=== 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: 627b89bbdc49bd82d8905e4d10d261ed7123330b ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2023.05.001, 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/01 job id: 48408766 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/02 job id: 48408767 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/03 job id: 48408768 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/04 job id: 48408769 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/05 job id: 48408770 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/06 job id: 48408771 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/07 job id: 48408772 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/08 job id: 48408773 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/09 job id: 48408774 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/10 job id: 48408775 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/11 job id: 48408776 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/12 job id: 48408777 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/13 job id: 48408779 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/14 job id: 48408780 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/15 job id: 48408781 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/16 job id: 48408782 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/17 job id: 48408783 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/18 job id: 48408784 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/19 job id: 48408785 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/20 job id: 48408786 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/21 job id: 48408787 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/22 job id: 48408788 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/23 job id: 48408791 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/24 job id: 48408792 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/25 job id: 48408794 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/26 job id: 48408796 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/27 job id: 48408797 --- 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/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 0.022 0.035 133.234 133.235 farming_run 1 2.0 132.662 132.665 133.205 133.208 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.485222E+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 0.025 114.672 114.672 qs_energies 1 2.0 0.000 0.000 114.471 114.472 mp2_main 1 3.0 0.000 0.000 112.342 112.343 mp2_gpw_main 1 4.0 0.031 0.037 111.313 111.315 mp2_ri_gpw_compute_in 1 5.0 0.171 0.175 92.251 92.716 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.675 55.141 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.165 41.058 46.202 get_2c_integrals 1 6.0 0.013 0.018 36.691 37.404 integrate_v_rspace 273 8.0 0.439 0.453 24.647 29.506 pw_transfer 6555 10.6 0.381 0.394 26.781 27.374 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 25.413 26.048 grid_integrate_task_list 273 9.0 20.555 25.865 20.555 25.865 fft_wrap_pw1pw2_100 2178 12.4 1.117 1.294 22.973 23.588 rpa_ri_compute_en 1 5.0 0.033 0.037 18.941 19.124 compute_2c_integrals 1 7.0 0.003 0.003 19.003 19.009 compute_2c_integrals_loop_lm 1 8.0 0.008 0.013 18.571 18.733 mp2_eri_2c_integrate_gpw 1 9.0 2.378 2.409 18.563 18.720 cp_fm_cholesky_decompose 12 8.2 17.623 18.369 17.623 18.369 cholesky_decomp 1 7.0 0.000 0.000 16.525 17.270 fft3d_s 5443 13.4 16.133 16.463 16.155 16.485 ao_to_mo_and_store_B_mult_1 272 7.0 10.780 15.319 10.780 15.319 calculate_wavefunction 272 8.0 5.405 5.555 12.320 12.974 rpa_num_int 1 6.0 0.000 0.002 10.704 10.714 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.654 10.684 calc_mat_Q 8 8.0 0.000 0.000 9.506 9.620 contract_S_to_Q 8 9.0 0.000 0.000 8.930 9.040 parallel_gemm_fm 14 9.1 0.000 0.000 8.504 8.660 parallel_gemm_fm_cosma 14 10.1 8.504 8.660 8.504 8.660 calc_potential_gpw 544 9.5 0.005 0.005 8.159 8.626 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.051 8.276 potential_pw2rs 545 10.0 0.108 0.109 7.530 8.111 create_integ_mat 1 6.0 0.016 0.029 7.661 7.670 collocate_single_gaussian 272 10.0 0.039 0.042 7.328 7.614 array2fm 1 7.0 0.000 0.000 6.797 7.110 pw_scatter_s 2720 13.7 4.333 4.571 4.333 4.571 pw_gather_s 2722 13.2 3.434 3.676 3.434 3.676 array2fm_buffer_send 1 8.0 2.945 3.157 2.945 3.157 pw_poisson_solve 545 10.5 1.124 1.204 2.195 2.358 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.312977, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2794.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 0.039 0.046 422.442 422.443 farming_run 1 2.0 421.551 421.567 422.392 422.396 ------------------------------------------------------------------------------- @@@@@@@@@@ Run number: 2 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 16777216 0.0% 0.0% 100.0% flops 14 x 32 x 32 565182464 0.0% 0.0% 100.0% flops 29 x 32 x 32 585367552 0.0% 0.0% 100.0% flops 14 x 14 x 32 626196480 0.0% 0.0% 100.0% flops 29 x 14 x 32 638582784 0.0% 0.0% 100.0% flops 14 x 29 x 32 638582784 0.0% 0.0% 100.0% flops 29 x 29 x 32 682057728 0.0% 0.0% 100.0% flops 14 x 32 x 14 897827141120 0.0% 0.0% 100.0% flops 29 x 32 x 14 929989394432 0.0% 0.0% 100.0% flops 14 x 32 x 29 929989394432 0.0% 0.0% 100.0% flops 29 x 32 x 29 963203301376 0.0% 0.0% 100.0% flops 32 x 32 x 14 1693481172992 0.0% 0.0% 100.0% flops 32 x 32 x 29 1753962643456 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 7.172206E+12 0.0% 0.0% 100.0% flops max/rank 150.696064E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 249788822 0.0% 0.0% 100.0% number of processed stacks 98736 0.0% 0.0% 100.0% average stack size 0.0 0.0 2529.9 marketing flops 7.174951E+12 ------------------------------------------------------------------------------- # multiplications 1140 max memory usage/rank 1.240383E+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 12032553. 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 0.012 0.032 217.920 217.920 qs_energies 1 2.0 0.003 0.013 217.626 217.641 scf_env_do_scf 1 3.0 0.000 0.000 115.447 115.447 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.529 114.537 rebuild_ks_matrix 4 6.0 0.000 0.000 114.528 114.536 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.062 114.528 114.536 hfx_ks_matrix 4 8.0 0.001 0.001 114.182 114.187 integrate_four_center 4 9.0 0.152 0.464 114.181 114.186 integrate_four_center_main 4 10.0 0.136 0.644 102.264 106.131 integrate_four_center_bin 263 11.0 102.128 105.610 102.128 105.610 mp2_main 1 3.0 0.006 0.036 101.876 101.882 mp2_gpw_main 1 4.0 0.055 0.096 100.990 101.002 init_scf_loop 1 4.0 0.000 0.000 97.266 97.266 mp2_ri_gpw_compute_in 1 5.0 0.087 0.172 73.950 75.045 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.754 54.798 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.488 46.275 integrate_v_rspace 95 8.0 0.396 0.566 27.965 32.583 pw_transfer 2240 10.6 0.145 0.159 29.488 29.865 ao_to_mo_and_store_B_mult_1 91 7.0 10.597 29.013 10.597 29.013 mp2_ri_gpw_compute_en 1 5.0 0.057 0.072 26.816 28.920 fft_wrap_pw1pw2 1868 11.4 0.018 0.021 28.491 28.881 grid_integrate_task_list 95 9.0 23.384 28.162 23.384 28.162 fft_wrap_pw1pw2_100 730 12.4 1.270 1.474 26.242 26.684 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.838 1.891 25.046 25.056 get_2c_integrals 1 6.0 0.001 0.008 20.079 20.116 compute_2c_integrals 1 7.0 0.003 0.004 19.058 19.073 compute_2c_integrals_loop_lm 1 8.0 0.003 0.012 18.602 18.916 mp2_eri_2c_integrate_gpw 1 9.0 1.731 1.867 18.599 18.915 fft3d_s 1823 13.4 18.511 18.790 18.524 18.803 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 18.179 18.179 calculate_wavefunction 91 8.0 2.005 2.030 9.614 9.838 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.589 8.835 9.395 potential_pw2rs 186 10.0 0.034 0.036 8.413 8.957 local_gemm 172 8.0 8.277 8.846 8.277 8.846 mp2_ri_gpw_compute_en_comm 22 7.0 0.497 0.514 7.996 8.503 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.138 8.487 calc_potential_gpw 182 9.5 0.002 0.002 7.897 8.175 collocate_single_gaussian 91 10.0 0.018 0.040 7.742 8.010 mp_sync 37 10.5 3.908 7.304 3.908 7.304 integrate_four_center_load 4 10.0 0.000 0.000 6.749 6.753 hfx_load_balance 1 11.0 0.000 0.000 6.749 6.753 mp_sendrecv_dm3 2068 8.0 6.033 6.526 6.033 6.526 mp2_ri_gpw_compute_en_ener 172 7.0 6.333 6.386 6.333 6.386 pw_gather_s 912 13.2 4.461 4.866 4.461 4.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.986006, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1498.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/03/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 29.277748E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 5055360 0.0% 0.0% 100.0% average stack size 0.0 0.0 29.1 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 452.542464E+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 592243. 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 0.028 0.046 51.411 51.413 qs_mol_dyn_low 1 2.0 0.004 0.006 51.007 51.018 qs_forces 11 3.9 0.004 0.014 50.935 50.938 qs_energies 11 4.9 0.018 0.033 49.447 49.463 scf_env_do_scf 11 5.9 0.001 0.004 43.386 43.386 scf_env_do_scf_inner_loop 108 6.5 0.003 0.010 41.339 41.341 dbcsr_multiply_generic 2286 12.5 0.095 0.106 31.539 31.914 qs_scf_new_mos 108 7.5 0.000 0.001 31.491 31.793 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.490 31.792 ot_scf_mini 108 9.5 0.002 0.003 29.920 30.107 multiply_cannon 2286 13.5 0.189 0.194 25.354 26.622 multiply_cannon_loop 2286 14.5 1.832 1.949 24.700 25.987 velocity_verlet 10 3.0 0.002 0.003 24.965 24.966 ot_mini 108 10.5 0.001 0.001 17.933 18.186 qs_ot_get_derivative 108 11.5 0.001 0.001 15.049 15.238 mp_waitall_1 245248 16.5 7.339 13.097 7.339 13.097 multiply_cannon_metrocomm3 54864 15.5 0.075 0.084 5.351 11.985 multiply_cannon_multrec 54864 15.5 3.730 6.006 7.687 11.112 qs_ot_get_p 119 10.4 0.001 0.002 7.657 7.941 rebuild_ks_matrix 119 8.3 0.000 0.000 7.776 7.919 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.013 7.776 7.919 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.845 6.977 multiply_cannon_sync_h2d 54864 15.5 5.294 6.036 5.294 6.036 mp_sum_l 7287 12.8 4.206 5.986 4.206 5.986 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.342 5.804 dbcsr_mm_accdrv_process 76910 16.1 1.818 2.872 3.870 5.521 qs_ot_p2m_diag 50 11.0 0.005 0.007 5.034 5.073 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.595 4.699 init_scf_run 11 5.9 0.002 0.009 4.675 4.677 scf_env_initial_rho_setup 11 6.9 0.011 0.018 4.673 4.677 sum_up_and_integrate 119 10.3 0.012 0.015 4.422 4.429 integrate_v_rspace 119 11.3 0.002 0.003 4.410 4.419 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.308 4.309 qs_rho_update_rho_low 119 7.7 0.001 0.003 4.035 4.171 calculate_rho_elec 119 8.7 0.011 0.017 4.035 4.170 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.110 4.110 cp_fm_redistribute_end 50 14.0 2.099 4.086 2.104 4.088 cp_fm_diag_elpa_base 50 14.0 1.979 3.984 1.983 3.991 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.906 3.102 apply_single 119 13.6 0.000 0.000 2.906 3.101 calculate_dm_sparse 119 9.5 0.000 0.001 2.841 2.998 acc_transpose_blocks 54864 15.5 0.232 0.262 2.245 2.788 ot_diis_step 108 11.5 0.006 0.006 2.669 2.669 jit_kernel_multiply 13 15.8 1.987 2.600 1.987 2.600 calculate_first_density_matrix 1 7.0 0.001 0.006 2.521 2.533 density_rs2pw 119 9.7 0.004 0.004 2.131 2.222 multiply_cannon_metrocomm1 54864 15.5 0.059 0.065 1.355 2.184 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.181 2.183 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.124 2.177 grid_integrate_task_list 119 12.3 2.020 2.103 2.020 2.103 wfi_extrapolate 11 7.9 0.001 0.001 2.065 2.065 init_scf_loop 11 6.9 0.001 0.008 2.027 2.032 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.927 1.977 mp_sum_d 4135 12.0 1.238 1.916 1.238 1.916 potential_pw2rs 119 12.3 0.004 0.004 1.793 1.806 make_m2s 4572 13.5 0.053 0.055 1.670 1.722 pw_transfer 1439 11.6 0.052 0.057 1.647 1.719 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.572 1.645 make_images 4572 14.5 0.133 0.139 1.588 1.640 transfer_rs2pw 487 10.6 0.005 0.006 1.387 1.544 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.428 1.449 acc_transpose_blocks_sync 164592 16.5 1.195 1.446 1.195 1.446 mp_waitany 12084 13.8 1.284 1.406 1.284 1.406 grid_collocate_task_list 119 9.7 1.350 1.403 1.350 1.403 mp_alltoall_d11v 2130 13.8 1.231 1.396 1.231 1.396 fft3d_ps 1201 14.6 0.374 0.482 1.285 1.350 transfer_pw2rs 487 13.2 0.006 0.006 1.305 1.313 fft_wrap_pw1pw2_140 487 13.2 0.142 0.156 1.214 1.291 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.663 1.114 acc_transpose_blocks_kernels 54864 16.5 0.250 0.389 0.772 1.040 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.413000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.239835 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.029632E+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 1778. 