 3.948 4.116 cp_fm_diag_elpa_base 48 14.0 4.049 4.071 4.102 4.115 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 3.841 4.010 wfi_extrapolate 11 7.9 0.001 0.001 3.732 3.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.586 3.588 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.350 3.440 grid_integrate_task_list 110 12.3 3.151 3.397 3.151 3.397 fft_wrap_pw1pw2_140 451 13.1 0.522 0.538 3.152 3.321 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.201 3.249 make_images_data 4110 15.4 0.046 0.051 2.552 3.120 hybrid_alltoall_any 4261 16.3 0.103 0.446 2.213 3.062 calculate_dm_sparse 110 9.5 0.001 0.001 3.018 3.053 cp_fm_cholesky_invert 11 10.9 3.018 3.025 3.018 3.025 fft3d_ps 1111 14.6 1.105 1.330 2.813 2.996 potential_pw2rs 110 12.3 0.008 0.009 2.820 2.851 calculate_first_density_matrix 1 7.0 0.000 0.000 2.822 2.825 jit_kernel_multiply 11 16.2 1.557 2.663 1.557 2.663 grid_collocate_task_list 110 9.6 2.096 2.544 2.096 2.544 mp_sum_l 6594 12.7 1.874 2.484 1.874 2.484 mp_alltoall_d11v 2046 13.8 1.947 2.358 1.947 2.358 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.015 2.038 mp_waitany 10164 13.8 1.291 1.929 1.291 1.929 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.903 1.905 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.874 1.887 mp_allgather_i34 2055 14.4 0.752 1.849 0.752 1.849 multiply_cannon_metrocomm4 22605 15.4 0.078 0.081 0.784 1.706 acc_transpose_blocks 24660 15.4 0.112 0.115 1.548 1.624 mp_irecv_dv 57340 16.2 0.657 1.582 0.657 1.582 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.567 1.581 cp_fm_cholesky_decompose 22 10.9 1.560 1.565 1.560 1.565 rs_pw_transfer_RS2PW_140 121 11.5 0.208 0.219 0.993 1.564 mp_alltoall_z22v 1111 16.6 1.423 1.540 1.423 1.540 dbcsr_complete_redistribute 325 12.2 0.237 0.290 1.250 1.526 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.386 1.491 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=73.748000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=555.000000, yerr=7.298817 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 662.528000E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.398 65.099 65.100 qs_mol_dyn_low 1 2.0 0.003 0.004 64.354 64.365 qs_forces 11 3.9 0.003 0.003 64.286 64.288 qs_energies 11 4.9 0.015 0.015 60.998 61.002 scf_env_do_scf 11 5.9 0.000 0.001 52.136 52.137 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 42.571 42.572 velocity_verlet 10 3.0 0.001 0.002 34.105 34.107 dbcsr_multiply_generic 2055 12.4 0.117 0.124 30.103 30.379 qs_scf_new_mos 99 7.5 0.001 0.001 26.756 26.837 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.756 26.836 ot_scf_mini 99 9.5 0.002 0.003 25.507 25.612 multiply_cannon 2055 13.4 0.213 0.222 23.084 24.328 multiply_cannon_loop 2055 14.4 0.616 0.630 21.835 22.828 ot_mini 99 10.5 0.001 0.001 14.510 14.612 rebuild_ks_matrix 110 8.3 0.000 0.000 12.835 12.971 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.835 12.971 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.285 11.410 multiply_cannon_multrec 16440 15.4 4.019 4.995 10.399 11.318 qs_ot_get_derivative 99 11.5 0.001 0.001 10.104 10.207 mp_waitall_1 139946 16.5 7.475 10.169 7.475 10.169 init_scf_loop 11 6.9 0.000 0.000 9.526 9.527 prepare_preconditioner 11 7.9 0.000 0.000 7.394 7.410 make_preconditioner 11 8.9 0.000 0.001 7.394 7.410 multiply_cannon_metrocomm3 16440 15.4 0.042 0.046 4.331 7.134 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.667 7.048 sum_up_and_integrate 110 10.3 0.061 0.062 6.924 6.937 integrate_v_rspace 110 11.3 0.003 0.003 6.863 6.877 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.442 6.452 calculate_rho_elec 110 8.6 0.059 0.060 6.442 6.452 dbcsr_mm_accdrv_process 34862 16.1 4.992 5.890 6.233 6.423 init_scf_run 11 5.9 0.000 0.001 6.388 6.388 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.388 6.388 qs_ot_get_p 110 10.4 0.001 0.001 6.108 6.257 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.966 5.362 apply_single 110 13.6 0.000 0.000 4.966 5.361 make_m2s 4110 13.4 0.050 0.051 4.592 4.997 make_images 4110 14.4 0.395 0.514 4.477 4.882 density_rs2pw 110 9.6 0.004 0.005 3.461 4.626 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.385 4.390 ot_diis_step 99 11.5 0.011 0.011 4.377 4.378 multiply_cannon_sync_h2d 16440 15.4 3.725 4.330 3.725 4.330 rs_pw_transfer 902 11.9 0.011 0.012 3.179 4.312 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.469 4.117 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.023 4.024 pw_transfer 1331 11.6 0.066 0.073 3.877 3.894 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.769 3.789 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.555 3.556 grid_integrate_task_list 110 12.3 3.160 3.424 3.160 3.424 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.392 3.403 cp_fm_diag_elpa_base 48 14.0 3.314 3.355 3.390 3.401 make_images_data 4110 15.4 0.043 0.047 2.602 3.227 fft_wrap_pw1pw2_140 451 13.1 0.640 0.648 3.166 3.185 calculate_first_density_matrix 1 7.0 0.000 0.000 3.174 3.176 hybrid_alltoall_any 4261 16.3 0.106 0.378 2.318 3.124 wfi_extrapolate 11 7.9 0.001 0.001 3.110 3.110 cp_fm_cholesky_invert 11 10.9 3.056 3.062 3.056 3.062 mp_sum_l 6594 12.7 2.125 2.869 2.125 2.869 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.750 2.789 potential_pw2rs 110 12.3 0.011 0.011 2.609 2.628 fft3d_ps 1111 14.6 1.092 1.104 2.587 2.601 calculate_dm_sparse 110 9.5 0.001 0.001 2.551 2.585 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.462 2.525 mp_alltoall_d11v 2046 13.8 1.951 2.512 1.951 2.512 grid_collocate_task_list 110 9.6 2.130 2.499 2.130 2.499 dbcsr_complete_redistribute 325 12.2 0.788 0.830 2.020 2.494 mp_waitany 17072 13.8 1.250 2.475 1.250 2.475 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.851 2.226 mp_irecv_dv 48980 15.7 0.782 2.106 0.782 2.106 rs_pw_transfer_RS2PW_140 121 11.5 0.179 0.183 0.934 2.089 jit_kernel_multiply 8 16.5 0.856 2.069 0.856 2.069 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 1.998 2.000 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.522 1.980 cp_fm_upper_to_full 70 14.2 1.403 1.815 1.403 1.815 cp_fm_cholesky_decompose 22 10.9 1.764 1.786 1.764 1.786 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.746 1.759 mp_allgather_i34 2055 14.4 0.569 1.658 0.569 1.658 acc_transpose_blocks 16440 15.4 0.073 0.076 1.307 1.540 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.488 1.500 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.002 1.365 1.482 rs_gather_matrices 110 12.3 0.232 0.260 1.015 1.481 mp_alltoall_z22v 1111 16.6 1.284 1.303 1.284 1.303 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=65.100000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=627.272727, yerr=8.853257 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 736.563200E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.031 69.666 69.668 qs_mol_dyn_low 1 2.0 0.003 0.004 69.330 69.339 qs_forces 11 3.9 0.003 0.004 69.134 69.135 qs_energies 11 4.9 0.016 0.064 65.280 65.283 scf_env_do_scf 11 5.9 0.011 0.055 56.235 56.250 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 43.757 43.757 velocity_verlet 10 3.0 0.001 0.002 38.660 38.667 dbcsr_multiply_generic 2055 12.4 0.113 0.118 30.381 30.576 qs_scf_new_mos 99 7.5 0.001 0.001 28.012 28.157 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.011 28.157 ot_scf_mini 99 9.5 0.003 0.004 26.365 26.478 multiply_cannon 2055 13.4 0.244 0.258 22.877 23.982 multiply_cannon_loop 2055 14.4 0.888 0.913 21.477 22.019 ot_mini 99 10.5 0.001 0.001 14.683 14.821 multiply_cannon_multrec 24660 15.4 4.247 6.910 12.834 14.116 rebuild_ks_matrix 110 8.3 0.000 0.000 13.025 13.129 qs_ks_build_kohn_sham_matrix 110 9.3 0.060 0.375 13.025 13.129 init_scf_loop 11 6.9 0.002 0.006 12.425 12.448 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.179 11.270 prepare_preconditioner 11 7.9 0.000 0.000 10.553 10.570 make_preconditioner 11 8.9 0.001 0.004 10.553 10.570 qs_ot_get_derivative 99 11.5 0.001 0.001 10.251 10.372 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.753 10.238 dbcsr_mm_accdrv_process 52304 16.0 7.258 8.633 8.431 9.397 sum_up_and_integrate 110 10.3 0.068 0.071 6.879 6.895 integrate_v_rspace 110 11.3 0.003 0.004 6.811 6.825 qs_ot_get_p 110 10.4 0.001 0.001 6.588 6.728 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.547 6.566 calculate_rho_elec 110 8.6 0.078 0.082 6.546 6.565 