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 0.031 0.058 39.594 39.598 qs_mol_dyn_low 1 2.0 0.003 0.006 39.211 39.219 qs_forces 11 3.9 0.004 0.008 39.112 39.113 qs_energies 11 4.9 0.004 0.011 37.415 37.420 scf_env_do_scf 11 5.9 0.001 0.002 31.887 31.887 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 29.426 29.427 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.275 22.621 qs_scf_new_mos 108 7.5 0.001 0.001 20.719 20.943 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.718 20.943 ot_scf_mini 108 9.5 0.003 0.003 19.807 19.968 multiply_cannon 2286 13.5 0.210 0.218 17.175 18.645 velocity_verlet 10 3.0 0.007 0.027 18.495 18.500 multiply_cannon_loop 2286 14.5 1.195 1.260 15.999 17.443 ot_mini 108 10.5 0.001 0.001 12.205 12.435 mp_waitall_1 200699 16.5 5.644 10.953 5.644 10.953 qs_ot_get_derivative 108 11.5 0.001 0.001 9.842 10.003 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.120 9.658 multiply_cannon_multrec 27432 15.5 1.821 4.049 6.281 9.092 rebuild_ks_matrix 119 8.3 0.000 0.000 6.820 6.957 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.016 6.820 6.957 dbcsr_mm_accdrv_process 47894 16.0 3.616 5.789 4.379 6.514 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.985 6.110 qs_ot_get_p 119 10.4 0.001 0.001 4.687 4.904 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.781 4.631 init_scf_run 11 5.9 0.000 0.001 4.156 4.157 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.156 4.157 mp_sum_l 7287 12.8 2.041 4.080 2.041 4.080 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.007 4.064 apply_single 119 13.6 0.000 0.000 3.007 4.064 sum_up_and_integrate 119 10.3 0.025 0.028 3.741 3.747 integrate_v_rspace 119 11.3 0.002 0.003 3.716 3.724 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.670 3.702 calculate_rho_elec 119 8.7 0.021 0.024 3.670 3.702 qs_ot_p2m_diag 50 11.0 0.013 0.031 3.032 3.049 make_m2s 4572 13.5 0.052 0.053 2.628 2.960 make_images 4572 14.5 0.205 0.243 2.541 2.876 calculate_first_density_matrix 1 7.0 0.001 0.003 2.634 2.636 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.629 2.634 init_scf_loop 11 6.9 0.003 0.014 2.440 2.442 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.313 2.400 ot_diis_step 108 11.5 0.011 0.011 2.315 2.316 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.293 2.293 cp_fm_redistribute_end 50 14.0 1.162 2.267 1.165 2.270 multiply_cannon_sync_h2d 27432 15.5 1.689 2.230 1.689 2.230 cp_fm_diag_elpa_base 50 14.0 1.072 2.174 1.100 2.219 calculate_dm_sparse 119 9.5 0.000 0.001 2.138 2.217 acc_transpose_blocks 27432 15.5 0.113 0.121 1.701 2.081 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.004 2.068 2.070 density_rs2pw 119 9.7 0.004 0.004 1.962 2.030 grid_integrate_task_list 119 12.3 1.844 1.944 1.844 1.944 pw_transfer 1439 11.6 0.065 0.070 1.856 1.891 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.818 1.855 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.765 1.802 jit_kernel_multiply 8 16.4 0.704 1.631 0.704 1.631 make_images_data 4572 15.5 0.047 0.054 1.244 1.613 prepare_preconditioner 11 7.9 0.000 0.001 1.528 1.553 make_preconditioner 11 8.9 0.000 0.002 1.528 1.553 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.430 1.487 potential_pw2rs 119 12.3 0.006 0.006 1.474 1.484 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.108 1.470 wfi_extrapolate 11 7.9 0.001 0.001 1.462 1.462 fft_wrap_pw1pw2_140 487 13.2 0.160 0.169 1.400 1.437 fft3d_ps 1201 14.6 0.523 0.579 1.400 1.430 mp_alltoall_d11v 2130 13.8 1.217 1.357 1.217 1.357 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.337 1.345 grid_collocate_task_list 119 9.7 1.286 1.337 1.286 1.337 transfer_rs2pw 487 10.6 0.005 0.006 1.217 1.306 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.199 1.244 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 1.055 1.057 mp_sum_d 4135 12.0 0.561 1.053 0.561 1.053 acc_transpose_blocks_kernels 27432 16.5 0.191 0.281 0.745 1.005 mp_allgather_i34 2286 14.5 0.595 0.998 0.595 0.998 acc_transpose_blocks_sync 82296 16.5 0.817 0.941 0.817 0.941 transfer_pw2rs 487 13.2 0.004 0.005 0.930 0.937 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.598000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.545455, yerr=1.876342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 519.127040E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.035 0.066 32.704 32.706 qs_mol_dyn_low 1 2.0 0.003 0.003 32.284 32.347 qs_forces 11 3.9 0.003 0.003 32.223 32.225 qs_energies 11 4.9 0.002 0.006 30.665 30.668 scf_env_do_scf 11 5.9 0.001 0.002 25.920 25.920 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 23.405 23.409 dbcsr_multiply_generic 2286 12.5 0.096 0.099 17.151 17.233 qs_scf_new_mos 108 7.5 0.001 0.001 15.656 15.671 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.655 15.670 velocity_verlet 10 3.0 0.004 0.008 15.271 15.272 ot_scf_mini 108 9.5 0.002 0.003 14.891 14.902 multiply_cannon 2286 13.5 0.194 0.198 13.719 14.431 multiply_cannon_loop 2286 14.5 0.856 0.901 12.920 13.660 ot_mini 108 10.5 0.001 0.001 9.137 9.153 qs_ot_get_derivative 108 11.5 0.001 0.001 7.640 7.650 multiply_cannon_multrec 18288 15.5 1.884 2.855 6.960 7.246 rebuild_ks_matrix 119 8.3 0.000 0.000 6.028 6.041 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.027 6.040 dbcsr_mm_accdrv_process 38222 16.0 4.923 5.800 4.983 5.867 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.317 5.331 mp_waitall_1 158411 16.6 2.647 3.749 2.647 3.749 sum_up_and_integrate 119 10.3 0.030 0.031 3.602 3.607 integrate_v_rspace 119 11.3 0.003 0.003 3.571 3.580 qs_ot_get_p 119 10.4 0.005 0.038 3.556 3.577 init_scf_run 11 5.9 0.000 0.001 3.562 3.562 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.562 3.562 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.876 3.469 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.295 3.300 calculate_rho_elec 119 8.7 0.031 0.032 3.294 3.299 init_scf_loop 11 6.9 0.000 0.002 2.497 2.502 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.413 2.406 qs_ot_p2m_diag 50 11.0 0.019 0.065 2.350 2.358 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.002 2.342 apply_single 119 13.6 0.000 0.000 2.002 2.342 calculate_first_density_matrix 1 7.0 0.001 0.005 2.320 2.321 make_m2s 4572 13.5 0.044 0.045 1.942 2.094 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.043 2.050 make_images 4572 14.5 0.192 0.203 1.858 2.009 density_rs2pw 119 9.7 0.004 0.004 1.858 1.980 grid_integrate_task_list 119 12.3 1.810 1.895 1.810 1.895 calculate_dm_sparse 119 9.5 0.000 0.001 1.862 1.871 pw_transfer 1439 11.6 0.065 0.069 1.857 1.869 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.786 1.787 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.764 1.779 acc_transpose_blocks 18288 15.5 0.080 0.082 1.684 1.758 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.748 1.754 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.740 1.749 cp_fm_diag_elpa_base 50 14.0 1.716 1.728 1.738 1.747 prepare_preconditioner 11 7.9 0.000 0.000 1.655 1.658 make_preconditioner 11 8.9 0.002 0.016 1.655 1.658 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.637 1.643 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.519 1.602 mp_sum_l 7287 12.8 1.134 1.520 1.134 1.520 ot_diis_step 108 11.5 0.011 0.012 1.481 1.481 fft_wrap_pw1pw2_140 487 13.2 0.212 0.218 1.412 1.426 potential_pw2rs 119 12.3 0.007 0.008 1.366 1.372 fft3d_ps 1201 14.6 0.530 0.547 1.333 1.346 grid_collocate_task_list 119 9.7 1.236 1.320 1.236 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.193 1.197 wfi_extrapolate 11 7.9 0.001 0.001 1.178 1.178 multiply_cannon_sync_h2d 18288 15.5 1.052 1.169 1.052 1.169 transfer_rs2pw 487 10.6 0.005 0.005 1.002 1.131 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.967 0.985 make_images_data 4572 15.5 0.047 0.051 0.813 0.975 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.954 0.956 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.711 0.903 acc_transpose_blocks_kernels 18288 16.5 0.220 0.226 0.849 0.859 transfer_pw2rs 487 13.2 0.004 0.004 0.832 0.837 acc_transpose_blocks_sync 54864 16.5 0.736 0.804 0.736 0.804 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.787 0.788 mp_alltoall_d11v 2130 13.8 0.681 0.762 0.681 0.762 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.681 0.735 mp_alltoall_z22v 1201 16.6 0.651 0.725 0.651 0.725 cp_fm_cholesky_invert 11 10.9 0.705 0.709 0.705 0.709 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.706000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.181818, yerr=0.833196 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 556.355584E+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 0.016 0.055 36.921 36.923 qs_mol_dyn_low 1 2.0 0.003 0.003 36.697 36.705 qs_forces 11 3.9 0.002 0.003 36.624 36.630 qs_energies 11 4.9 0.001 0.002 34.917 34.925 scf_env_do_scf 11 5.9 0.000 0.001 29.649 29.650 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.271 26.271 dbcsr_multiply_generic 2286 12.5 0.102 0.126 20.309 20.459 velocity_verlet 10 3.0 0.001 0.002 18.515 18.517 qs_scf_new_mos 108 7.5 0.001 0.001 18.265 18.320 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.264 18.319 ot_scf_mini 108 9.5 0.002 0.003 17.231 17.279 multiply_cannon 2286 13.5 0.219 0.229 16.428 16.942 multiply_cannon_loop 2286 14.5 1.556 1.649 15.481 15.838 ot_mini 108 10.5 0.001 0.001 10.711 10.775 multiply_cannon_multrec 27432 15.5 2.504 3.341 9.075 9.586 qs_ot_get_derivative 108 11.5 0.001 0.001 8.815 8.864 dbcsr_mm_accdrv_process 47916 15.9 5.861 7.766 6.467 7.829 rebuild_ks_matrix 119 8.3 0.000 0.000 6.294 6.345 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.294 6.345 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.577 5.621 init_scf_run 11 5.9 0.000 0.001 3.825 3.826 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.825 3.825 qs_ot_get_p 119 10.4 0.001 0.001 3.646 3.712 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.104 3.568 sum_up_and_integrate 119 10.3 0.035 0.037 3.490 3.500 integrate_v_rspace 119 11.3 0.003 0.003 3.455 3.465 init_scf_loop 11 6.9 0.000 0.000 3.359 3.359 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.310 3.342 calculate_rho_elec 119 8.7 0.040 0.046 3.309 3.342 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.187 2.664 apply_single 119 13.6 0.000 0.000 2.187 2.664 acc_transpose_blocks 27432 15.5 0.121 0.125 2.486 2.586 prepare_preconditioner 11 7.9 0.000 0.000 2.524 2.532 make_preconditioner 11 8.9 0.002 0.012 2.524 2.532 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.130 2.460 make_m2s 4572 13.5 0.054 0.057 2.309 2.416 calculate_first_density_matrix 1 7.0 0.000 0.000 2.396 2.399 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.366 2.393 mp_waitall_1 137007 16.6 1.903 2.378 1.903 2.378 make_images 4572 14.5 0.273 0.334 2.201 2.307 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.210 2.221 calculate_dm_sparse 119 9.5 0.000 0.000 2.161 2.211 pw_transfer 1439 11.6 0.065 0.069 1.927 1.960 grid_integrate_task_list 119 12.3 1.825 1.924 1.825 1.924 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.918 1.919 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.835 1.872 density_rs2pw 119 9.7 0.004 0.004 1.773 1.862 ot_diis_step 108 11.5 0.012 0.012 1.850 1.850 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.794 1.794 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.737 1.755 mp_sum_l 7287 12.8 1.147 1.751 1.147 1.751 acc_transpose_blocks_sync 82296 16.5 1.464 1.560 1.464 1.560 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.546 1.556 cp_fm_diag_elpa_base 50 14.0 1.513 1.531 1.544 1.554 fft_wrap_pw1pw2_140 487 13.2 0.246 0.259 1.511 1.550 jit_kernel_multiply 8 15.7 0.540 1.510 0.540 1.510 fft3d_ps 1201 14.6 0.559 0.609 1.353 1.379 wfi_extrapolate 11 7.9 0.001 0.001 1.376 1.376 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.317 1.333 grid_collocate_task_list 119 9.7 1.251 1.328 1.251 1.328 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.849 1.307 potential_pw2rs 119 12.3 0.009 0.010 1.271 1.275 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.255 1.263 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.182 1.189 cp_fm_upper_to_full 72 14.2 0.807 1.151 0.807 1.151 transfer_rs2pw 487 10.6 0.004 0.005 0.971 1.082 dbcsr_complete_redistribute 329 12.2 0.120 0.155 0.758 1.042 make_images_data 4572 15.5 0.047 0.052 0.845 0.959 hybrid_alltoall_any 4725 16.4 0.065 0.154 0.738 0.897 acc_transpose_blocks_kernels 27432 16.5 0.275 0.284 0.871 0.888 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.800 0.882 mp_alltoall_d11v 2130 13.8 0.747 0.854 0.747 0.854 copy_fm_to_dbcsr 176 11.2 0.001 0.002 0.567 0.842 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.806 0.812 multiply_cannon_metrocomm1 27432 15.5 0.036 0.037 0.253 0.784 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=36.923000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=528.363636, yerr=3.053246 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 619.003904E+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 0.030 29.829 29.832 qs_mol_dyn_low 1 2.0 0.003 0.004 29.619 29.629 qs_forces 11 3.9 0.002 0.004 29.491 29.493 qs_energies 11 4.9 0.001 0.002 27.755 27.758 scf_env_do_scf 11 5.9 0.000 0.001 22.972 22.973 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 20.342 20.343 velocity_verlet 10 3.0 0.020 0.024 15.316 15.322 dbcsr_multiply_generic 2286 12.5 0.093 0.098 13.545 13.697 qs_scf_new_mos 108 7.5 0.001 0.001 12.459 12.490 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.458 12.489 ot_scf_mini 108 9.5 0.002 0.002 11.727 11.749 multiply_cannon 2286 13.5 0.224 0.227 10.752 11.389 multiply_cannon_loop 2286 14.5 0.645 0.669 9.808 10.033 ot_mini 108 10.5 0.001 0.001 6.761 6.787 multiply_cannon_multrec 9144 15.5 1.668 1.852 5.962 6.156 rebuild_ks_matrix 119 8.3 0.000 0.000 5.966 5.996 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.965 5.995 qs_ot_get_derivative 108 11.5 0.001 0.001 5.378 5.400 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.323 5.351 dbcsr_mm_accdrv_process 12550 15.8 3.506 4.181 4.183 4.275 sum_up_and_integrate 119 10.3 0.037 0.041 3.581 3.589 integrate_v_rspace 119 11.3 0.003 0.003 3.543 3.552 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.427 3.434 calculate_rho_elec 119 8.7 0.060 0.061 3.426 3.433 init_scf_run 11 5.9 0.000 0.001 3.329 3.329 scf_env_initial_rho_setup 11 6.9 0.002 0.007 3.329 3.329 qs_ot_get_p 119 10.4 0.001 0.001 3.000 3.027 init_scf_loop 11 6.9 0.000 0.000 2.611 2.612 calculate_first_density_matrix 1 7.0 0.000 0.000 2.171 2.173 mp_waitall_1 115863 16.7 1.722 2.127 1.722 2.127 pw_transfer 1439 11.6 0.065 0.068 2.037 2.050 make_m2s 4572 13.5 0.034 0.035 1.880 2.034 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.974 1.977 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.944 1.959 make_images 4572 14.5 0.269 0.299 1.792 1.943 grid_integrate_task_list 119 12.3 1.866 1.917 1.866 1.917 density_rs2pw 119 9.7 0.004 0.004 1.800 1.861 prepare_preconditioner 11 7.9 0.000 0.000 1.823 1.843 make_preconditioner 11 8.9 0.000 0.000 1.823 1.843 calculate_dm_sparse 119 9.5 0.000 0.000 1.773 1.802 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.780 1.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.717 1.751 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.728 1.729 fft_wrap_pw1pw2_140 487 13.2 0.322 0.333 1.607 1.622 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.603 1.604 acc_transpose_blocks 9144 15.5 0.041 0.043 1.444 1.480 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.436 1.444 cp_fm_diag_elpa_base 50 14.0 1.406 1.426 1.435 1.442 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.414 1.430 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.412 1.423 grid_collocate_task_list 119 9.7 1.298 1.379 1.298 1.379 jit_kernel_multiply 8 15.7 0.638 1.372 0.638 1.372 fft3d_ps 1201 14.6 0.558 0.571 1.358 1.372 ot_diis_step 108 11.5 0.013 0.013 1.368 1.368 potential_pw2rs 119 12.3 0.010 0.011 1.317 1.325 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.272 1.323 apply_single 119 13.6 0.000 0.000 1.272 1.323 qs_energies_init_hamiltonians 11 5.9 0.002 0.004 1.240 1.242 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.137 1.144 wfi_extrapolate 11 7.9 0.001 0.001 1.107 1.107 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.780 1.001 mp_alltoall_d11v 2130 13.8 0.860 0.975 0.860 0.975 make_images_data 4572 15.5 0.042 0.046 0.799 0.971 transfer_rs2pw 487 10.6 0.004 0.005 0.872 0.947 cp_fm_cholesky_invert 11 10.9 0.927 0.930 0.927 0.930 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.873 0.924 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.861 0.867 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.768 0.771 acc_transpose_blocks_sync 27432 16.5 0.719 0.753 0.719 0.753 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.412 0.736 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.684 0.728 mp_alltoall_z22v 1201 16.6 0.674 0.703 0.674 0.703 mp_allgather_i34 2286 14.5 0.256 0.702 0.256 0.702 transfer_pw2rs 487 13.2 0.003 0.004 0.699 0.701 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 0.668 0.671 multiply_cannon_metrocomm1 9144 15.5 0.024 0.024 0.485 0.664 mp_waitany 5200 13.7 0.525 0.600 0.525 0.600 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.832000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=586.363636, yerr=4.753424 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 764.309504E+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 0.020 0.036 42.883 42.895 qs_mol_dyn_low 1 2.0 0.003 0.003 42.615 42.623 qs_forces 11 3.9 0.003 0.004 42.525 42.528 qs_energies 11 4.9 0.002 0.003 40.514 40.519 scf_env_do_scf 11 5.9 0.001 0.001 34.724 34.725 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.890 26.891 velocity_verlet 10 3.0 0.002 0.002 24.159 24.166 dbcsr_multiply_generic 2286 12.5 0.101 0.105 19.111 19.357 qs_scf_new_mos 108 7.5 0.001 0.001 17.432 17.532 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.432 17.531 ot_scf_mini 108 9.5 0.002 0.002 16.264 16.367 multiply_cannon 2286 13.5 0.296 0.301 15.146 16.106 multiply_cannon_loop 2286 14.5 0.874 0.883 13.866 14.865 ot_mini 108 10.5 0.001 0.001 9.825 9.943 multiply_cannon_multrec 9144 15.5 3.390 4.731 8.795 8.886 qs_ot_get_derivative 108 11.5 0.001 0.001 7.768 7.872 init_scf_loop 11 6.9 0.000 0.000 7.806 7.808 rebuild_ks_matrix 119 8.3 0.000 0.001 7.164 7.309 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.164 7.308 prepare_preconditioner 11 7.9 0.000 0.000 6.837 6.851 make_preconditioner 11 8.9 0.000 0.000 6.837 6.851 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.408 6.730 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.454 6.584 dbcsr_mm_accdrv_process 12550 15.8 4.721 6.532 5.266 6.564 cp_fm_upper_to_full 72 14.2 3.140 4.512 3.140 4.512 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.034 4.039 calculate_rho_elec 119 8.7 0.118 0.121 4.033 4.038 sum_up_and_integrate 119 10.3 0.064 0.066 3.843 3.852 integrate_v_rspace 119 11.3 0.004 0.004 3.778 3.787 init_scf_run 11 5.9 0.000 0.001 3.688 3.688 scf_env_initial_rho_setup 11 6.9 0.001 0.003 3.688 3.688 qs_ot_get_p 119 10.4 0.001 0.001 3.485 3.627 mp_waitall_1 94719 16.7 2.309 3.361 2.309 3.361 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.847 3.240 dbcsr_complete_redistribute 329 12.2 0.287 0.292 1.979 2.815 pw_transfer 1439 11.6 0.069 0.070 2.615 2.627 make_m2s 4572 13.5 0.037 0.038 2.393 2.558 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.516 2.530 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.672 2.497 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.194 2.496 apply_single 119 13.6 0.000 0.000 2.194 2.496 make_images 4572 14.5 0.353 0.387 2.273 2.438 calculate_dm_sparse 119 9.5 0.000 0.000 2.271 2.284 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.337 2.248 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.423 2.241 calculate_first_density_matrix 1 7.0 0.000 0.000 2.221 2.223 mp_alltoall_i22 627 13.8 1.406 2.205 1.406 2.205 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.088 2.140 fft_wrap_pw1pw2_140 487 13.2 0.577 0.583 2.117 2.133 grid_integrate_task_list 119 12.3 2.074 2.121 2.074 2.121 density_rs2pw 119 9.7 0.004 0.004 2.064 2.091 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.055 2.056 ot_diis_step 108 11.5 0.014 0.014 2.032 2.032 acc_transpose_blocks 9144 15.5 0.044 0.045 1.883 1.939 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.901 1.902 qs_energies_init_hamiltonians 11 5.9 0.004 0.007 1.806 1.809 mp_sum_l 7287 12.8 1.033 1.782 1.033 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.693 1.694 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.642 1.685 fft3d_ps 1201 14.6 0.596 0.608 1.612 1.620 grid_collocate_task_list 119 9.7 1.517 1.558 1.517 1.558 cp_fm_cholesky_invert 11 10.9 1.473 1.477 1.473 1.477 wfi_extrapolate 11 7.9 0.001 0.001 1.402 1.402 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.397 1.397 cp_fm_diag_elpa_base 50 14.0 1.253 1.305 1.396 1.396 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.325 1.345 potential_pw2rs 119 12.3 0.014 0.015 1.289 1.292 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.080 1.283 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.272 1.278 mp_alltoall_d11v 2130 13.8 1.196 1.232 1.196 1.232 make_images_data 4572 15.5 0.046 0.049 1.018 1.229 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.131 1.144 acc_transpose_blocks_sync 27432 16.5 1.114 1.135 1.114 1.135 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.043 qs_create_task_list 11 7.9 0.027 0.029 0.958 0.970 generate_qs_task_list 11 8.9 0.367 0.386 0.931 0.942 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.921 0.935 mp_alltoall_z22v 1201 16.6 0.879 0.896 0.879 0.896 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.895000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=719.000000, yerr=10.144412 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 501.473280E+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 2347412. 