mp_waitall_1 121746 16.5 4.721 6.498 4.721 6.498 make_m2s 4110 13.4 0.059 0.060 5.862 6.166 init_scf_run 11 5.9 0.000 0.001 6.164 6.165 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.164 6.165 make_images 4110 14.4 0.575 0.694 5.722 6.022 cp_fm_upper_to_full 70 14.2 3.357 4.890 3.357 4.890 qs_ot_p2m_diag 48 11.0 0.055 0.066 4.690 4.705 ot_diis_step 99 11.5 0.011 0.011 4.393 4.393 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.186 4.187 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.033 4.115 apply_single 110 13.6 0.000 0.000 4.033 4.115 pw_transfer 1331 11.6 0.066 0.076 3.975 4.017 dbcsr_complete_redistribute 325 12.2 0.416 0.452 2.755 3.931 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.869 3.915 density_rs2pw 110 9.6 0.004 0.004 3.381 3.905 rs_pw_transfer 902 11.9 0.010 0.011 3.186 3.763 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.531 3.541 cp_fm_diag_elpa_base 48 14.0 3.371 3.432 3.529 3.539 grid_integrate_task_list 110 12.3 3.282 3.515 3.282 3.515 calculate_dm_sparse 110 9.5 0.001 0.001 3.418 3.476 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.401 3.450 make_images_data 4110 15.4 0.045 0.049 2.971 3.385 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.173 3.328 multiply_cannon_sync_h2d 24660 15.4 3.194 3.328 3.194 3.328 fft_wrap_pw1pw2_140 451 13.1 0.668 0.692 3.238 3.289 qs_ot_get_derivative_diag 47 12.0 0.001 0.002 3.174 3.242 hybrid_alltoall_any 4261 16.3 0.121 0.460 2.468 3.227 cp_fm_cholesky_invert 11 10.9 3.138 3.147 3.138 3.147 wfi_extrapolate 11 7.9 0.001 0.001 3.134 3.134 multiply_cannon_metrocomm3 24660 15.4 0.036 0.038 1.362 3.132 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.791 2.934 calculate_first_density_matrix 1 7.0 0.001 0.002 2.895 2.918 mp_alltoall_i22 605 13.7 1.715 2.880 1.715 2.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.812 2.814 fft3d_ps 1111 14.6 1.086 1.117 2.630 2.652 potential_pw2rs 110 12.3 0.013 0.013 2.471 2.494 grid_collocate_task_list 110 9.6 2.224 2.490 2.224 2.490 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.439 2.473 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.309 2.310 mp_alltoall_d11v 2046 13.8 1.903 2.130 1.903 2.130 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.092 2.104 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 2.013 2.054 mp_waitany 13376 13.8 1.471 2.023 1.471 2.023 cp_fm_cholesky_decompose 22 10.9 1.837 1.877 1.837 1.877 acc_transpose_blocks 24660 15.4 0.104 0.107 1.781 1.802 jit_kernel_multiply 9 16.0 0.842 1.801 0.842 1.801 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.746 1.776 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.011 1.615 1.708 mp_sum_l 6594 12.7 0.965 1.599 0.965 1.599 multiply_cannon_metrocomm4 20550 15.4 0.060 0.062 0.840 1.574 mp_allgather_i34 2055 14.4 0.500 1.553 0.500 1.553 mp_irecv_dv 62702 16.1 0.739 1.492 0.739 1.492 rs_pw_transfer_RS2PW_140 121 11.5 0.172 0.178 0.869 1.413 mp_alltoall_z22v 1111 16.6 1.328 1.400 1.328 1.400 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.668000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=698.090909, yerr=9.404588 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 830.341120E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.028 59.972 59.973 qs_mol_dyn_low 1 2.0 0.003 0.003 59.680 59.696 qs_forces 11 3.9 0.003 0.003 59.593 59.593 qs_energies 11 4.9 0.001 0.001 55.813 55.821 scf_env_do_scf 11 5.9 0.000 0.001 46.663 46.663 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 37.988 37.989 velocity_verlet 10 3.0 0.006 0.007 33.130 33.135 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.830 23.971 qs_scf_new_mos 99 7.5 0.001 0.001 22.197 22.237 qs_scf_loop_do_ot 99 8.5 0.001 0.001 22.196 22.236 ot_scf_mini 99 9.5 0.002 0.002 20.914 20.930 multiply_cannon 2055 13.4 0.243 0.255 17.842 19.382 multiply_cannon_loop 2055 14.4 0.322 0.336 16.417 16.823 rebuild_ks_matrix 110 8.3 0.000 0.000 12.360 12.384 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.360 12.383 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.947 10.964 ot_mini 99 10.5 0.001 0.001 10.952 10.959 init_scf_loop 11 6.9 0.000 0.000 8.621 8.623 multiply_cannon_multrec 8220 15.4 3.251 4.546 7.614 8.589 mp_waitall_1 103326 16.6 6.578 8.364 6.578 8.364 qs_ot_get_derivative 99 11.5 0.001 0.001 7.131 7.146 prepare_preconditioner 11 7.9 0.000 0.000 6.924 6.926 make_preconditioner 11 8.9 0.000 0.000 6.924 6.926 sum_up_and_integrate 110 10.3 0.079 0.081 6.856 6.872 integrate_v_rspace 110 11.3 0.003 0.003 6.776 6.792 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.738 6.750 calculate_rho_elec 110 8.6 0.114 0.115 6.737 6.750 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.478 6.568 qs_ot_get_p 110 10.4 0.001 0.001 5.923 5.936 init_scf_run 11 5.9 0.000 0.001 5.926 5.926 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.926 5.926 dbcsr_mm_accdrv_process 17442 15.9 2.944 3.957 4.233 5.122 make_m2s 4110 13.4 0.039 0.040 4.594 4.844 make_images 4110 14.4 0.637 0.699 4.464 4.711 qs_ot_p2m_diag 48 11.0 0.081 0.084 4.402 4.409 pw_transfer 1331 11.6 0.066 0.072 4.235 4.257 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.127 4.155 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 2.903 4.120 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.063 4.063 ot_diis_step 99 11.5 0.012 0.012 3.794 3.794 density_rs2pw 110 9.6 0.004 0.004 3.389 3.762 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.699 3.723 apply_single 110 13.6 0.000 0.000 3.699 3.723 grid_integrate_task_list 110 12.3 3.352 3.535 3.352 3.535 fft_wrap_pw1pw2_140 451 13.1 0.833 0.844 3.485 3.518 cp_fm_cholesky_invert 11 10.9 3.470 3.474 3.470 3.474 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.410 3.417 cp_fm_diag_elpa_base 48 14.0 3.346 3.374 3.408 3.415 multiply_cannon_sync_h2d 8220 15.4 2.908 3.067 2.908 3.067 hybrid_alltoall_any 4261 16.3 0.201 0.870 2.492 3.043 make_images_data 4110 15.4 0.038 0.043 2.573 3.030 rs_pw_transfer 902 11.9 0.010 0.010 2.591 2.972 wfi_extrapolate 11 7.9 0.001 0.001 2.922 2.922 calculate_first_density_matrix 1 7.0 0.000 0.000 2.895 2.896 calculate_dm_sparse 110 9.5 0.001 0.001 2.690 2.725 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.702 2.702 fft3d_ps 1111 14.6 1.140 1.166 2.673 2.695 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.619 2.620 grid_collocate_task_list 110 9.6 2.319 2.574 2.319 2.574 potential_pw2rs 110 12.3 0.015 0.016 2.342 2.353 mp_alltoall_d11v 2046 13.8 1.862 2.270 1.862 2.270 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.239 2.246 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.058 2.071 cp_fm_cholesky_decompose 22 10.9 1.986 2.008 1.986 2.008 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.778 1.986 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.969 1.978 mp_allgather_i34 2055 14.4 0.610 1.749 0.610 1.749 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.743 1.749 dbcsr_complete_redistribute 325 12.2 0.554 0.578 1.551 1.653 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.523 1.646 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.628 1.639 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 1.100 1.612 jit_kernel_multiply 8 15.8 0.979 1.572 0.979 1.572 mp_waitany 9240 13.8 1.124 1.548 1.124 1.548 qs_create_task_list 11 7.9 0.000 0.001 1.225 1.332 generate_qs_task_list 11 8.9 0.375 0.441 1.225 1.331 rs_gather_matrices 110 12.3 0.324 0.370 0.985 1.318 mp_alltoall_z22v 1111 16.6 1.298 1.310 1.298 1.310 rs_pw_transfer_RS2PW_140 121 11.5 0.171 0.179 0.832 1.240 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.216 1.238 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=59.973000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=783.909091, yerr=8.596559 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 1.390391E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 94.654 94.654 qs_mol_dyn_low 1 2.0 0.003 0.003 94.280 94.290 qs_forces 11 3.9 0.004 0.006 94.210 94.211 qs_energies 11 4.9 0.002 0.002 89.886 89.890 scf_env_do_scf 11 5.9 0.000 0.001 78.454 78.454 velocity_verlet 10 3.0 0.002 0.002 59.177 59.184 scf_env_do_scf_inner_loop 99 6.5 0.002 0.006 48.462 48.464 dbcsr_multiply_generic 2055 12.4 0.119 0.123 31.110 31.239 init_scf_loop 11 6.9 0.000 0.000 29.908 29.910 qs_scf_new_mos 99 7.5 0.001 0.001 29.148 29.165 qs_scf_loop_do_ot 99 