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 0.026 0.044 82.521 82.522 qs_mol_dyn_low 1 2.0 0.003 0.022 82.173 82.183 qs_forces 11 3.9 0.005 0.009 82.096 82.100 qs_energies 11 4.9 0.001 0.003 79.206 79.225 scf_env_do_scf 11 5.9 0.000 0.001 70.229 70.231 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.615 64.616 dbcsr_multiply_generic 2055 12.4 0.108 0.117 50.948 51.241 qs_scf_new_mos 99 7.5 0.001 0.001 47.344 47.457 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.343 47.456 ot_scf_mini 99 9.5 0.002 0.002 44.945 45.036 velocity_verlet 10 3.0 0.001 0.002 43.741 43.741 multiply_cannon 2055 13.4 0.184 0.188 42.490 43.245 multiply_cannon_loop 2055 14.4 1.801 1.837 41.449 42.219 ot_mini 99 10.5 0.001 0.001 26.461 26.557 qs_ot_get_derivative 99 11.5 0.001 0.001 19.690 19.818 multiply_cannon_multrec 49320 15.4 11.423 12.056 17.402 18.023 rebuild_ks_matrix 110 8.3 0.000 0.001 14.649 14.807 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.648 14.806 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.848 12.994 mp_waitall_1 220248 16.4 10.688 11.476 10.688 11.476 multiply_cannon_sync_h2d 49320 15.4 9.610 10.174 9.610 10.174 qs_ot_get_p 110 10.4 0.001 0.001 9.962 10.070 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.646 8.203 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.212 7.746 apply_single 110 13.6 0.000 0.000 7.212 7.745 multiply_cannon_metrocomm3 49320 15.4 0.085 0.091 6.411 7.450 sum_up_and_integrate 110 10.3 0.036 0.043 7.116 7.130 integrate_v_rspace 110 11.3 0.003 0.004 7.080 7.102 init_scf_run 11 5.9 0.000 0.001 6.878 6.879 scf_env_initial_rho_setup 11 6.9 0.002 0.005 6.878 6.879 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.650 6.672 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.466 6.582 calculate_rho_elec 110 8.6 0.021 0.024 6.466 6.581 ot_diis_step 99 11.5 0.006 0.006 6.581 6.581 dbcsr_mm_accdrv_process 87628 16.1 3.015 3.112 5.848 6.111 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.800 5.801 init_scf_loop 11 6.9 0.000 0.000 5.582 5.583 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.237 5.288 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.220 5.243 cp_fm_diag_elpa_base 48 14.0 5.207 5.232 5.218 5.242 mp_sum_l 6594 12.7 4.030 4.678 4.030 4.678 wfi_extrapolate 11 7.9 0.001 0.001 4.112 4.112 make_m2s 4110 13.4 0.060 0.066 3.968 4.112 make_images 4110 14.4 0.177 0.190 3.873 4.019 calculate_dm_sparse 110 9.5 0.001 0.001 3.776 3.877 density_rs2pw 110 9.6 0.004 0.005 3.484 3.668 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.600 3.604 grid_integrate_task_list 110 12.3 3.265 3.431 3.265 3.431 prepare_preconditioner 11 7.9 0.000 0.000 3.358 3.377 make_preconditioner 11 8.9 0.000 0.000 3.358 3.377 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.242 3.282 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.212 3.280 pw_transfer 1331 11.6 0.056 0.069 3.176 3.266 multiply_cannon_metrocomm1 49320 15.4 0.068 0.070 2.321 3.215 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.161 3.213 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.087 3.176 potential_pw2rs 110 12.3 0.006 0.006 2.701 2.732 fft_wrap_pw1pw2_140 451 13.1 0.388 0.436 2.613 2.719 calculate_first_density_matrix 1 7.0 0.000 0.000 2.664 2.667 acc_transpose_blocks 49320 15.4 0.217 0.227 2.576 2.644 jit_kernel_multiply 13 15.9 2.554 2.567 2.554 2.567 mp_alltoall_d11v 2046 13.8 2.082 2.513 2.082 2.513 fft3d_ps 1111 14.6 0.793 0.886 2.383 2.468 grid_collocate_task_list 110 9.6 2.154 2.279 2.154 2.279 transfer_rs2pw 451 10.6 0.005 0.006 1.997 2.187 mp_waitany 14300 13.8 1.820 2.111 1.820 2.111 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.945 make_images_data 4110 15.4 0.044 0.047 1.791 1.929 mp_sum_d 3889 11.9 1.352 1.915 1.352 1.915 transfer_pw2rs 451 13.1 0.006 0.007 1.840 1.857 cp_fm_cholesky_invert 11 10.9 1.831 1.835 1.831 1.835 hybrid_alltoall_any 4261 16.3 0.085 0.487 1.550 1.807 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.659 1.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.522000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.766386 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 592.863232E+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 3309342. 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 0.028 0.058 69.817 69.819 qs_mol_dyn_low 1 2.0 0.003 0.004 69.333 69.344 qs_forces 11 3.9 0.004 0.012 69.244 69.248 qs_energies 11 4.9 0.001 0.002 65.919 65.927 scf_env_do_scf 11 5.9 0.000 0.001 57.184 57.187 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 49.724 49.724 dbcsr_multiply_generic 2055 12.4 0.115 0.118 38.684 38.894 velocity_verlet 10 3.0 0.001 0.002 35.892 35.893 qs_scf_new_mos 99 7.5 0.001 0.001 33.850 33.975 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.850 33.975 multiply_cannon 2055 13.4 0.225 0.246 31.641 33.103 ot_scf_mini 99 9.5 0.003 0.003 32.169 32.283 multiply_cannon_loop 2055 14.4 1.167 1.199 30.340 31.606 ot_mini 99 10.5 0.001 0.001 18.992 19.120 multiply_cannon_multrec 24660 15.4 6.989 8.556 14.043 15.673 rebuild_ks_matrix 110 8.3 0.000 0.001 13.550 13.698 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.549 13.697 qs_ot_get_derivative 99 11.5 0.001 0.001 13.156 13.275 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.918 12.042 mp_waitall_1 176588 16.5 7.719 10.865 7.719 10.865 multiply_cannon_metrocomm3 24660 15.4 0.072 0.075 5.326 8.204 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.840 7.768 apply_single 110 13.6 0.000 0.001 6.839 7.768 multiply_cannon_sync_h2d 24660 15.4 6.366 7.494 6.366 7.494 init_scf_loop 11 6.9 0.000 0.000 7.426 7.426 dbcsr_mm_accdrv_process 52282 16.1 5.611 6.420 6.883 7.207 qs_ot_get_p 110 10.4 0.001 0.001 6.525 6.657 sum_up_and_integrate 110 10.3 0.053 0.059 6.379 6.391 integrate_v_rspace 110 11.3 0.002 0.003 6.326 6.342 init_scf_run 11 5.9 0.000 0.001 6.332 6.332 scf_env_initial_rho_setup 11 6.9 0.001 0.001 6.331 6.332 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.982 5.951 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.891 5.897 calculate_rho_elec 110 8.6 0.039 0.047 5.890 5.896 ot_diis_step 99 11.5 0.010 0.010 5.787 5.787 prepare_preconditioner 11 7.9 0.000 0.000 5.423 5.444 make_preconditioner 11 8.9 0.000 0.000 5.423 5.444 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.015 5.171 make_m2s 4110 13.4 0.056 0.061 4.328 4.795 make_images 4110 14.4 0.405 0.468 4.221 4.687 qs_ot_p2m_diag 48 11.0 0.029 0.045 4.628 4.649 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.146 4.147 pw_transfer 1331 11.6 0.066 0.075 3.542 3.682 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.578 3.590 cp_fm_diag_elpa_base 48 14.0 3.531 3.557 3.575 3.588 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.435 3.579 wfi_extrapolate 11 7.9 0.001 0.001 3.496 3.496 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.366 3.368 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.284 3.353 density_rs2pw 110 9.6 0.004 0.005 3.132 3.329 grid_integrate_task_list 110 12.3 3.168 3.304 3.168 3.304 fft_wrap_pw1pw2_140 451 13.1 0.459 0.477 2.950 3.095 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.033 3.080 mp_sum_l 6594 12.7 2.101 3.066 2.101 3.066 calculate_dm_sparse 110 9.5 0.001 0.001 2.984 3.011 make_images_data 4110 15.4 0.049 0.053 2.312 2.797 hybrid_alltoall_any 4261 16.3 0.107 0.452 2.026 2.782 calculate_first_density_matrix 1 7.0 0.000 0.000 2.738 2.740 fft3d_ps 1111 14.6 1.115 1.338 2.533 2.692 acc_transpose_blocks 24660 15.4 0.114 0.118 2.078 2.596 cp_fm_cholesky_invert 11 10.9 2.531 2.537 2.531 2.537 grid_collocate_task_list 110 9.6 2.172 2.318 2.172 2.318 potential_pw2rs 110 12.3 0.008 0.009 2.226 2.244 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.969 1.991 mp_alltoall_d11v 2046 13.8 1.730 1.956 1.730 1.956 qs_energies_init_hamiltonians 11 5.9 0.002 0.017 1.897 1.902 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.787 1.808 multiply_cannon_metrocomm4 22605 15.4 0.078 0.082 0.776 1.768 mp_allgather_i34 2055 14.4 0.606 1.659 0.606 1.659 transfer_rs2pw 451 10.6 0.006 0.007 1.453 1.620 jit_kernel_multiply 8 16.4 0.908 1.609 0.908 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.538 1.549 mp_irecv_dv 57340 16.2 0.648 1.544 0.648 1.544 acc_transpose_blocks_kernels 24660 16.4 0.322 0.331 1.011 1.520 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.388 1.492 mp_waitany 10164 13.8 1.197 1.401 1.197 1.401 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.819000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.545455, yerr=6.959149 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 668.221440E+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 0.019 0.030 60.659 60.661 qs_mol_dyn_low 1 2.0 0.003 0.003 60.369 60.378 qs_forces 11 3.9 0.003 0.003 60.283 60.288 qs_energies 11 4.9 0.001 0.002 57.108 57.115 scf_env_do_scf 11 5.9 0.000 0.001 49.122 49.122 scf_env_do_scf_inner_loop 99 6.5 0.007 0.027 40.674 40.675 velocity_verlet 10 3.0 0.005 0.008 32.448 32.450 dbcsr_multiply_generic 2055 12.4 0.106 0.110 29.572 29.858 qs_scf_new_mos 99 7.5 0.001 0.001 26.016 26.158 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.015 26.158 ot_scf_mini 99 9.5 0.002 0.002 24.766 24.925 multiply_cannon 2055 13.4 0.214 0.225 23.316 24.732 multiply_cannon_loop 2055 14.4 0.811 0.839 22.102 23.221 ot_mini 99 10.5 0.001 0.001 14.429 14.589 rebuild_ks_matrix 110 8.3 0.000 0.000 12.192 12.357 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.191 12.356 multiply_cannon_multrec 16440 15.4 3.502 4.274 10.166 11.021 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.732 10.882 mp_waitall_1 139946 16.5 7.386 10.357 7.386 10.357 qs_ot_get_derivative 99 11.5 0.001 0.001 9.889 10.047 init_scf_loop 11 6.9 0.000 0.000 8.414 8.415 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.783 7.731 dbcsr_mm_accdrv_process 34862 16.1 5.779 6.322 6.508 6.734 prepare_preconditioner 11 7.9 0.000 0.000 6.660 6.675 make_preconditioner 11 8.9 0.000 0.000 6.660 6.675 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.967 6.309 sum_up_and_integrate 110 10.3 0.061 0.061 6.279 6.296 integrate_v_rspace 110 11.3 0.003 0.003 6.218 6.235 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.704 5.717 calculate_rho_elec 110 8.6 0.058 0.059 5.703 5.717 qs_ot_get_p 110 10.4 0.001 0.001 5.465 5.628 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.988 5.528 apply_single 110 13.6 0.000 0.000 4.988 5.528 init_scf_run 11 5.9 0.000 0.001 5.525 5.526 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.525 5.525 make_m2s 4110 13.4 0.049 0.052 4.189 4.601 ot_diis_step 99 11.5 0.011 0.011 4.511 4.511 make_images 4110 14.4 0.397 0.526 4.075 4.487 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.231 3.835 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.821 3.825 multiply_cannon_sync_h2d 16440 15.4 3.225 3.791 3.225 3.791 pw_transfer 1331 11.6 0.066 0.072 3.539 3.548 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.452 3.453 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.431 3.443 grid_integrate_task_list 110 12.3 3.215 3.373 3.215 3.373 density_rs2pw 110 9.6 0.004 0.005 2.942 3.090 fft_wrap_pw1pw2_140 451 13.1 0.580 0.586 2.979 2.991 wfi_extrapolate 11 7.9 0.001 0.001 2.970 2.970 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.909 2.917 cp_fm_diag_elpa_base 48 14.0 2.842 2.873 2.906 2.914 make_images_data 4110 15.4 0.045 0.050 2.376 2.877 hybrid_alltoall_any 4261 16.3 0.109 0.381 2.133 2.850 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.843 2.844 cp_fm_cholesky_invert 11 10.9 2.574 2.580 2.574 2.580 calculate_dm_sparse 110 9.5 0.001 0.001 2.534 2.566 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.485 2.536 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.443 2.535 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.921 2.528 calculate_first_density_matrix 1 7.0 0.000 0.000 2.467 2.467 grid_collocate_task_list 110 9.6 2.223 2.431 2.223 2.431 mp_irecv_dv 48980 15.7 0.846 2.387 0.846 2.387 fft3d_ps 1111 14.6 1.091 1.112 2.371 2.383 mp_sum_l 6594 12.7 1.455 2.119 1.455 2.119 potential_pw2rs 110 12.3 0.011 0.011 2.091 2.096 mp_alltoall_d11v 2046 13.8 1.778 2.037 1.778 2.037 qs_energies_init_hamiltonians 11 5.9 0.064 0.075 2.015 2.021 dbcsr_complete_redistribute 325 12.2 0.350 0.387 1.522 1.944 cp_fm_upper_to_full 70 14.2 1.352 1.719 1.352 1.719 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.651 1.670 acc_transpose_blocks 16440 15.4 0.077 0.080 1.577 1.662 mp_allgather_i34 2055 14.4 0.524 1.630 0.524 1.630 cp_fm_cholesky_decompose 22 10.9 1.550 1.571 1.550 1.571 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.470 1.484 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.368 1.469 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.024 1.454 transfer_rs2pw 451 10.6 0.005 0.006 1.250 1.398 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.283 1.288 mp_waitany 17072 13.8 1.117 1.279 1.117 1.279 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.661000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=633.181818, yerr=7.917968 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 737.480704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.038 66.003 66.005 qs_mol_dyn_low 1 2.0 0.003 0.003 65.648 65.659 qs_forces 11 3.9 0.100 0.112 65.510 65.512 qs_energies 11 4.9 0.002 0.009 62.035 62.058 scf_env_do_scf 11 5.9 0.001 0.001 53.176 53.178 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.710 41.710 velocity_verlet 10 3.0 0.001 0.002 36.855 36.860 dbcsr_multiply_generic 2055 12.4 0.113 0.121 30.871 31.065 qs_scf_new_mos 99 7.5 0.001 0.001 27.304 27.393 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.304 27.392 ot_scf_mini 99 9.5 0.002 0.003 25.617 25.704 multiply_cannon 2055 13.4 0.239 0.254 23.954 24.910 multiply_cannon_loop 2055 14.4 1.425 1.496 22.629 23.275 ot_mini 99 10.5 0.001 0.001 14.762 14.866 multiply_cannon_multrec 24660 15.4 4.110 6.777 13.402 14.516 rebuild_ks_matrix 110 8.3 0.000 0.000 11.850 11.951 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.018 11.849 11.950 init_scf_loop 11 6.9 0.001 0.003 11.426 11.427 qs_ot_get_derivative 99 11.5 0.001 0.001 10.583 10.670 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.466 10.553 dbcsr_mm_accdrv_process 52304 16.0 8.106 9.436 9.133 10.127 prepare_preconditioner 11 7.9 0.000 0.000 9.677 9.693 make_preconditioner 11 8.9 0.000 0.001 9.677 9.693 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.917 9.381 sum_up_and_integrate 110 10.3 0.068 0.072 6.204 6.217 integrate_v_rspace 110 11.3 0.003 0.003 6.136 6.151 init_scf_run 11 5.9 0.000 0.002 6.032 6.033 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.032 6.032 qs_ot_get_p 110 10.4 0.001 0.001 5.793 5.928 mp_waitall_1 121746 16.5 4.122 5.824 4.122 5.824 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.746 5.754 calculate_rho_elec 110 8.6 0.078 0.081 5.745 5.753 make_m2s 4110 13.4 0.059 0.061 5.338 5.613 make_images 4110 14.4 0.577 0.696 5.198 5.467 cp_fm_upper_to_full 70 14.2 3.325 4.833 3.325 4.833 ot_diis_step 99 11.5 0.011 0.012 4.142 4.142 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.060 4.138 apply_single 110 13.6 0.000 0.000 4.060 4.138 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.955 3.972 dbcsr_complete_redistribute 325 12.2 0.422 0.466 2.625 3.749 pw_transfer 1331 11.6 0.065 0.074 3.610 3.641 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.503 3.539 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.489 3.535 calculate_dm_sparse 110 9.5 0.001 0.001 3.409 3.441 grid_integrate_task_list 110 12.3 3.302 3.427 3.302 3.427 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.348 3.396 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.391 3.392 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.101 3.218 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.363 3.087 fft_wrap_pw1pw2_140 451 13.1 0.608 0.630 3.024 3.061 density_rs2pw 110 9.6 0.004 0.004 2.874 2.996 calculate_first_density_matrix 1 7.0 0.001 0.009 2.977 2.982 wfi_extrapolate 11 7.9 0.001 0.001 2.945 2.945 make_images_data 4110 15.4 0.049 0.053 2.601 2.890 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.861 2.872 cp_fm_diag_elpa_base 48 14.0 2.714 2.773 2.858 2.870 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.751 2.856 hybrid_alltoall_any 4261 16.3 0.123 0.464 2.276 2.840 mp_alltoall_i22 605 13.7 1.647 2.823 1.647 2.823 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.790 2.792 cp_fm_cholesky_invert 11 10.9 2.577 2.585 2.577 2.585 acc_transpose_blocks 24660 15.4 0.114 0.118 2.468 2.584 multiply_cannon_sync_h2d 24660 15.4 2.393 2.481 2.393 2.481 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.443 2.474 fft3d_ps 1111 14.6 1.096 1.128 2.388 2.403 grid_collocate_task_list 110 9.6 2.269 2.372 2.269 2.372 qs_energies_init_hamiltonians 11 5.9 0.030 0.043 2.329 2.352 potential_pw2rs 110 12.3 0.013 0.013 1.989 1.997 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.965 1.980 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.796 1.834 mp_alltoall_d11v 2046 13.8 1.695 1.813 1.695 1.813 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.003 1.615 1.713 cp_fm_cholesky_decompose 22 10.9 1.602 1.647 1.602 1.647 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.625 1.635 acc_transpose_blocks_sync 73980 16.4 1.413 1.525 1.413 1.525 mp_sum_l 6594 12.7 0.899 1.475 0.899 1.475 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.843 1.464 mp_allgather_i34 2055 14.4 0.410 1.409 0.410 1.409 mp_irecv_dv 62702 16.1 0.738 1.383 0.738 1.383 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.229 1.355 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.005000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.545455, yerr=10.272325 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 846.692352E+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 0.019 0.032 54.772 54.773 qs_mol_dyn_low 1 2.0 0.003 0.003 54.479 54.503 qs_forces 11 3.9 0.003 0.005 54.404 54.406 qs_energies 11 4.9 0.001 0.002 50.758 50.764 scf_env_do_scf 11 5.9 0.000 0.001 42.462 42.463 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.833 34.834 velocity_verlet 10 3.0 0.002 0.008 30.948 30.951 dbcsr_multiply_generic 2055 12.4 0.106 0.110 23.053 23.221 qs_scf_new_mos 99 7.5 0.001 0.001 20.432 20.514 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.431 20.514 ot_scf_mini 99 9.5 0.002 0.002 19.191 19.255 multiply_cannon 2055 13.4 0.241 0.250 17.617 18.785 multiply_cannon_loop 2055 14.4 0.610 0.628 16.403 16.737 rebuild_ks_matrix 110 8.3 0.000 0.000 11.529 11.605 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.528 11.605 ot_mini 99 10.5 0.001 0.001 10.612 10.681 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.254 10.321 multiply_cannon_multrec 8220 15.4 3.137 4.760 7.558 8.820 init_scf_loop 11 6.9 0.000 0.000 7.583 7.584 mp_waitall_1 103326 16.6 5.662 7.294 5.662 7.294 qs_ot_get_derivative 99 11.5 0.001 0.001 6.830 6.897 sum_up_and_integrate 110 10.3 0.080 0.082 6.177 6.189 integrate_v_rspace 110 11.3 0.004 0.004 6.097 6.109 prepare_preconditioner 11 7.9 0.000 0.000 5.942 5.947 make_preconditioner 11 8.9 0.000 0.000 5.942 5.947 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.898 5.912 calculate_rho_elec 110 8.6 0.114 0.114 5.897 5.911 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.517 5.596 dbcsr_mm_accdrv_process 17442 15.9 3.076 4.138 4.281 5.186 init_scf_run 11 5.9 0.000 0.001 5.167 5.167 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.167 5.167 qs_ot_get_p 110 10.4 0.001 0.001 4.661 4.730 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.047 4.516 make_m2s 4110 13.4 0.038 0.039 4.109 4.298 make_images 4110 14.4 0.647 0.704 3.979 4.166 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.810 3.886 apply_single 110 13.6 0.000 0.000 3.809 3.885 pw_transfer 1331 11.6 0.066 0.073 3.789 3.807 ot_diis_step 99 11.5 0.012 0.012 3.760 3.760 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.681 3.703 grid_integrate_task_list 110 12.3 3.372 3.528 3.372 3.528 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.318 3.322 fft_wrap_pw1pw2_140 451 13.1 0.775 0.786 3.200 3.230 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.983 2.984 density_rs2pw 110 9.6 0.004 0.004 2.867 2.974 cp_fm_cholesky_invert 11 10.9 2.843 2.847 2.843 2.847 wfi_extrapolate 11 7.9 0.001 0.001 2.686 2.686 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 2.676 2.678 hybrid_alltoall_any 4261 16.3 0.201 0.853 2.214 2.595 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.567 2.568 calculate_dm_sparse 110 9.5 0.001 0.001 2.477 2.516 multiply_cannon_sync_h2d 8220 15.4 2.349 2.501 2.349 2.501 grid_collocate_task_list 110 9.6 2.367 2.450 2.367 2.450 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.419 2.425 cp_fm_diag_elpa_base 48 14.0 2.363 2.384 2.417 2.424 make_images_data 4110 15.4 0.040 0.046 2.169 2.420 calculate_first_density_matrix 1 7.0 0.000 0.000 2.387 2.388 fft3d_ps 1111 14.6 1.140 1.174 2.348 2.365 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.168 2.187 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.102 2.148 mp_alltoall_d11v 2046 13.8 1.754 2.026 1.754 2.026 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.785 2.011 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.905 1.939 potential_pw2rs 110 12.3 0.016 0.016 1.926 1.928 cp_fm_cholesky_decompose 22 10.9 1.649 1.669 1.649 1.669 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.502 1.621 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.602 1.611 dbcsr_complete_redistribute 325 12.2 0.585 0.635 1.457 1.554 acc_transpose_blocks 8220 15.4 0.039 0.040 1.503 1.534 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.440 1.451 mp_allgather_i34 2055 14.4 0.380 1.400 0.380 1.400 multiply_cannon_metrocomm4 6165 15.4 0.019 0.020 0.481 1.398 mp_irecv_dv 24056 15.7 0.454 1.351 0.454 1.351 qs_create_task_list 11 7.9 0.001 0.002 1.211 1.313 generate_qs_task_list 11 8.9 0.374 0.443 1.210 1.312 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.211 1.248 