8.5 0.001 0.001 29.147 29.164 prepare_preconditioner 11 7.9 0.000 0.000 27.811 27.815 make_preconditioner 11 8.9 0.000 0.000 27.811 27.815 ot_scf_mini 99 9.5 0.002 0.002 27.321 27.337 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.878 27.266 multiply_cannon 2055 13.4 0.340 0.364 23.383 24.572 multiply_cannon_loop 2055 14.4 0.341 0.343 21.421 22.198 cp_fm_upper_to_full 70 14.2 12.861 18.501 12.861 18.501 ot_mini 99 10.5 0.001 0.001 14.933 14.961 rebuild_ks_matrix 110 8.3 0.000 0.000 14.726 14.758 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.726 14.758 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.299 13.328 dbcsr_complete_redistribute 325 12.2 1.103 1.145 7.754 11.120 multiply_cannon_multrec 8220 15.4 4.789 5.023 10.308 10.538 qs_ot_get_derivative 99 11.5 0.001 0.001 10.275 10.287 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.652 10.026 mp_waitall_1 84994 16.7 8.585 9.574 8.585 9.574 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.918 9.265 mp_alltoall_i22 605 13.7 5.551 8.902 5.551 8.902 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.355 8.392 calculate_rho_elec 110 8.6 0.224 0.224 8.354 8.391 sum_up_and_integrate 110 10.3 0.150 0.152 7.780 7.792 integrate_v_rspace 110 11.3 0.004 0.004 7.629 7.640 init_scf_run 11 5.9 0.000 0.001 7.063 7.064 scf_env_initial_rho_setup 11 6.9 0.000 0.000 7.063 7.063 qs_ot_get_p 110 10.4 0.001 0.001 6.933 6.957 make_m2s 4110 13.4 0.042 0.043 5.722 6.191 cp_fm_cholesky_invert 11 10.9 6.130 6.135 6.130 6.135 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 5.409 6.102 make_images 4110 14.4 0.879 0.931 5.533 6.000 dbcsr_mm_accdrv_process 11614 15.7 3.290 3.614 5.374 5.781 pw_transfer 1331 11.6 0.075 0.076 5.503 5.514 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.385 5.396 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.764 5.211 apply_single 110 13.6 0.000 0.000 4.764 5.210 qs_ot_p2m_diag 48 11.0 0.151 0.156 5.108 5.112 fft_wrap_pw1pw2_140 451 13.1 1.336 1.341 4.607 4.618 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.618 4.618 ot_diis_step 99 11.5 0.015 0.016 4.611 4.611 density_rs2pw 110 9.6 0.004 0.004 4.258 4.293 multiply_cannon_sync_h2d 8220 15.4 3.951 3.952 3.951 3.952 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.903 3.903 cp_fm_diag_elpa_base 48 14.0 3.340 3.546 3.900 3.900 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.847 3.849 make_images_data 4110 15.4 0.041 0.044 3.081 3.767 hybrid_alltoall_any 4261 16.3 0.256 0.552 3.095 3.762 grid_integrate_task_list 110 12.3 3.694 3.758 3.694 3.758 qs_energies_init_hamiltonians 11 5.9 0.005 0.012 3.720 3.721 wfi_extrapolate 11 7.9 0.001 0.001 3.642 3.642 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.115 3.566 fft3d_ps 1111 14.6 1.300 1.307 3.317 3.335 calculate_first_density_matrix 1 7.0 0.000 0.000 3.285 3.286 calculate_dm_sparse 110 9.5 0.001 0.001 3.221 3.246 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.186 3.194 rs_pw_transfer 902 11.9 0.010 0.011 2.794 2.821 potential_pw2rs 110 12.3 0.021 0.022 2.769 2.779 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.711 2.716 grid_collocate_task_list 110 9.6 2.682 2.693 2.682 2.693 cp_fm_cholesky_decompose 22 10.9 2.410 2.436 2.410 2.436 mp_alltoall_d11v 2046 13.8 2.301 2.425 2.301 2.425 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.218 2.277 jit_kernel_multiply 10 15.3 1.880 2.180 1.880 2.180 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.120 2.178 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.033 2.035 qs_create_task_list 11 7.9 0.001 0.002 1.903 1.952 generate_qs_task_list 11 8.9 0.731 0.784 1.903 1.950 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=94.654000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1238.363636, yerr=66.870666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 632.860672E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 1884503. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.032 220.136 220.137 qs_mol_dyn_low 1 2.0 0.003 0.004 219.609 219.622 qs_forces 11 3.9 0.005 0.005 219.434 219.435 qs_energies 11 4.9 0.001 0.002 213.690 213.709 scf_env_do_scf 11 5.9 0.001 0.001 194.568 194.572 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 172.235 172.237 velocity_verlet 10 3.0 0.001 0.002 130.101 130.103 qs_scf_new_mos 118 7.6 0.001 0.001 129.810 129.995 qs_scf_loop_do_ot 118 8.6 0.001 0.001 129.810 129.994 dbcsr_multiply_generic 2527 12.6 0.180 0.186 127.374 128.101 ot_scf_mini 118 9.6 0.003 0.003 123.131 123.338 multiply_cannon 2527 13.6 0.241 0.248 103.054 104.532 multiply_cannon_loop 2527 14.6 2.111 2.159 100.632 101.802 ot_mini 118 10.6 0.001 0.001 67.469 67.677 multiply_cannon_multrec 60648 15.6 33.404 35.117 42.241 43.561 qs_ot_get_derivative 118 11.6 0.001 0.001 42.605 42.787 rebuild_ks_matrix 129 8.3 0.001 0.001 35.011 35.334 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.016 35.010 35.333 qs_ot_get_p 129 10.4 0.001 0.001 33.468 33.701 mp_waitall_1 269254 16.5 29.743 33.497 29.743 33.497 qs_ks_update_qs_env 129 7.6 0.001 0.001 31.469 31.715 multiply_cannon_sync_h2d 60648 15.6 27.510 29.137 27.510 29.137 qs_ot_p2m_diag 83 11.4 0.079 0.092 26.438 26.502 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.515 25.281 apply_single 129 13.6 0.001 0.001 24.515 25.281 ot_diis_step 118 11.6 0.008 0.008 24.639 24.640 cp_dbcsr_syevd 83 12.4 0.005 0.005 23.887 23.889 init_scf_loop 11 6.9 0.000 0.001 22.248 22.250 cp_fm_diag_elpa 83 13.4 0.000 0.001 20.462 20.490 cp_fm_diag_elpa_base 83 14.4 20.358 20.397 20.457 20.486 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.949 20.125 multiply_cannon_metrocomm3 60648 15.6 0.113 0.117 15.929 18.185 prepare_preconditioner 11 7.9 0.000 0.000 17.532 17.579 make_preconditioner 11 8.9 0.000 0.000 17.532 17.579 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.760 16.970 sum_up_and_integrate 129 10.3 0.090 0.110 15.350 15.366 integrate_v_rspace 129 11.3 0.004 0.005 15.260 15.278 qs_rho_update_rho_low 129 7.7 0.001 0.001 15.113 15.262 calculate_rho_elec 129 8.7 0.046 0.065 15.113 15.262 make_m2s 5054 13.6 0.105 0.113 14.673 15.052 make_images 5054 14.6 0.409 0.429 14.495 14.885 init_scf_run 11 5.9 0.000 0.001 14.809 14.810 scf_env_initial_rho_setup 11 6.9 0.000 0.001 14.809 14.809 density_rs2pw 129 9.7 0.006 0.007 8.235 11.568 rs_pw_transfer 1054 12.0 0.017 0.020 7.137 10.554 cp_fm_cholesky_invert 11 10.9 10.148 10.156 10.148 10.156 wfi_extrapolate 11 7.9 0.001 0.001 9.347 9.347 mp_sum_l 8010 12.9 8.371 9.270 8.371 9.270 calculate_dm_sparse 129 9.5 0.001 0.001 8.958 9.066 dbcsr_mm_accdrv_process 125468 16.2 3.184 3.316 8.398 8.991 pw_transfer 1559 11.6 0.076 0.097 8.527 8.799 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 8.611 8.714 fft_wrap_pw1pw2 1301 12.7 0.011 0.013 8.322 8.597 make_images_data 5054 15.6 0.067 0.072 7.161 8.216 qs_ot_get_orbitals 118 10.6 0.001 0.001 8.081 8.155 multiply_cannon_metrocomm1 60648 15.6 0.090 0.093 6.289 8.059 grid_integrate_task_list 129 12.3 7.123 7.720 7.123 7.720 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 7.674 7.688 hybrid_alltoall_any 5240 16.5 0.294 2.292 6.216 7.567 fft_wrap_pw1pw2_140 527 13.2 1.277 1.334 6.920 7.192 fft3d_ps 1301 14.7 2.157 2.793 6.073 6.470 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.275 6.364 mp_alltoall_d11v 2423 14.1 4.600 6.168 4.600 6.168 mp_waitany 16140 13.9 2.712 6.080 2.712 6.080 grid_collocate_task_list 129 9.7 4.716 6.008 4.716 6.008 potential_pw2rs 129 12.3 0.009 0.011 5.533 5.650 rs_pw_transfer_RS2PW_140 140 11.5 0.284 0.310 2.172 5.542 calculate_first_density_matrix 1 7.0 0.000 0.000 5.269 5.279 cp_fm_cholesky_decompose 22 10.9 5.176 5.189 5.176 5.189 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=220.137000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=598.181818, yerr=6.833317 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 833.294336E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6106579. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.075 199.404 199.405 qs_mol_dyn_low 1 2.0 0.003 0.004 198.790 198.823 qs_forces 11 3.9 0.053 0.207 198.258 198.260 qs_energies 11 4.9 0.006 0.028 191.074 191.086 scf_env_do_scf 11 5.9 0.001 0.003 174.274 174.285 scf_env_do_scf_inner_loop 117 6.6 0.006 0.020 139.513 139.516 velocity_verlet 10 3.0 0.001 0.002 124.706 124.717 dbcsr_multiply_generic 2507 12.6 0.188 0.193 98.167 99.409 qs_scf_new_mos 117 7.6 0.001 0.001 98.237 98.629 qs_scf_loop_do_ot 117 8.6 0.001 0.001 98.236 98.629 ot_scf_mini 117 9.6 0.004 0.005 93.409 93.861 multiply_cannon 2507 13.6 0.476 0.523 77.484 81.990 multiply_cannon_loop 2507 14.6 1.240 1.284 74.160 76.570 ot_mini 117 10.6 0.001 0.001 50.485 50.893 mp_waitall_1 214728 16.6 24.986 38.908 24.986 38.908 multiply_cannon_multrec 30084 15.6 22.116 26.979 31.733 37.114 init_scf_loop 11 6.9 0.002 0.005 34.659 34.661 rebuild_ks_matrix 128 8.3 0.001 0.001 34.025 34.487 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 34.025 34.486 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.537 30.963 prepare_preconditioner 11 7.9 0.000 0.001 30.176 30.256 make_preconditioner 11 8.9 0.000 0.001 30.176 30.256 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.828 29.448 qs_ot_get_derivative 117 11.6 0.001 0.002 28.510 28.954 multiply_cannon_metrocomm3 30084 15.6 0.091 0.097 15.556 28.183 qs_ot_get_p 128 10.4 0.001 0.001 24.553 25.051 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.049 23.253 apply_single 128 13.6 0.001 0.001 22.048 23.253 multiply_cannon_sync_h2d 30084 15.6 19.352 22.413 19.352 22.413 ot_diis_step 117 11.6 0.014 0.015 21.802 21.805 qs_ot_p2m_diag 83 11.4 0.190 0.224 19.561 19.603 cp_dbcsr_syevd 83 12.4 0.006 0.006 18.340 18.344 cp_fm_cholesky_invert 11 10.9 17.780 17.793 17.780 17.793 make_m2s 5014 13.6 0.091 0.097 14.845 16.390 make_images 5014 14.6 1.164 1.351 14.633 16.180 sum_up_and_integrate 128 10.3 0.116 0.131 15.278 15.308 integrate_v_rspace 128 11.3 0.004 0.004 15.162 15.194 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.838 14.869 calculate_rho_elec 128 8.7 0.089 0.106 14.837 14.869 cp_fm_diag_elpa 83 13.4 0.000 0.001 14.762 14.798 cp_fm_diag_elpa_base 83 14.4 14.472 14.607 14.756 14.786 init_scf_run 11 5.9 0.000 0.001 11.857 11.859 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.857 11.859 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 11.300 11.628 make_images_data 5014 15.6 0.064 0.073 8.906 10.696 multiply_cannon_metrocomm4 27577 15.6 0.104 0.117 3.754 10.615 density_rs2pw 128 9.7 0.006 0.007 8.064 10.500 mp_irecv_dv 69486 16.3 3.554 10.227 3.554 10.227 hybrid_alltoall_any 5200 16.5 0.345 1.535 7.610 10.191 dbcsr_mm_accdrv_process 62242 16.2 4.628 5.308 9.073 9.671 pw_transfer 1547 11.6 0.087 0.099 9.058 9.116 rs_pw_transfer 1046 11.9 0.015 0.017 6.657 9.075 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 8.833 8.897 wfi_extrapolate 11 7.9 0.001 0.001 8.665 8.665 grid_integrate_task_list 128 12.3 7.187 7.792 7.187 7.792 fft_wrap_pw1pw2_140 523 13.2 1.337 1.359 7.618 7.703 cp_fm_cholesky_decompose 22 10.9 7.447 7.523 7.447 7.523 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.325 7.081 calculate_dm_sparse 128 9.5 0.001 0.001 6.518 6.661 fft3d_ps 1291 14.7 2.810 2.977 6.137 6.203 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.194 6.202 mp_sum_l 7950 12.9 4.084 5.924 4.084 5.924 grid_collocate_task_list 128 9.7 4.784 5.918 4.784 5.918 potential_pw2rs 128 12.3 0.015 0.017 5.517 5.552 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.414 5.477 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.313 5.437 mp_alltoall_d11v 2415 14.1 4.533 5.275 4.533 5.275 mp_allgather_i34 2507 14.6 1.888 5.078 1.888 5.078 mp_waitany 11748 13.9 2.554 5.006 2.554 5.006 rs_pw_transfer_RS2PW_140 139 11.5 0.355 0.377 2.120 4.527 dbcsr_complete_redistribute 395 12.7 0.773 0.855 3.311 4.196 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=199.405000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=795.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420243808256 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 11528908111872 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 93.514772E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755942624 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 950.857728E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531401248 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57952. MP_Allreduce 11127 998. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.036 187.031 187.032 qs_mol_dyn_low 1 2.0 0.003 0.004 186.292 186.565 qs_forces 11 3.9 0.004 0.004 185.731 185.736 qs_energies 11 4.9 0.002 0.005 178.968 178.980 scf_env_do_scf 11 5.9 0.001 0.002 162.408 162.409 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 124.370 124.371 velocity_verlet 10 3.0 0.155 0.180 119.987 119.998 qs_scf_new_mos 117 7.6 0.001 0.001 84.773 85.106 qs_scf_loop_do_ot 117 8.6 0.001 0.001 84.772 85.105 dbcsr_multiply_generic 2507 12.6 0.179 0.185 83.493 84.661 ot_scf_mini 117 9.6 0.004 0.008 80.555 80.923 multiply_cannon 2507 13.6 0.502 0.521 62.323 67.485 multiply_cannon_loop 2507 14.6 0.852 0.883 58.995 61.955 ot_mini 117 10.6 0.001 0.001 43.855 44.210 init_scf_loop 11 6.9 0.001 0.004 37.935 37.935 mp_waitall_1 170520 16.6 26.529 36.237 26.529 36.237 prepare_preconditioner 11 7.9 0.000 0.000 33.790 33.834 make_preconditioner 11 8.9 0.000 0.001 33.790 33.834 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.259 32.772 rebuild_ks_matrix 128 8.3 0.001 0.001 31.620 32.100 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 31.620 32.099 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.448 28.893 multiply_cannon_metrocomm3 20056 15.6 0.057 0.061 15.590 25.263 multiply_cannon_multrec 20056 15.6 13.453 16.459 22.168 25.101 qs_ot_get_derivative 117 11.6 0.001 0.002 23.979 24.333 qs_ot_get_p 128 10.4 0.001 0.001 21.751 22.170 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.923 20.833 apply_single 128 13.6 0.001 0.001 19.922 20.833 ot_diis_step 117 11.6 0.018 0.020 19.739 19.740 qs_ot_p2m_diag 83 11.4 0.265 0.273 17.259 17.289 make_m2s 5014 13.6 0.081 0.085 15.662 16.681 make_images 5014 14.6 1.181 1.293 15.429 16.448 cp_fm_cholesky_invert 11 10.9 16.231 16.244 16.231 16.244 cp_dbcsr_syevd 83 12.4 0.005 0.006 16.137 16.139 multiply_cannon_sync_h2d 20056 15.6 14.137 15.512 14.137 15.512 sum_up_and_integrate 128 10.3 0.134 0.145 15.241 15.269 integrate_v_rspace 128 11.3 0.004 0.004 15.107 15.133 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.079 15.117 calculate_rho_elec 128 8.7 0.132 0.148 15.078 15.117 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.576 12.595 cp_fm_diag_elpa_base 83 14.4 12.125 12.299 12.571 12.590 init_scf_run 11 5.9 0.000 0.001 11.471 11.471 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.471 11.471 make_images_data 5014 15.6 0.059 0.067 9.780 11.131 hybrid_alltoall_any 5200 16.5 0.437 1.997 8.382 10.427 density_rs2pw 128 9.7 0.006 0.007 7.827 9.868 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.536 9.792 multiply_cannon_metrocomm4 17549 15.6 0.062 0.074 3.462 9.302 pw_transfer 1547 11.6 0.087 0.105 9.158 9.269 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.933 9.048 mp_irecv_dv 50230 16.2 3.342 9.045 3.342 9.045 dbcsr_mm_accdrv_process 41502 16.2 4.983 5.301 8.187 8.299 cp_fm_cholesky_decompose 22 10.9 8.225 8.253 8.225 8.253 rs_pw_transfer 1046 11.9 0.014 0.015 6.085 8.070 fft_wrap_pw1pw2_140 523 13.2 1.413 1.441 7.746 7.874 wfi_extrapolate 11 7.9 0.001 0.001 7.793 7.793 grid_integrate_task_list 128 12.3 7.257 7.767 7.257 7.767 cp_fm_upper_to_full 105 14.8 5.795 7.481 5.795 7.481 dbcsr_complete_redistribute 395 12.7 1.177 1.212 4.928 6.733 fft3d_ps 1291 14.7 2.738 2.971 6.054 6.136 calculate_dm_sparse 128 9.5 0.001 0.001 5.878 5.991 grid_collocate_task_list 128 9.7 4.997 5.818 4.997 5.818 mp_alltoall_d11v 2415 14.1 4.644 5.757 4.644 5.757 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.581 5.588 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.698 5.500 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.733 5.410 potential_pw2rs 128 12.3 0.021 0.023 5.275 5.303 mp_allgather_i34 2507 14.6 1.842 5.183 1.842 5.183 mp_sum_l 7950 12.9 3.699 5.164 3.699 5.164 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.613 4.740 mp_waitany 11748 13.9 2.416 4.537 2.416 4.537 transfer_fm_to_dbcsr 11 9.9 0.019 0.027 2.511 4.277 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.183 4.213 mp_alltoall_i22 716 14.1 2.137 4.122 2.137 4.122 rs_pw_transfer_RS2PW_140 139 11.5 0.336 0.357 1.946 3.962 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.777 3.779 