transfer_rs2pw 451 10.6 0.005 0.005 1.098 1.227 jit_kernel_multiply 7 15.6 0.894 1.216 0.894 1.216 mp_waitany 9240 13.8 1.037 1.166 1.037 1.166 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.773000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=800.090909, yerr=11.049102 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 1.376182E+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 0.023 0.039 86.970 86.971 qs_mol_dyn_low 1 2.0 0.003 0.003 86.407 86.642 qs_forces 11 3.9 0.004 0.004 86.274 86.287 qs_energies 11 4.9 0.002 0.004 82.197 82.211 scf_env_do_scf 11 5.9 0.001 0.001 72.177 72.177 velocity_verlet 10 3.0 0.002 0.004 55.405 55.413 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.918 43.920 dbcsr_multiply_generic 2055 12.4 0.121 0.126 29.457 29.560 init_scf_loop 11 6.9 0.000 0.000 28.188 28.191 qs_scf_new_mos 99 7.5 0.001 0.001 26.725 26.779 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.724 26.778 prepare_preconditioner 11 7.9 0.000 0.000 26.181 26.190 make_preconditioner 11 8.9 0.000 0.000 26.181 26.190 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.328 25.661 ot_scf_mini 99 9.5 0.002 0.002 24.921 24.968 multiply_cannon 2055 13.4 0.337 0.376 22.485 23.118 multiply_cannon_loop 2055 14.4 0.809 0.821 20.726 21.010 cp_fm_upper_to_full 70 14.2 12.762 18.342 12.762 18.342 ot_mini 99 10.5 0.001 0.001 14.090 14.138 rebuild_ks_matrix 110 8.3 0.001 0.001 13.613 13.665 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.612 13.665 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.347 12.393 dbcsr_complete_redistribute 325 12.2 1.022 1.042 7.391 10.747 multiply_cannon_multrec 8220 15.4 4.069 4.212 9.778 9.900 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.397 9.759 qs_ot_get_derivative 99 11.5 0.001 0.001 9.467 9.517 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.837 9.157 mp_alltoall_i22 605 13.7 5.465 8.857 5.465 8.857 mp_waitall_1 84994 16.7 7.432 8.287 7.432 8.287 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.000 7.039 calculate_rho_elec 110 8.6 0.225 0.226 6.999 7.038 sum_up_and_integrate 110 10.3 0.150 0.151 6.803 6.824 integrate_v_rspace 110 11.3 0.004 0.004 6.653 6.673 dbcsr_mm_accdrv_process 11614 15.7 3.784 4.157 5.560 5.786 init_scf_run 11 5.9 0.000 0.001 5.746 5.747 scf_env_initial_rho_setup 11 6.9 0.001 0.002 5.746 5.746 make_m2s 4110 13.4 0.043 0.044 5.225 5.699 qs_ot_get_p 110 10.4 0.001 0.001 5.533 5.592 make_images 4110 14.4 0.885 0.939 5.036 5.511 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.765 5.251 apply_single 110 13.6 0.000 0.000 4.764 5.251 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.917 5.221 cp_fm_cholesky_invert 11 10.9 5.134 5.138 5.134 5.138 pw_transfer 1331 11.6 0.075 0.075 4.872 4.886 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.754 4.769 ot_diis_step 99 11.5 0.015 0.015 4.603 4.603 fft_wrap_pw1pw2_140 451 13.1 1.280 1.286 4.162 4.176 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.971 3.979 grid_integrate_task_list 110 12.3 3.678 3.745 3.678 3.745 qs_energies_init_hamiltonians 11 5.9 0.017 0.024 3.663 3.676 density_rs2pw 110 9.6 0.004 0.004 3.458 3.474 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.994 3.463 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.457 3.458 hybrid_alltoall_any 4261 16.3 0.264 0.570 2.788 3.407 make_images_data 4110 15.4 0.045 0.048 2.687 3.295 wfi_extrapolate 11 7.9 0.001 0.001 3.288 3.289 calculate_dm_sparse 110 9.5 0.001 0.001 3.187 3.210 multiply_cannon_sync_h2d 8220 15.4 3.151 3.180 3.151 3.180 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.901 2.904 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.878 2.878 cp_fm_diag_elpa_base 48 14.0 2.335 2.534 2.876 2.876 fft3d_ps 1111 14.6 1.300 1.306 2.806 2.819 grid_collocate_task_list 110 9.6 2.670 2.682 2.670 2.682 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.633 2.663 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.398 2.417 calculate_first_density_matrix 1 7.0 0.000 0.000 2.344 2.345 potential_pw2rs 110 12.3 0.022 0.022 2.223 2.237 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.173 2.223 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.088 2.161 mp_alltoall_d11v 2046 13.8 1.999 2.080 1.999 2.080 cp_fm_cholesky_decompose 22 10.9 1.936 1.959 1.936 1.959 qs_create_task_list 11 7.9 0.001 0.001 1.879 1.926 generate_qs_task_list 11 8.9 0.730 0.781 1.878 1.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.865 1.871 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.733 1.776 jit_kernel_multiply 10 15.3 1.582 1.753 1.582 1.753 acc_transpose_blocks 8220 15.4 0.040 0.041 1.713 1.740 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.971000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1243.909091, yerr=57.631201 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 1.086553E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 11851392 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.8 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 630.669312E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10258080 MPI messages size (bytes): total size 4.456715E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.459031E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3542056 94711185408 32768 < size <= 131072 1282176 73356279808 131072 < size <= 4194304 5107038 3151762421624 4194304 < size <= 16777216 259842 1136842803272 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 57774. MP_Allreduce 11082 797. MP_Sync 86 MP_Alltoall 2210 3298361. MP_SendRecv 24130 18752. MP_ISendRecv 24130 18752. MP_Wait 42150 MP_ISend 15900 108037. MP_IRecv 15900 108037. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.028 205.722 205.723 qs_mol_dyn_low 1 2.0 0.003 0.007 205.229 205.277 qs_forces 11 3.9 0.005 0.006 204.157 204.163 qs_energies 11 4.9 0.001 0.002 198.556 198.587 scf_env_do_scf 11 5.9 0.001 0.001 182.002 182.006 scf_env_do_scf_inner_loop 116 6.6 0.002 0.007 161.519 161.520 dbcsr_multiply_generic 2485 12.5 0.187 0.192 124.398 125.519 qs_scf_new_mos 116 7.6 0.001 0.001 122.826 123.109 qs_scf_loop_do_ot 116 8.6 0.001 0.001 122.826 123.109 velocity_verlet 10 3.0 0.001 0.001 122.793 122.804 ot_scf_mini 116 9.6 0.003 0.003 116.174 116.397 multiply_cannon 2485 13.5 0.241 0.255 100.563 103.048 multiply_cannon_loop 2485 14.5 2.386 2.441 98.405 101.061 ot_mini 116 10.6 0.001 0.001 65.704 65.967 multiply_cannon_multrec 59640 15.5 31.475 33.545 41.114 42.854 qs_ot_get_derivative 116 11.6 0.001 0.001 41.130 41.349 rebuild_ks_matrix 127 8.3 0.001 0.001 33.072 33.459 qs_ks_build_kohn_sham_matrix 127 9.3 0.015 0.017 33.071 33.458 mp_waitall_1 264810 16.5 28.421 31.093 28.421 31.093 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.664 30.013 qs_ot_get_p 127 10.4 0.001 0.001 28.462 28.718 multiply_cannon_sync_h2d 59640 15.5 26.059 28.314 26.059 28.314 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 24.062 24.708 apply_single 127 13.6 0.001 0.001 24.062 24.708 ot_diis_step 116 11.6 0.008 0.008 24.334 24.335 qs_ot_p2m_diag 82 11.4 0.078 0.090 21.942 22.008 init_scf_loop 11 6.9 0.000 0.000 20.411 20.412 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 19.747 19.914 cp_dbcsr_syevd 82 12.4 0.005 0.005 19.281 19.282 multiply_cannon_metrocomm3 59640 15.5 0.119 0.123 15.931 18.350 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.417 16.450 cp_fm_diag_elpa_base 82 14.4 16.352 16.384 16.413 16.447 prepare_preconditioner 11 7.9 0.000 0.000 15.880 15.920 make_preconditioner 11 8.9 0.000 0.000 15.880 15.920 make_full_inverse_cholesky 11 9.9 0.000 0.000 15.130 15.310 make_m2s 4970 13.5 0.104 0.111 13.579 13.917 sum_up_and_integrate 127 10.3 0.089 0.106 13.754 13.768 make_images 4970 14.5 0.394 0.413 13.398 13.748 integrate_v_rspace 127 11.3 0.004 0.004 13.665 13.683 qs_rho_update_rho_low 127 7.7 0.001 0.001 12.870 13.028 calculate_rho_elec 127 8.7 0.045 0.063 12.869 13.027 init_scf_run 11 5.9 0.000 0.001 12.396 12.397 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.396 12.397 mp_sum_l 7884 12.9 8.996 10.584 8.996 10.584 dbcsr_mm_accdrv_process 123452 16.2 4.625 4.787 9.215 9.797 cp_fm_cholesky_invert 11 10.9 9.106 9.114 9.106 9.114 wfi_extrapolate 11 7.9 0.001 0.001 9.048 9.048 calculate_dm_sparse 127 9.5 0.001 0.001 8.469 8.607 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.132 8.301 qs_ot_get_orbitals 116 10.6 0.001 0.001 8.046 8.134 multiply_cannon_metrocomm1 59640 15.5 0.097 0.101 6.186 7.777 pw_transfer 1535 11.6 0.073 0.092 7.444 7.621 make_images_data 4970 15.5 0.067 0.074 6.671 7.574 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 7.243 7.413 grid_integrate_task_list 127 12.3 7.018 7.400 7.018 7.400 hybrid_alltoall_any 5155 16.4 0.293 2.242 5.849 7.157 density_rs2pw 127 9.7 0.006 0.007 6.539 7.120 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.752 6.760 fft_wrap_pw1pw2_140 519 13.2 1.130 1.176 6.398 6.563 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.815 5.979 fft3d_ps 1281 14.7 2.179 2.837 5.250 5.580 mp_alltoall_d11v 2401 14.1 4.211 5.270 4.211 5.270 grid_collocate_task_list 127 9.7 4.801 5.231 4.801 5.231 cp_fm_cholesky_decompose 22 10.9 4.596 4.611 4.596 4.611 potential_pw2rs 127 12.3 0.009 0.011 4.343 4.360 mp_sum_d 4451 12.1 3.454 4.245 3.454 4.245 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.723000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.272727, yerr=5.986210 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 848.039936E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3999 58256. MP_Allreduce 11076 960. MP_Sync 86 MP_Alltoall 1955 5064256. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.078 0.178 191.495 191.497 qs_mol_dyn_low 1 2.0 0.003 0.004 189.706 190.816 qs_forces 11 3.9 0.005 0.007 189.397 189.402 qs_energies 11 4.9 0.007 0.026 182.647 182.659 scf_env_do_scf 11 5.9 0.001 0.002 165.988 166.000 scf_env_do_scf_inner_loop 116 6.6 0.004 0.008 133.041 133.044 velocity_verlet 10 3.0 0.001 0.002 118.557 118.562 dbcsr_multiply_generic 2485 12.5 0.188 0.194 96.414 97.723 qs_scf_new_mos 116 7.6 0.001 0.001 94.716 95.278 qs_scf_loop_do_ot 116 8.6 0.001 0.001 94.715 95.277 ot_scf_mini 116 9.6 0.004 0.005 89.948 90.579 multiply_cannon 2485 13.5 0.502 0.572 76.375 80.999 multiply_cannon_loop 2485 14.5 1.563 1.631 73.403 76.215 ot_mini 116 10.6 0.001 0.001 49.913 50.472 mp_waitall_1 212858 16.6 23.975 38.003 23.975 38.003 multiply_cannon_multrec 29820 15.5 20.828 25.204 31.335 36.186 init_scf_loop 11 6.9 0.002 0.008 32.858 32.861 rebuild_ks_matrix 127 8.3 0.001 0.001 32.069 32.592 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 32.069 32.592 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.825 29.302 qs_ot_get_derivative 116 11.6 0.001 0.002 27.982 28.600 prepare_preconditioner 11 7.9 0.000 0.000 28.534 28.593 make_preconditioner 11 8.9 0.000 0.001 28.533 28.593 multiply_cannon_metrocomm3 29820 15.5 0.097 0.103 15.611 28.457 make_full_inverse_cholesky 11 9.9 0.000 0.002 27.190 27.702 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.016 23.221 apply_single 127 13.6 0.001 0.001 22.016 23.221 qs_ot_get_p 127 10.4 0.001 0.001 22.317 22.931 ot_diis_step 116 11.6 0.014 0.015 21.755 21.757 multiply_cannon_sync_h2d 29820 15.5 17.825 19.805 17.825 19.805 qs_ot_p2m_diag 82 11.4 0.187 0.215 17.442 17.476 cp_fm_cholesky_invert 11 10.9 16.676 16.688 16.676 16.688 cp_dbcsr_syevd 82 12.4 0.005 0.006 16.226 16.227 make_m2s 4970 13.5 0.089 0.094 14.084 15.340 make_images 4970 14.5 1.156 1.354 13.876 15.132 sum_up_and_integrate 127 10.3 0.116 0.134 14.048 14.079 integrate_v_rspace 127 11.3 0.004 0.004 13.932 13.968 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.366 13.392 calculate_rho_elec 127 8.7 0.086 0.103 13.365 13.391 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.035 13.065 cp_fm_diag_elpa_base 82 14.4 12.781 12.882 13.029 13.055 init_scf_run 11 5.9 0.000 0.001 11.723 11.724 scf_env_initial_rho_setup 11 6.9 0.001 0.004 11.722 11.724 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.035 11.474 multiply_cannon_metrocomm4 27335 15.5 0.104 0.119 3.861 10.927 mp_irecv_dv 68888 16.3 3.657 10.524 3.657 10.524 dbcsr_mm_accdrv_process 61726 16.2 5.377 6.049 9.954 10.523 make_images_data 4970 15.5 0.067 0.076 8.296 9.807 hybrid_alltoall_any 5155 16.4 0.350 1.550 7.058 9.094 pw_transfer 1535 11.6 0.084 0.099 8.574 8.628 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.352 8.406 wfi_extrapolate 11 7.9 0.001 0.001 8.350 8.350 density_rs2pw 127 9.7 0.006 0.007 7.117 7.543 grid_integrate_task_list 127 12.3 7.095 7.468 7.095 7.468 fft_wrap_pw1pw2_140 519 13.2 1.216 1.244 7.382 7.451 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.246 7.005 cp_fm_cholesky_decompose 22 10.9 6.923 7.001 6.923 7.001 calculate_dm_sparse 127 9.5 0.001 0.001 6.503 6.667 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.158 6.168 fft3d_ps 1281 14.7 2.821 2.972 5.916 5.950 mp_sum_l 7884 12.9 4.171 5.909 4.171 5.909 grid_collocate_task_list 127 9.7 4.964 5.501 4.964 5.501 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.350 5.424 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.228 5.402 mp_alltoall_d11v 2401 14.1 4.214 5.008 4.214 5.008 potential_pw2rs 127 12.3 0.015 0.017 4.546 4.575 mp_allgather_i34 2485 14.5 1.496 4.557 1.496 4.557 dbcsr_complete_redistribute 393 12.7 0.791 0.917 3.282 4.143 mp_sum_d 4450 12.1 2.716 4.015 2.716 4.015 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.497000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=806.363636, yerr=5.788275 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 940.937216E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172769472 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4007 58135. MP_Allreduce 11095 999. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.041 0.098 175.228 175.230 qs_mol_dyn_low 1 2.0 0.003 0.003 174.548 174.562 qs_forces 11 3.9 0.016 0.025 174.443 174.449 qs_energies 11 4.9 0.006 0.026 167.900 167.916 scf_env_do_scf 11 5.9 0.002 0.008 152.250 152.252 scf_env_do_scf_inner_loop 116 6.6 0.019 0.032 117.287 117.289 velocity_verlet 10 3.0 0.002 0.005 111.379 111.381 dbcsr_multiply_generic 2485 12.5 0.178 0.184 81.557 82.683 qs_scf_new_mos 116 7.6 0.001 0.001 80.527 80.860 qs_scf_loop_do_ot 116 8.6 0.001 0.001 80.526 80.859 ot_scf_mini 116 9.6 0.017 0.060 76.328 76.673 multiply_cannon 2485 13.5 0.499 0.517 61.848 66.372 multiply_cannon_loop 2485 14.5 1.116 1.189 59.172 62.220 ot_mini 116 10.6 0.001 0.001 42.533 42.935 init_scf_loop 11 6.9 0.001 0.005 34.859 34.860 mp_waitall_1 169034 16.6 24.840 33.864 24.840 33.864 prepare_preconditioner 11 7.9 0.000 0.000 30.827 30.873 make_preconditioner 11 8.9 0.002 0.011 30.827 30.873 rebuild_ks_matrix 127 8.3 0.001 0.001 29.978 30.514 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.017 29.977 30.513 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.494 29.882 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.012 27.502 multiply_cannon_multrec 19880 15.5 12.886 15.868 22.287 25.155 multiply_cannon_metrocomm3 19880 15.5 0.062 0.066 15.495 24.834 qs_ot_get_derivative 116 11.6 0.001 0.002 22.956 23.315 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.627 20.646 apply_single 127 13.6 0.001 0.001 19.626 20.645 qs_ot_get_p 127 10.4 0.001 0.001 19.254 19.713 ot_diis_step 116 11.6 0.018 0.020 19.465 19.465 multiply_cannon_sync_h2d 19880 15.5 13.578 15.508 13.578 15.508 make_m2s 4970 13.5 0.079 0.084 14.693 15.437 make_images 4970 14.5 1.137 1.231 14.463 15.207 qs_ot_p2m_diag 82 11.4 0.263 0.270 15.030 15.051 cp_fm_cholesky_invert 11 10.9 14.458 14.467 14.458 14.467 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.978 13.979 sum_up_and_integrate 127 10.3 0.132 0.144 13.917 13.941 integrate_v_rspace 127 11.3 0.004 0.004 13.784 13.810 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.449 13.479 calculate_rho_elec 127 8.7 0.129 0.144 13.448 13.478 cp_fm_diag_elpa 82 13.4 0.000 0.001 10.790 10.807 cp_fm_diag_elpa_base 82 14.4 10.385 10.530 10.787 10.803 init_scf_run 11 5.9 0.000 0.001 10.582 10.583 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.582 10.582 make_images_data 4970 15.5 0.062 0.071 8.832 9.978 multiply_cannon_metrocomm4 17395 15.5 0.066 0.075 3.518 9.403 hybrid_alltoall_any 5155 16.4 0.447 2.025 7.560 9.298 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 8.910 9.166 mp_irecv_dv 49801 16.2 3.390 9.140 3.390 9.140 dbcsr_mm_accdrv_process 41158 16.2 5.664 6.068 8.860 8.985 pw_transfer 1535 11.6 0.085 0.102 8.627 8.729 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.404 8.511 grid_integrate_task_list 127 12.3 7.301 7.721 7.301 7.721 fft_wrap_pw1pw2_140 519 13.2 1.303 1.353 7.457 7.570 wfi_extrapolate 11 7.9 0.001 0.001 7.457 7.457 cp_fm_cholesky_decompose 22 10.9 7.323 7.366 7.323 7.366 density_rs2pw 127 9.7 0.006 0.006 6.857 7.337 cp_fm_upper_to_full 104 14.8 5.632 7.137 5.632 7.137 dbcsr_complete_redistribute 393 12.7 1.199 1.265 4.649 6.366 calculate_dm_sparse 127 9.5 0.001 0.001 5.806 5.887 fft3d_ps 1281 14.7 2.745 2.959 5.789 5.857 grid_collocate_task_list 127 9.7 5.127 5.630 5.127 5.630 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.551 5.558 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.724 5.319 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.472 5.190 mp_alltoall_d11v 2401 14.1 4.230 5.011 4.230 5.011 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.577 4.746 mp_sum_l 7884 12.9 3.211 4.661 3.211 4.661 potential_pw2rs 127 12.3 0.020 0.023 4.345 4.363 mp_allgather_i34 2485 14.5 1.180 4.297 1.180 4.297 transfer_fm_to_dbcsr 11 9.9 0.019 0.023 2.310 4.017 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.962 4.001 qs_energies_init_hamiltonians 11 5.9 0.001 0.006 3.780 3.794 mp_alltoall_i22 712 14.1 1.924 3.775 1.924 3.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.462 3.509 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=175.230000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=890.909091, yerr=8.575383 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019187724288 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019187724288 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796577E+12 0.0% 0.0% 100.0% flops max/rank 4.320341E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500544 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.154404E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+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 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837747920 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3955 58883. MP_Allreduce 10965 1087. MP_Sync 86 MP_Alltoall 1700 12496304. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.037 0.082 189.955 189.960 qs_mol_dyn_low 1 2.0 0.003 0.004 189.136 189.152 qs_forces 11 3.9 0.004 0.005 188.992 189.003 qs_energies 11 4.9 0.005 0.013 181.869 181.878 scf_env_do_scf 11 5.9 0.003 0.015 164.285 164.299 velocity_verlet 10 3.0 0.002 0.002 124.501 124.505 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 117.697 117.698 qs_scf_new_mos 116 7.6 0.001 0.001 81.602 81.970 qs_scf_loop_do_ot 116 8.6 0.001 0.001 81.601 81.969 dbcsr_multiply_generic 2485 12.5 0.189 0.200 80.230 80.946 ot_scf_mini 116 9.6 0.003 0.004 77.068 77.411 multiply_cannon 2485 13.5 0.546 0.590 55.351 57.527 multiply_cannon_loop 2485 14.5 1.819 1.924 51.851 53.597 init_scf_loop 11 6.9 0.002 0.010 46.461 46.463 ot_mini 116 10.6 0.001 0.001 42.735 43.068 prepare_preconditioner 11 7.9 0.000 0.002 42.306 42.326 make_preconditioner 11 8.9 0.001 0.002 42.305 42.326 make_full_inverse_cholesky 11 9.9 0.014 0.023 35.952 40.984 multiply_cannon_multrec 29820 15.5 13.594 18.402 26.516 31.121 rebuild_ks_matrix 127 8.3 0.001 0.001 29.253 29.561 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 29.253 29.561 mp_waitall_1 146592 16.7 17.034 26.921 17.034 26.921 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.381 26.652 qs_ot_get_derivative 116 11.6 0.001 0.002 23.253 23.588 make_m2s 4970 13.5 0.096 0.100 20.477 21.189 make_images 4970 14.5 1.965 2.259 20.172 20.884 qs_ot_get_p 127 10.4 0.001 0.001 20.520 20.873 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.926 19.522 apply_single 127 13.6 0.001 0.001 18.925 19.522 ot_diis_step 116 11.6 0.017 0.018 19.351 19.353 cp_fm_upper_to_full 104 14.8 11.420 16.848 11.420 16.848 qs_ot_p2m_diag 82 11.4 0.339 0.385 16.325 16.377 cp_fm_cholesky_invert 11 10.9 16.192 16.200 16.192 16.200 multiply_cannon_metrocomm3 29820 15.5 0.049 0.053 6.181 15.251 cp_dbcsr_syevd 82 12.4 0.005 0.005 14.919 14.920 sum_up_and_integrate 127 10.3 0.139 0.151 14.068 14.094 integrate_v_rspace 127 11.3 0.004 0.004 13.928 13.959 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.665 13.715 calculate_rho_elec 127 8.7 0.171 0.187 13.664 13.714 dbcsr_mm_accdrv_process 61748 16.2 8.413 9.172 12.493 12.941 dbcsr_complete_redistribute 393 12.7 1.518 1.624 9.146 12.925 make_images_data 4970 15.5 0.066 0.073 10.967 12.591 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.659 11.672 cp_fm_diag_elpa_base 82 14.4 10.699 11.010 11.652 11.666 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.752 11.512 multiply_cannon_sync_h2d 29820 15.5 10.420 11.491 10.420 11.491 init_scf_run 11 5.9 0.000 0.001 11.451 11.452 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.450 11.452 hybrid_alltoall_any 5155 16.4 0.530 2.208 9.805 11.437 transfer_fm_to_dbcsr 11 9.9 0.002 0.008 6.330 9.980 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.701 9.960 mp_alltoall_i22 712 14.1 5.609 9.321 5.609 9.321 pw_transfer 1535 11.6 0.086 0.100 8.965 9.039 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.740 8.819 fft_wrap_pw1pw2_140 519 13.2 1.448 1.480 7.768 7.847 wfi_extrapolate 11 7.9 0.001 0.001 7.758 7.758 grid_integrate_task_list 127 12.3 7.470 7.742 7.470 7.742 cp_fm_cholesky_decompose 22 10.9 7.577 7.694 7.577 7.694 density_rs2pw 127 9.7 0.006 0.006 6.760 7.064 multiply_cannon_metrocomm4 24850 15.5 0.084 0.096 2.739 7.025 mp_irecv_dv 75445 16.2 2.586 6.745 2.586 6.745 calculate_dm_sparse 127 9.5 0.001 0.001 6.336 6.399 fft3d_ps 1281 14.7 2.822 2.894 5.923 5.993 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.625 5.671 grid_collocate_task_list 127 9.7 5.254 5.585 5.254 5.585 mp_alltoall_d11v 2401 14.1 4.917 5.239 4.917 5.239 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.505 4.597 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.498 4.595 qs_energies_init_hamiltonians 11 5.9 0.005 0.011 4.574 4.578 potential_pw2rs 127 12.3 0.023 0.024 4.334 4.351 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.244 4.313 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.960000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1071.545455, yerr=22.386537 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 