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=187.032000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=904.000000, yerr=9.486833 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/18/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 4.353791E+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 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.148412E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.031 196.473 196.474 qs_mol_dyn_low 1 2.0 0.003 0.004 196.013 196.039 qs_forces 11 3.9 0.004 0.004 195.854 195.858 qs_energies 11 4.9 0.070 0.078 188.572 188.583 scf_env_do_scf 11 5.9 0.001 0.001 169.361 169.368 velocity_verlet 10 3.0 0.001 0.002 128.594 128.599 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 121.228 121.229 qs_scf_new_mos 117 7.6 0.001 0.001 82.786 83.048 qs_scf_loop_do_ot 117 8.6 0.001 0.001 82.785 83.047 dbcsr_multiply_generic 2507 12.6 0.185 0.190 80.993 81.732 ot_scf_mini 117 9.6 0.003 0.003 78.276 78.571 multiply_cannon 2507 13.6 0.556 0.596 56.059 59.272 multiply_cannon_loop 2507 14.6 1.180 1.207 52.131 53.851 init_scf_loop 11 6.9 0.000 0.000 48.009 48.010 prepare_preconditioner 11 7.9 0.000 0.000 43.869 43.907 make_preconditioner 11 8.9 0.000 0.000 43.869 43.907 ot_mini 117 10.6 0.001 0.001 43.025 43.313 make_full_inverse_cholesky 11 9.9 0.011 0.023 37.563 42.595 multiply_cannon_multrec 30084 15.6 14.259 19.494 26.696 31.577 rebuild_ks_matrix 128 8.3 0.001 0.001 30.415 30.701 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.020 30.415 30.700 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.378 27.636 mp_waitall_1 147882 16.7 17.646 27.322 17.646 27.322 qs_ot_get_derivative 117 11.6 0.002 0.002 23.281 23.579 make_m2s 5014 13.6 0.096 0.100 20.561 21.868 qs_ot_get_p 128 10.4 0.001 0.001 21.422 21.735 make_images 5014 14.6 1.942 2.249 20.255 21.563 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.119 19.690 apply_single 128 13.6 0.001 0.001 19.119 19.690 ot_diis_step 117 11.6 0.018 0.018 19.613 19.614 qs_ot_p2m_diag 83 11.4 0.343 0.390 17.226 17.278 cp_fm_cholesky_invert 11 10.9 17.231 17.240 17.231 17.240 cp_fm_upper_to_full 105 14.8 11.548 17.005 11.548 17.005 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.897 15.899 sum_up_and_integrate 128 10.3 0.140 0.152 15.238 15.265 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.184 15.214 calculate_rho_elec 128 8.7 0.175 0.192 15.183 15.214 integrate_v_rspace 128 11.3 0.004 0.004 15.098 15.129 multiply_cannon_metrocomm3 30084 15.6 0.048 0.050 6.369 14.893 init_scf_run 11 5.9 0.000 0.001 12.826 12.827 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.825 12.827 dbcsr_complete_redistribute 395 12.7 1.522 1.645 9.039 12.791 make_images_data 5014 15.6 0.061 0.066 11.041 12.621 cp_fm_diag_elpa 83 13.4 0.000 0.000 12.586 12.601 cp_fm_diag_elpa_base 83 14.4 11.583 11.915 12.578 12.591 multiply_cannon_sync_h2d 30084 15.6 11.647 12.550 11.647 12.550 dbcsr_mm_accdrv_process 62264 16.2 7.686 8.741 12.006 12.520 hybrid_alltoall_any 5200 16.5 0.527 2.215 9.922 11.987 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.609 11.391 transfer_fm_to_dbcsr 11 9.9 0.001 0.006 6.283 10.002 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.625 9.848 pw_transfer 1547 11.6 0.087 0.102 9.385 9.476 mp_alltoall_i22 716 14.1 5.529 9.269 5.529 9.269 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.158 9.256 density_rs2pw 128 9.7 0.006 0.006 7.469 8.806 cp_fm_cholesky_decompose 22 10.9 8.131 8.219 8.131 8.219 fft_wrap_pw1pw2_140 523 13.2 1.558 1.591 8.022 8.126 grid_integrate_task_list 128 12.3 7.460 8.002 7.460 8.002 wfi_extrapolate 11 7.9 0.001 0.001 7.875 7.875 multiply_cannon_metrocomm4 25070 15.6 0.080 0.090 2.766 6.992 calculate_dm_sparse 128 9.5 0.001 0.001 6.822 6.886 rs_pw_transfer 1046 11.9 0.013 0.014 5.351 6.744 mp_irecv_dv 76098 16.2 2.618 6.725 2.618 6.725 mp_alltoall_d11v 2415 14.1 5.126 6.241 5.126 6.241 fft3d_ps 1291 14.7 2.842 2.935 6.085 6.160 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.956 6.012 grid_collocate_task_list 128 9.7 5.153 5.868 5.153 5.868 potential_pw2rs 128 12.3 0.024 0.024 4.960 4.981 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.826 4.895 calculate_first_density_matrix 1 7.0 0.000 0.000 4.735 4.738 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.581 4.667 qs_energies_init_hamiltonians 11 5.9 0.028 0.031 4.495 4.496 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.285 4.342 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=196.474000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1079.090909, yerr=19.888117 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/19/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 5.865089E+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 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.531679E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+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 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.033 174.049 174.050 qs_mol_dyn_low 1 2.0 0.003 0.003 173.610 173.623 qs_forces 11 3.9 0.004 0.004 173.468 173.473 qs_energies 11 4.9 0.004 0.005 165.863 165.869 scf_env_do_scf 11 5.9 0.001 0.001 148.054 148.059 velocity_verlet 10 3.0 0.002 0.002 113.951 113.957 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 112.113 112.114 dbcsr_multiply_generic 2507 12.6 0.179 0.191 73.805 74.266 qs_scf_new_mos 117 7.6 0.001 0.001 73.852 73.972 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.851 73.972 ot_scf_mini 117 9.6 0.003 0.004 69.450 69.505 multiply_cannon 2507 13.6 0.587 0.612 54.382 58.533 multiply_cannon_loop 2507 14.6 0.447 0.457 49.689 50.322 ot_mini 117 10.6 0.001 0.001 39.474 39.525 init_scf_loop 11 6.9 0.000 0.000 35.793 35.795 mp_waitall_1 125778 16.7 25.221 32.229 25.221 32.229 prepare_preconditioner 11 7.9 0.000 0.000 31.892 31.912 make_preconditioner 11 8.9 0.000 0.000 31.892 31.912 make_full_inverse_cholesky 11 9.9 0.018 0.028 29.801 30.072 rebuild_ks_matrix 128 8.3 0.001 0.001 29.632 29.673 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 29.632 29.672 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.832 26.869 multiply_cannon_multrec 10028 15.6 10.469 15.589 17.972 21.536 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.536 19.900 apply_single 128 13.6 0.001 0.001 19.536 19.900 ot_diis_step 117 11.6 0.020 0.021 19.899 19.899 qs_ot_get_derivative 117 11.6 0.002 0.002 19.506 19.562 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.146 19.004 make_m2s 5014 13.6 0.066 0.070 15.865 18.325 cp_fm_cholesky_invert 11 10.9 18.244 18.250 18.244 18.250 make_images 5014 14.6 2.321 2.792 15.558 18.016 qs_ot_get_p 128 10.4 0.001 0.001 17.163 17.231 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.294 15.340 calculate_rho_elec 128 8.7 0.258 0.269 15.294 15.339 sum_up_and_integrate 128 10.3 0.179 0.189 14.931 14.971 integrate_v_rspace 128 11.3 0.004 0.004 14.752 14.801 qs_ot_p2m_diag 83 11.4 0.496 0.501 13.549 13.562 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.408 12.410 multiply_cannon_sync_h2d 10028 15.6 11.594 12.355 11.594 12.355 make_images_data 5014 15.6 0.051 0.060 9.508 11.970 hybrid_alltoall_any 5200 16.5 0.795 3.630 9.381 11.844 init_scf_run 11 5.9 0.000 0.001 10.666 10.666 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.666 10.666 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.381 9.393 cp_fm_diag_elpa_base 83 14.4 9.143 9.223 9.378 9.390 pw_transfer 1547 11.6 0.086 0.095 9.348 9.370 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.124 9.152 grid_integrate_task_list 128 12.3 7.766 8.174 7.766 8.174 fft_wrap_pw1pw2_140 523 13.2 1.917 1.958 8.086 8.115 cp_fm_cholesky_decompose 22 10.9 7.954 8.088 7.954 8.088 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 7.832 7.873 dbcsr_mm_accdrv_process 20762 16.1 2.744 3.492 7.129 7.790 density_rs2pw 128 9.7 0.006 0.006 6.778 7.685 wfi_extrapolate 11 7.9 0.001 0.001 7.503 7.503 multiply_cannon_metrocomm1 10028 15.6 0.029 0.029 4.315 7.276 mp_allgather_i34 2507 14.6 2.879 7.084 2.879 7.084 calculate_dm_sparse 128 9.5 0.001 0.001 6.117 6.212 grid_collocate_task_list 128 9.7 5.468 6.122 5.468 6.122 fft3d_ps 1291 14.7 2.768 2.846 5.596 5.627 qs_energies_init_hamiltonians 11 5.9 0.208 0.272 5.553 5.555 mp_alltoall_d11v 2415 14.1 4.986 5.551 4.986 5.551 dbcsr_complete_redistribute 395 12.7 2.116 2.191 5.118 5.541 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.315 5.330 multiply_cannon_metrocomm4 7521 15.6 0.023 0.027 1.873 5.086 rs_pw_transfer 1046 11.9 0.013 0.013 4.101 5.022 mp_irecv_dv 28860 15.9 1.837 4.995 1.837 4.995 potential_pw2rs 128 12.3 0.027 0.027 4.371 4.383 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.258 4.290 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.573 3.857 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.721 3.765 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.698 3.701 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.332 3.640 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.541 3.611 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=174.050000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1434.818182, yerr=54.139245 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 3.164959E+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 4057 58363. MP_Allreduce 11219 1499. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.042 308.514 308.515 qs_mol_dyn_low 1 2.0 0.003 0.004 307.928 307.954 qs_forces 11 3.9 0.005 0.009 307.365 307.368 qs_energies 11 4.9 0.002 0.002 297.954 297.964 scf_env_do_scf 11 5.9 0.001 0.001 274.734 274.750 velocity_verlet 10 3.0 0.002 0.003 221.215 221.282 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 147.380 147.383 init_scf_loop 11 6.9 0.000 0.000 127.098 127.102 prepare_preconditioner 11 7.9 0.000 0.000 121.752 121.774 make_preconditioner 11 8.9 0.000 0.000 121.752 121.774 make_full_inverse_cholesky 11 9.9 0.037 0.039 97.609 118.852 qs_scf_new_mos 118 7.6 0.001 0.001 94.982 95.101 qs_scf_loop_do_ot 118 8.6 0.001 0.001 94.981 95.100 ot_scf_mini 118 9.6 0.004 0.004 90.115 90.133 dbcsr_multiply_generic 2535 12.6 0.212 0.222 85.643 86.168 cp_fm_upper_to_full 106 14.8 53.510 76.683 53.510 76.683 multiply_cannon 2535 13.6 0.716 0.798 59.704 61.015 multiply_cannon_loop 2535 14.6 0.476 0.484 55.558 56.816 ot_mini 118 10.6 0.001 0.001 46.130 46.148 dbcsr_complete_redistribute 397 12.7 3.975 4.029 29.896 42.784 rebuild_ks_matrix 129 8.3 0.001 0.001 39.526 39.557 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 39.525 39.557 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.496 39.412 transfer_fm_to_dbcsr 11 9.9 0.029 0.034 24.098 36.886 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.466 36.489 mp_alltoall_i22 720 14.1 21.871 34.904 21.871 34.904 cp_fm_cholesky_invert 11 10.9 34.444 34.451 34.444 34.451 mp_waitall_1 104820 16.8 29.529 33.960 29.529 33.960 qs_ot_get_p 129 10.4 0.001 0.001 28.327 28.399 qs_ot_get_derivative 118 11.6 0.002 0.002 25.114 25.133 qs_ot_p2m_diag 84 11.4 0.889 0.894 24.097 24.125 make_m2s 5070 13.6 0.076 0.078 21.337 22.661 cp_dbcsr_syevd 84 12.4 0.006 0.006 22.286 22.288 make_images 5070 14.6 3.782 3.915 20.854 22.182 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.314 21.331 calculate_rho_elec 129 8.7 0.483 0.484 21.313 21.330 ot_diis_step 118 11.6 0.022 0.024 20.981 20.982 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.207 20.432 apply_single 129 13.6 0.001 0.001 20.206 20.432 sum_up_and_integrate 129 10.3 0.324 0.328 20.278 20.362 multiply_cannon_metrocomm3 10140 15.6 0.023 0.026 18.844 20.078 integrate_v_rspace 129 11.3 0.004 0.004 19.953 20.038 cp_fm_diag_elpa 84 13.4 0.000 0.000 18.848 18.849 cp_fm_diag_elpa_base 84 14.4 14.245 15.944 18.844 18.844 multiply_cannon_multrec 10140 15.6 10.604 12.406 18.250 18.353 multiply_cannon_sync_h2d 10140 15.6 15.789 15.813 15.789 15.813 hybrid_alltoall_any 5257 16.5 1.314 3.081 11.744 14.116 make_images_data 5070 15.6 0.058 0.064 11.723 14.102 init_scf_run 11 5.9 0.000 0.001 12.668 12.668 scf_env_initial_rho_setup 11 6.9 0.000 0.000 12.668 12.668 pw_transfer 1559 11.6 0.094 0.095 12.134 12.141 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.896 11.903 fft_wrap_pw1pw2_140 527 13.2 3.101 3.150 10.489 10.498 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 9.943 9.963 mp_alltoall_d11v 2429 14.1 8.877 9.847 8.877 9.847 cp_fm_cholesky_decompose 22 10.9 9.603 9.629 9.603 9.629 wfi_extrapolate 11 7.9 0.001 0.001 9.297 9.298 dbcsr_mm_accdrv_process 20958 16.1 3.883 5.820 7.405 9.295 grid_integrate_task_list 129 12.3 8.659 8.863 8.659 8.863 density_rs2pw 129 9.7 0.005 0.005 8.548 8.699 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.454 8.455 calculate_dm_sparse 129 9.5 0.001 0.001 6.907 7.024 fft3d_ps 1301 14.7 2.875 2.896 6.930 6.962 grid_collocate_task_list 129 9.7 6.493 6.573 6.493 6.573 rs_scatter_matrices 140 9.7 3.702 4.624 6.241 6.463 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.369 6.439 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.104 6.187 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=308.515000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2881.909091, yerr=160.940675 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 1.259848E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 9239733. MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.043 86.004 86.005 qs_energies 1 2.0 0.000 0.000 84.939 84.946 ls_scf 1 3.0 0.000 0.000 83.883 83.890 dbcsr_multiply_generic 111 6.7 0.015 0.016 72.743 72.929 multiply_cannon 111 7.7 0.017 0.021 55.955 57.259 multiply_cannon_loop 111 8.7 0.208 0.220 52.505 53.812 ls_scf_main 1 4.0 0.000 0.000 52.534 52.534 density_matrix_trs4 2 5.0 0.002 0.003 46.931 47.008 ls_scf_init_scf 1 4.0 0.000 0.000 28.307 28.308 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.169 27.222 mp_waitall_1 11031 10.9 22.581 26.158 22.581 26.158 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.004 25.054 25.072 multiply_cannon_multrec 2664 9.7 8.216 8.850 15.604 17.212 multiply_cannon_sync_h2d 2664 9.7 13.574 15.195 13.574 15.195 make_m2s 222 7.7 0.009 0.012 13.174 13.695 make_images 222 8.7 0.101 0.112 13.153 13.676 multiply_cannon_metrocomm1 2664 9.7 0.009 0.010 9.764 13.179 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.387 9.293 make_images_data 222 9.7 0.005 0.006 7.770 8.399 hybrid_alltoall_any 227 10.6 0.219 1.847 6.556 8.076 dbcsr_mm_accdrv_process 4760 10.4 0.510 0.641 7.007 7.980 dbcsr_mm_accdrv_process_sort 4760 11.4 6.298 7.175 6.298 7.175 calculate_norms 4752 9.8 5.517 6.167 5.517 6.167 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.030 5.141 mp_sum_l 887 5.1 3.084 4.626 3.084 4.626 multiply_cannon_metrocomm4 2442 9.7 0.011 0.014 2.041 3.895 mp_irecv_dv 6231 10.9 2.025 3.865 2.025 3.865 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.319 3.572 make_images_sizes 222 9.7 0.000 0.000 0.796 3.459 mp_alltoall_i44 222 10.7 0.796 3.458 0.796 3.458 arnoldi_extremal 4 6.8 0.000 0.000 3.235 3.264 arnoldi_normal_ev 4 7.8 0.002 0.012 3.235 3.264 build_subspace 16 8.4 0.009 0.012 3.131 3.133 ls_scf_post 1 4.0 0.000 0.000 3.042 3.049 ls_scf_store_result 1 5.0 0.000 0.000 2.861 2.905 dbcsr_special_finalize 555 9.7 0.005 0.006 2.244 2.774 dbcsr_merge_single_wm 555 10.7 0.452 0.575 2.236 2.765 make_images_pack 222 9.7 2.215 2.630 2.217 2.632 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.318 2.577 dbcsr_sort_data 658 11.4 2.042 2.531 2.042 2.531 dbcsr_matrix_vector_mult_local 304 10.0 2.067 2.476 2.069 2.478 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.371 2.453 buffer_matrices_ensure_size 222 8.7 1.746 2.089 1.746 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.803 1.805 rebuild_ks_matrix 3 7.3 0.000 0.000 1.794 1.795 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.794 1.795 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.005000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1139.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 2.090025E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.054 91.957 91.958 qs_energies 1 2.0 0.000 0.000 91.335 91.339 ls_scf 1 3.0 0.000 0.000 89.935 89.938 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.722 76.026 multiply_cannon 111 7.7 0.030 0.044 53.632 57.130 ls_scf_main 1 4.0 0.000 0.000 55.206 55.212 multiply_cannon_loop 111 8.7 0.116 0.123 50.320 53.137 density_matrix_trs4 2 5.0 0.002 0.003 49.424 49.631 ls_scf_init_scf 1 4.0 0.000 0.000 31.178 31.179 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.935 30.043 mp_waitall_1 9105 10.9 21.204 29.099 21.204 29.099 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.446 27.467 multiply_cannon_multrec 1332 9.7 13.304 16.794 22.629 27.216 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.930 20.122 make_m2s 222 7.7 0.008 0.009 15.284 15.881 make_images 222 8.7 1.585 1.932 15.253 15.850 dbcsr_mm_accdrv_process 4041 10.4 0.286 0.447 8.917 10.516 dbcsr_mm_accdrv_process_sort 4041 11.4 8.512 10.077 8.512 10.077 make_images_data 222 9.7 0.004 0.005 8.717 9.686 hybrid_alltoall_any 227 10.6 0.522 2.428 8.152 9.054 mp_sum_l 887 5.1 5.514 8.098 5.514 8.098 multiply_cannon_metrocomm4 1221 9.7 0.006 0.009 3.253 7.713 mp_irecv_dv 3311 11.0 3.233 7.658 3.233 7.658 calculate_norms 2376 9.8 5.989 6.765 5.989 6.765 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.204 6.508 multiply_cannon_sync_h2d 1332 9.7 4.744 5.838 4.744 5.838 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.075 5.250 arnoldi_extremal 4 6.8 0.000 0.000 4.701 4.723 arnoldi_normal_ev 4 7.8 0.001 0.005 4.701 4.723 build_subspace 16 8.4 0.014 0.021 4.443 4.447 ls_scf_post 1 4.0 0.000 0.000 3.551 3.556 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.199 3.428 ls_scf_store_result 1 5.0 0.000 0.000 3.251 3.375 dbcsr_matrix_vector_mult_local 304 10.0 2.777 3.269 2.779 3.271 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.189 2.843 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.636 2.713 make_images_pack 222 9.7 2.026 2.464 2.029 2.466 mp_allgather_i34 111 8.7 1.004 2.305 1.004 2.305 dbcsr_sort_data 436 11.2 1.803 2.073 1.803 2.073 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.889 1.891 rebuild_ks_matrix 3 7.3 0.000 0.000 1.876 1.878 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.006 1.876 1.878 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.958000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1850.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 2.776252E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.050 94.254 94.255 qs_energies 1 2.0 0.000 0.000 93.583 93.655 ls_scf 1 3.0 0.000 0.000 91.977 92.048 dbcsr_multiply_generic 111 6.7 0.016 0.016 76.227 76.481 ls_scf_main 1 4.0 0.000 0.000 57.384 57.389 multiply_cannon 111 7.7 0.037 0.074 52.690 56.864 multiply_cannon_loop 111 8.7 0.100 0.105 49.055 52.721 density_matrix_trs4 2 5.0 0.002 0.003 51.381 51.591 mp_waitall_1 7281 11.0 23.133 33.028 23.133 33.028 ls_scf_init_scf 1 4.0 0.000 0.000 30.835 30.838 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.628 29.710 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.076 27.092 multiply_cannon_multrec 888 9.7 12.697 15.107 21.331 24.532 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.709 22.168 make_m2s 222 7.7 0.007 0.008 16.884 18.214 make_images 222 8.7 1.986 2.303 16.845 18.174 hybrid_alltoall_any 227 10.6 0.625 2.867 9.138 10.548 make_images_data 222 9.7 0.004 0.005 9.518 10.536 dbcsr_mm_accdrv_process 3754 10.4 0.232 0.431 8.155 9.396 dbcsr_mm_accdrv_process_sort 3754 11.4 7.767 8.965 7.767 8.965 mp_sum_l 887 5.1 5.095 8.505 5.095 8.505 multiply_cannon_sync_h2d 888 9.7 6.002 7.722 6.002 7.722 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.481 7.174 mp_irecv_dv 2335 11.1 2.465 7.135 2.465 7.135 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.926 6.673 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.533 6.141 arnoldi_extremal 4 6.8 0.000 0.000 5.153 5.170 arnoldi_normal_ev 4 7.8 0.001 0.005 5.153 5.170 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.969 5.144 build_subspace 16 8.4 0.014 0.020 4.851 4.857 calculate_norms 1584 9.8 4.345 4.758 4.345 4.758 ls_scf_post 1 4.0 0.000 0.000 3.758 3.832 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.482 3.825 mp_allgather_i34 111 8.7 1.456 3.740 1.456 3.740 dbcsr_matrix_vector_mult_local 304 10.0 3.053 3.638 3.055 3.640 ls_scf_store_result 1 5.0 0.000 0.000 3.301 3.386 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.889 3.010 dbcsr_sort_data 325 11.1 1.909 2.181 1.909 2.181 make_images_pack 222 9.7 1.820 2.101 1.823 2.104 dbcsr_data_release 9322 10.9 1.331 1.977 1.331 1.977 make_images_sizes 222 9.7 0.000 0.000 0.982 1.958 mp_alltoall_i44 222 10.7 0.981 1.958 0.981 1.958 dbcsr_finalize 304 7.8 0.027 0.032 1.618 1.904 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.900 1.902 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.255000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2198.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 3.343626E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.088 99.188 99.189 qs_energies 1 2.0 0.000 0.000 98.411 98.416 ls_scf 1 3.0 0.000 0.000 96.705 96.712 dbcsr_multiply_generic 111 6.7 0.018 0.020 79.083 79.330 ls_scf_main 1 4.0 0.001 0.013 59.260 59.263 multiply_cannon 111 7.7 0.068 0.180 52.084 56.867 density_matrix_trs4 2 5.0 0.015 0.161 53.049 53.216 multiply_cannon_loop 111 8.7 0.123 0.150 46.965 50.437 ls_scf_init_scf 1 4.0 0.000 0.001 34.141 34.143 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.343 32.424 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.800 29.825 mp_waitall_1 6369 11.0 22.659 28.921 22.659 28.921 multiply_cannon_multrec 1332 9.7 14.256 17.626 22.215 25.303 make_m2s 222 7.7 0.021 0.053 20.973 22.439 make_images 222 8.7 3.152 3.629 20.907 22.390 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.294 17.523 make_images_data 222 9.7 0.004 0.005 11.587 13.229 hybrid_alltoall_any 227 10.6 0.803 3.767 10.879 12.656 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.409 7.586 9.104 dbcsr_mm_accdrv_process_sort 3641 11.4 7.212 8.692 7.212 8.692 mp_sum_l 887 5.1 4.435 8.207 4.435 8.207 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.363 6.440 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.092 6.140 mp_irecv_dv 3229 10.9 2.068 6.052 2.068 6.052 multiply_cannon_sync_h2d 1332 9.7 5.499 6.021 5.499 6.021 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.611 5.350 arnoldi_extremal 4 6.8 0.000 0.000 5.305 5.333 arnoldi_normal_ev 4 7.8 0.001 0.005 5.304 5.333 build_subspace 16 8.4 0.014 0.021 4.947 4.955 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.660 4.881 mp_allgather_i34 111 8.7 2.199 4.564 2.199 4.564 calculate_norms 2376 9.8 4.203 4.533 4.203 4.533 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.644 3.932 dbcsr_matrix_vector_mult_local 304 10.0 3.227 3.724 3.229 3.726 dbcsr_sort_data 658 11.4 3.060 3.469 3.060 3.469 ls_scf_post 1 4.0 0.000 0.001 3.304 3.309 dbcsr_special_finalize 555 9.7 0.006 0.007 2.800 3.277 dbcsr_merge_single_wm 555 10.7 0.533 0.656 2.792 3.269 ls_scf_store_result 1 5.0 0.000 0.000 3.014 3.106 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.048 3.095 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.442 2.443 dbcsr_data_release 10477 10.7 1.604 2.406 1.604 2.406 rebuild_ks_matrix 3 7.3 0.000 0.000 2.035 2.036 qs_ks_build_kohn_sham_matrix 3 8.3 0.003 0.012 2.035 2.036 dbcsr_finalize 304 7.8 0.050 0.062 1.821 1.985 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=99.189000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 4.685431E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.058 100.093 100.094 qs_energies 1 2.0 0.000 0.000 99.047 99.051 ls_scf 1 3.0 0.000 0.000 97.108 97.111 dbcsr_multiply_generic 111 6.7 0.018 0.018 78.166 78.361 ls_scf_main 1 4.0 0.000 0.000 62.430 62.431 multiply_cannon 111 7.7 0.101 0.184 55.713 60.551 density_matrix_trs4 2 5.0 0.002 0.003 55.242 55.350 multiply_cannon_loop 111 8.7 0.069 0.078 51.123 52.760 mp_waitall_1 5436 11.0 26.545 32.281 26.545 32.281 ls_scf_init_scf 1 4.0 0.000 0.000 30.926 30.926 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.660 29.694 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.372 27.384 multiply_cannon_multrec 444 9.7 14.085 16.085 21.214 24.318 make_m2s 222 7.7 0.006 0.006 17.557 20.119 make_images 222 8.7 3.738 4.450 17.494 20.058 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.754 16.798 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 5.777 14.867 make_images_data 222 9.7 0.003 0.004 9.813 12.241 hybrid_alltoall_any 227 10.6 0.790 3.771 9.612 12.035 multiply_cannon_sync_h2d 444 9.7 6.562 8.662 6.562 8.662 dbcsr_mm_accdrv_process 3003 10.4 0.160 0.401 6.834 7.933 dbcsr_mm_accdrv_process_sort 3003 11.4 6.491 7.577 6.491 7.577 mp_allgather_i34 111 8.7 2.809 7.061 2.809 7.061 arnoldi_extremal 4 6.8 0.000 0.000 6.008 6.020 arnoldi_normal_ev 4 7.8 0.001 0.005 6.007 6.020 build_subspace 16 8.4 0.015 0.019 5.616 5.629 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.663 4.824 mp_sum_l 887 5.1 2.936 4.655 2.936 4.655 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.262 4.474 dbcsr_matrix_vector_mult_local 304 10.0 3.733 4.217 3.735 4.219 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.573 3.961 mp_irecv_dv 1241 11.2 1.555 3.914 1.555 3.914 ls_scf_post 1 4.0 0.000 0.000 3.752 3.756 calculate_norms 792 9.8 3.561 3.753 3.561 3.753 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.606 3.684 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.965 3.637 ls_scf_store_result 1 5.0 0.000 0.000 3.398 3.448 make_images_sizes 222 9.7 0.000 0.000 1.031 3.309 mp_alltoall_i44 222 10.7 1.030 3.309 1.030 3.309 dbcsr_finalize 304 7.8 0.062 0.078 2.206 2.283 dbcsr_merge_all 275 8.9 0.475 0.527 2.053 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.084 2.085 rebuild_ks_matrix 3 7.3 0.000 0.000 2.050 2.051 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.050 2.051 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.094000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3723.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_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 8.715518E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.087 0.102 110.820 110.820 qs_energies 1 2.0 0.000 0.000 109.327 109.337 ls_scf 1 3.0 0.000 0.000 106.373 106.383 dbcsr_multiply_generic 111 6.7 0.023 0.027 78.969 79.080 ls_scf_main 1 4.0 0.000 0.000 66.505 66.507 density_matrix_trs4 2 5.0 0.002 0.003 57.169 57.238 multiply_cannon 111 7.7 0.159 0.244 50.640 52.953 multiply_cannon_loop 111 8.7 0.068 0.070 46.970 48.150 ls_scf_init_scf 1 4.0 0.000 0.000 35.806 35.807 ls_scf_init_matrix_S 1 5.0 0.000 0.000 34.075 34.088 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 30.656 30.663 mp_waitall_1 4527 11.1 22.652 27.007 22.652 27.007 make_m2s 222 7.7 0.007 0.007 24.511 25.688 make_images 222 8.7 4.595 5.003 24.401 25.576 multiply_cannon_multrec 444 9.7 17.856 18.535 22.787 23.752 make_images_data 222 9.7 0.003 0.004 13.376 16.358 hybrid_alltoall_any 227 10.6 1.669 3.662 13.108 16.268 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.566 11.695 multiply_cannon_sync_h2d 444 9.7 8.845 8.891 8.845 8.891 arnoldi_extremal 4 6.8 0.000 0.000 7.655 7.663 arnoldi_normal_ev 4 7.8 0.003 0.009 7.655 7.663 build_subspace 16 8.4 0.025 0.035 7.084 7.094 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.653 5.923 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.629 5.826 dbcsr_matrix_vector_mult_local 304 10.0 5.082 5.410 5.084 5.412 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.242 5.333 dbcsr_mm_accdrv_process 1814 10.4 0.444 0.871 4.764 5.148 dbcsr_mm_accdrv_process_sort 1814 11.4 4.180 4.306 4.180 4.306 mp_allgather_i34 111 8.7 1.293 4.161 1.293 4.161 ls_scf_post 1 4.0 0.000 0.000 4.061 4.071 make_images_sizes 222 9.7 0.000 0.000 1.484 3.669 mp_alltoall_i44 222 10.7 1.484 3.669 1.484 3.669 ls_scf_store_result 1 5.0 0.000 0.000 3.434 3.444 calculate_norms 792 9.8 3.242 3.281 3.242 3.281 dbcsr_finalize 304 7.8 0.082 0.089 3.091 3.188 dbcsr_merge_all 275 8.9 0.893 0.922 2.876 2.967 qs_energies_init_hamiltonians 1 3.0 0.001 0.001 2.924 2.924 dbcsr_complete_redistribute 5 7.6 1.443 1.495 2.775 2.898 dbcsr_data_release 12724 10.6 2.332 2.870 2.332 2.870 matrix_ls_to_qs 2 6.0 0.000 0.000 2.428 2.568 dbcsr_sort_data 325 11.1 2.444 2.515 2.444 2.515 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.490 2.491 mp_sum_l 887 5.1 1.518 2.473 1.518 2.473 rebuild_ks_matrix 3 7.3 0.000 0.000 2.423 2.425 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.423 2.425 dbcsr_new_transposed 4 7.5 0.277 0.320 2.387 2.410 dbcsr_frobenius_norm 74 6.6 2.056 2.135 2.196 2.244 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.820000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6831.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/1501a5cde42d9e664f251d02093fe2fc81c3abfc_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 590.200832E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4037197. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.113 0.281 268.384 268.386 qs_mol_dyn_low 1 2.0 0.004 0.034 266.729 266.750 qs_forces 5 3.8 0.006 0.046 266.549 266.552 qs_energies 5 4.8 0.004 0.037 262.815 262.843 scf_env_do_scf 5 5.8 0.001 0.019 244.791 244.794 scf_env_do_scf_inner_loop 105 6.6 0.006 0.092 213.000 213.020 qs_scf_new_mos 105 7.6 0.000 0.001 166.514 166.662 qs_scf_loop_do_ot 105 8.6 0.001 0.001 166.513 166.661 ot_scf_mini 105 9.6 0.003 0.010 156.717 156.852 dbcsr_multiply_generic 1445 12.2 0.125 0.135 136.296 136.924 multiply_cannon 1445 13.2 0.275 0.293 116.055 118.675 multiply_cannon_loop 1445 14.2 2.395 2.516 113.671 115.551 velocity_verlet 4 3.0 0.002 0.011 111.746 111.747 qs_ot_get_p 112 10.4 0.001 0.034 65.907 66.225 ot_mini 105 10.6 0.001 0.011 60.749 60.865 qs_ot_p2m_diag 40 11.0 0.020 0.031 54.233 54.303 cp_dbcsr_syevd 40 12.0 0.002 0.010 50.931 50.933 cp_fm_syevd 40 13.0 0.001 0.017 44.988 45.151 multiply_cannon_multrec 69360 15.2 31.607 36.888 39.420 45.016 mp_waitall_1 488190 16.1 35.022 42.090 35.022 42.090 cp_fm_redistribute_end 40 14.0 19.814 39.520 19.821 39.525 cp_fm_syevd_base 40 14.0 19.691 39.412 19.691 39.412 qs_ot_get_derivative 55 11.6 0.001 0.009 38.904 39.028 multiply_cannon_sync_h2d 69360 15.2 30.509 34.855 30.509 34.855 multiply_cannon_metrocomm3 69360 15.2 0.197 0.206 24.870 32.826 rebuild_ks_matrix 110 8.4 0.000 0.000 31.705 31.889 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.023 31.705 31.888 init_scf_loop 7 6.6 0.001 0.013 31.750 31.753 qs_ks_update_qs_env 112 7.6 0.001 0.001 28.664 28.832 prepare_preconditioner 7 7.6 0.000 0.000 26.391 26.420 make_preconditioner 7 8.6 0.000 0.002 26.391 26.420 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.137 23.380 apply_single 62 13.6 0.000 0.000 23.137 23.380 qs_rho_update_rho_low 110 7.6 0.002 0.121 21.872 22.178 calculate_rho_elec 110 8.6 0.030 0.033 21.870 22.177 ot_new_cg_direction 55 11.6 0.001 0.019 21.132 21.133 make_full_inverse_cholesky 7 9.6 0.000 0.000 17.883 17.964 rs_pw_transfer 690 11.5 0.011 0.012 16.674 17.907 density_rs2pw 110 9.6 0.006 0.008 15.434 16.731 init_scf_run 5 5.8 0.000 0.001 15.159 15.160 scf_env_initial_rho_setup 5 6.8 0.000 0.002 15.159 15.160 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.794 15.007 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 13.853 13.929 pw_transfer 1645 12.4 0.086 0.114 13.219 13.543 fft_wrap_pw1pw2 1425 13.5 0.014 0.018 13.073 13.402 mp_sum_l 4764 12.2 12.158 12.994 12.158 12.994 cp_fm_cholesky_invert 7 10.6 12.032 12.046 12.032 12.046 calculate_dm_sparse 110 9.5 0.000 0.001 11.425 11.629 fft_wrap_pw1pw2_240 915 15.0 1.252 1.353 11.020 11.340 sum_up_and_integrate 60 10.3 0.028 0.031 10.905 10.936 integrate_v_rspace 60 11.3 0.002 0.004 10.877 10.910 check_diag 80 13.5 8.587 8.859 10.643 10.802 qs_vxc_create 110 10.4 0.002 0.006 10.701 10.755 qs_ot_get_derivative_diag 18 12.0 0.000 0.002 10.551 10.609 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 10.237 10.253 calculate_first_density_matrix 1 7.0 0.000 0.004 9.912 10.026 acc_transpose_blocks 69360 15.2 0.364 0.385 9.391 9.811 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 9.642 9.715 fft3d_pb 915 16.0 2.389 2.761 9.013 9.455 dbcsr_mm_accdrv_process 154766 15.8 4.017 4.170 7.676 8.498 make_full_single_inverse 7 9.6 0.015 0.318 8.179 8.493 make_m2s 2890 13.2 0.082 0.091 7.697 8.405 make_images 2890 14.2 0.243 0.262 7.586 8.293 acc_transpose_blocks_kernels 69360 16.2 0.852 0.906 7.452 7.753 xc_rho_set_and_dset_create 110 12.4 0.077 0.097 7.409 7.664 xc_vxc_pw_create 60 11.3 0.039 0.050 7.205 7.258 multiply_cannon_metrocomm1 69360 15.2 0.093 0.099 4.232 7.113 potential_pw2rs 60 12.3 0.003 0.003 6.891 6.934 jit_kernel_transpose 5 15.0 6.600 6.858 6.600 6.858 xc_pw_derive 510 13.4 0.006 0.007 6.402 6.474 mp_alltoall_z22v 2340 17.7 6.057 6.438 6.057 6.438 rs_pw_transfer_RS2PW_30 110 11.6 1.605 1.708 5.643 5.792 mp_waitany 7680 13.5 4.392 5.731 4.392 5.731 mp_sendrecv_dv 168740 12.6 5.244 5.378 5.244 5.378 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=268.386000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=559.800000, yerr=2.993326 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 1501a5cde42d9e664f251d02093fe2fc81c3abfc Summary: empty Status: OK