1.593844E+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 0.070 0.091 169.363 169.367 qs_mol_dyn_low 1 2.0 0.003 0.004 168.749 168.779 qs_forces 11 3.9 0.010 0.019 168.563 168.574 qs_energies 11 4.9 0.010 0.023 161.131 161.145 scf_env_do_scf 11 5.9 0.001 0.001 143.418 143.429 velocity_verlet 10 3.0 0.002 0.003 111.074 111.082 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 107.921 107.923 dbcsr_multiply_generic 2507 12.6 0.187 0.197 72.305 72.778 qs_scf_new_mos 117 7.6 0.001 0.001 71.352 71.466 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.351 71.465 ot_scf_mini 117 9.6 0.003 0.004 66.954 67.073 multiply_cannon 2507 13.6 0.584 0.623 53.534 56.740 multiply_cannon_loop 2507 14.6 0.807 0.843 50.532 51.567 ot_mini 117 10.6 0.001 0.001 37.597 37.716 init_scf_loop 11 6.9 0.000 0.000 35.347 35.349 mp_waitall_1 125778 16.7 25.027 31.710 25.027 31.710 prepare_preconditioner 11 7.9 0.000 0.000 31.506 31.531 make_preconditioner 11 8.9 0.000 0.000 31.506 31.531 make_full_inverse_cholesky 11 9.9 0.015 0.025 29.422 29.663 rebuild_ks_matrix 128 8.3 0.001 0.001 28.962 29.102 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.962 29.101 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.412 26.540 multiply_cannon_multrec 10028 15.6 10.306 15.008 17.990 22.665 qs_ot_get_derivative 117 11.6 0.001 0.002 20.442 20.558 multiply_cannon_metrocomm3 10028 15.6 0.026 0.027 12.848 20.535 cp_fm_cholesky_invert 11 10.9 18.055 18.061 18.055 18.061 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 17.204 17.390 apply_single 128 13.6 0.001 0.001 17.204 17.390 ot_diis_step 117 11.6 0.020 0.020 17.087 17.087 qs_ot_get_p 128 10.4 0.001 0.001 16.615 16.748 make_m2s 5014 13.6 0.066 0.069 14.919 15.750 make_images 5014 14.6 2.160 2.595 14.614 15.447 sum_up_and_integrate 128 10.3 0.181 0.191 14.096 14.146 integrate_v_rspace 128 11.3 0.004 0.004 13.914 13.972 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.880 13.929 calculate_rho_elec 128 8.7 0.256 0.267 13.879 13.928 qs_ot_p2m_diag 83 11.4 0.496 0.501 12.914 12.930 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.752 11.752 multiply_cannon_sync_h2d 10028 15.6 10.669 11.355 10.669 11.355 init_scf_run 11 5.9 0.000 0.001 10.843 10.843 scf_env_initial_rho_setup 11 6.9 0.006 0.013 10.843 10.843 make_images_data 5014 15.6 0.055 0.065 8.618 9.852 hybrid_alltoall_any 5200 16.5 0.845 3.791 8.422 9.826 pw_transfer 1547 11.6 0.085 0.093 9.158 9.176 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.935 8.961 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.914 8.924 cp_fm_diag_elpa_base 83 14.4 8.672 8.752 8.906 8.916 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.998 8.086 grid_integrate_task_list 128 12.3 7.749 8.079 7.749 8.079 dbcsr_mm_accdrv_process 20762 16.1 3.353 4.421 7.337 8.054 fft_wrap_pw1pw2_140 523 13.2 1.818 1.857 7.954 7.976 cp_fm_cholesky_decompose 22 10.9 7.808 7.913 7.808 7.913 multiply_cannon_metrocomm1 10028 15.6 0.031 0.032 4.354 7.755 wfi_extrapolate 11 7.9 0.001 0.001 7.424 7.424 density_rs2pw 128 9.7 0.006 0.006 6.580 6.908 calculate_dm_sparse 128 9.5 0.001 0.001 6.071 6.176 grid_collocate_task_list 128 9.7 5.563 5.810 5.563 5.810 dbcsr_complete_redistribute 395 12.7 2.198 2.347 5.374 5.759 fft3d_ps 1291 14.7 2.766 2.838 5.646 5.695 multiply_cannon_metrocomm4 7521 15.6 0.026 0.030 1.877 5.683 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.573 5.622 mp_irecv_dv 28860 15.9 1.836 5.598 1.836 5.598 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 5.298 5.312 mp_alltoall_d11v 2415 14.1 4.652 5.100 4.652 5.100 mp_allgather_i34 2507 14.6 1.153 4.610 1.153 4.610 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.301 4.375 potential_pw2rs 128 12.3 0.027 0.028 4.154 4.167 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.580 3.928 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.526 3.842 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.610 3.702 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.615 3.657 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.435 3.575 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.452 3.469 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=169.367000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1456.818182, yerr=41.873876 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430457200640 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986252263424 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992001093632 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613077360640 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239162695680 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239162695680 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 94.232994E+12 0.0% 0.0% 100.0% flops max/rank 11.786057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806382528 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.178742E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144969E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318395E+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 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592214928 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.086 0.145 293.298 293.300 qs_mol_dyn_low 1 2.0 0.003 0.004 292.113 292.447 qs_forces 11 3.9 0.013 0.022 290.645 290.649 qs_energies 11 4.9 0.002 0.002 281.835 281.845 scf_env_do_scf 11 5.9 0.001 0.001 259.345 259.351 velocity_verlet 10 3.0 0.002 0.002 211.669 211.842 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 135.921 135.923 init_scf_loop 11 6.9 0.000 0.000 123.149 123.152 prepare_preconditioner 11 7.9 0.000 0.000 118.293 118.325 make_preconditioner 11 8.9 0.000 0.000 118.293 118.325 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.791 115.426 qs_scf_new_mos 118 7.6 0.001 0.001 92.035 92.170 qs_scf_loop_do_ot 118 8.6 0.001 0.001 92.034 92.169 ot_scf_mini 118 9.6 0.004 0.004 87.042 87.127 dbcsr_multiply_generic 2529 12.6 0.216 0.228 83.872 84.658 cp_fm_upper_to_full 106 14.8 52.638 75.141 52.638 75.141 multiply_cannon 2529 13.6 0.685 0.739 60.217 61.306 multiply_cannon_loop 2529 14.6 1.062 1.091 55.969 57.973 ot_mini 118 10.6 0.001 0.001 45.341 45.453 dbcsr_complete_redistribute 397 12.7 4.010 4.042 29.433 42.551 copy_fm_to_dbcsr 210 11.7 0.001 0.002 25.929 39.032 transfer_fm_to_dbcsr 11 9.9 0.030 0.031 23.458 36.403 rebuild_ks_matrix 129 8.3 0.001 0.001 34.461 34.596 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 34.461 34.596 mp_waitall_1 104580 16.8 29.001 34.209 29.001 34.209 mp_alltoall_i22 720 14.1 21.247 34.169 21.247 34.169 cp_fm_cholesky_invert 11 10.9 33.196 33.202 33.196 33.202 qs_ks_update_qs_env 129 7.6 0.001 0.001 32.029 32.150 qs_ot_get_p 129 10.4 0.001 0.001 26.265 26.383 qs_ot_get_derivative 118 11.6 0.002 0.002 24.994 25.087 qs_ot_p2m_diag 84 11.4 0.890 0.896 22.049 22.080 multiply_cannon_metrocomm3 10116 15.6 0.026 0.028 19.575 21.601 make_m2s 5058 13.6 0.077 0.080 18.827 20.434 ot_diis_step 118 11.6 0.022 0.023 20.299 20.299 cp_dbcsr_syevd 84 12.4 0.006 0.006 20.230 20.232 make_images 5058 14.6 3.083 3.273 18.343 19.948 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.447 19.783 apply_single 129 13.6 0.001 0.001 19.447 19.783 multiply_cannon_multrec 10116 15.6 10.401 12.301 18.252 18.332 cp_fm_diag_elpa 84 13.4 0.000 0.000 17.030 17.031 cp_fm_diag_elpa_base 84 14.4 12.543 14.110 17.026 17.027 qs_rho_update_rho_low 129 7.7 0.001 0.001 16.467 16.475 calculate_rho_elec 129 8.7 0.485 0.486 16.466 16.474 sum_up_and_integrate 129 10.3 0.323 0.325 16.298 16.384 integrate_v_rspace 129 11.3 0.005 0.005 15.974 16.060 multiply_cannon_sync_h2d 10116 15.6 14.405 14.437 14.405 14.437 hybrid_alltoall_any 5245 16.5 1.319 3.092 10.961 13.333 make_images_data 5058 15.6 0.065 0.070 10.707 12.986 init_scf_run 11 5.9 0.000 0.001 12.192 12.193 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.192 12.193 pw_transfer 1559 11.6 0.094 0.095 11.322 11.331 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 11.084 11.091 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 9.958 10.023 fft_wrap_pw1pw2_140 527 13.2 3.077 3.112 9.897 9.907 dbcsr_mm_accdrv_process 20934 16.1 4.336 6.300 7.599 9.560 wfi_extrapolate 11 7.9 0.001 0.001 8.963 8.963 cp_fm_cholesky_decompose 22 10.9 8.827 8.855 8.827 8.855 grid_integrate_task_list 129 12.3 8.618 8.808 8.618 8.808 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 8.121 8.131 density_rs2pw 129 9.7 0.006 0.006 7.454 7.500 mp_alltoall_d11v 2429 14.1 6.822 7.007 6.822 7.007 calculate_dm_sparse 129 9.5 0.001 0.001 6.920 6.992 grid_collocate_task_list 129 9.7 6.463 6.543 6.463 6.543 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.438 6.471 fft3d_ps 1301 14.7 2.875 2.884 6.284 6.346 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.182 6.252 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=293.300000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2809.090909, yerr=179.148214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 1.260003E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.102 85.377 85.379 qs_energies 1 2.0 0.002 0.014 84.629 84.647 ls_scf 1 3.0 0.006 0.045 83.680 83.691 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.496 72.668 multiply_cannon 111 7.7 0.017 0.020 55.791 57.147 multiply_cannon_loop 111 8.7 0.226 0.240 52.408 53.819 ls_scf_main 1 4.0 0.000 0.004 52.254 52.255 density_matrix_trs4 2 5.0 0.002 0.003 46.651 46.733 ls_scf_init_scf 1 4.0 0.001 0.009 28.322 28.324 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.169 27.236 mp_waitall_1 11031 10.9 22.419 25.836 22.419 25.836 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 25.030 25.043 multiply_cannon_multrec 2664 9.7 8.141 9.022 15.525 17.395 multiply_cannon_sync_h2d 2664 9.7 13.514 16.059 13.514 16.059 make_m2s 222 7.7 0.009 0.011 13.007 13.502 make_images 222 8.7 0.098 0.108 12.985 13.480 multiply_cannon_metrocomm1 2664 9.7 0.010 0.012 9.663 12.667 make_images_data 222 9.7 0.004 0.005 7.625 8.122 dbcsr_mm_accdrv_process 4760 10.4 0.590 0.702 7.001 7.951 hybrid_alltoall_any 227 10.6 0.216 1.830 6.586 7.710 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.461 7.211 dbcsr_mm_accdrv_process_sort 4760 11.4 6.209 7.095 6.209 7.095 calculate_norms 4752 9.8 5.520 6.181 5.520 6.181 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.061 5.207 mp_sum_l 887 5.1 3.178 4.451 3.178 4.451 make_images_sizes 222 9.7 0.000 0.000 0.658 3.780 mp_alltoall_i44 222 10.7 0.657 3.780 0.657 3.780 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.065 3.770 mp_irecv_dv 6231 10.9 2.048 3.737 2.048 3.737 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.325 3.432 arnoldi_extremal 4 6.8 0.000 0.000 3.273 3.298 arnoldi_normal_ev 4 7.8 0.001 0.007 3.273 3.297 build_subspace 16 8.4 0.009 0.012 3.135 3.143 ls_scf_post 1 4.0 0.002 0.018 3.099 3.113 ls_scf_store_result 1 5.0 0.000 0.000 2.897 2.942 dbcsr_special_finalize 555 9.7 0.005 0.006 2.372 2.779 dbcsr_merge_single_wm 555 10.7 0.460 0.606 2.363 2.770 make_images_pack 222 9.7 2.206 2.645 2.208 2.646 dbcsr_matrix_vector_mult 304 9.0 0.006 0.014 2.327 2.584 dbcsr_matrix_vector_mult_local 304 10.0 2.070 2.476 2.072 2.478 dbcsr_sort_data 658 11.4 2.160 2.464 2.160 2.464 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.351 2.430 buffer_matrices_ensure_size 222 8.7 1.758 2.120 1.758 2.120 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.779 1.786 rebuild_ks_matrix 3 7.3 0.000 0.000 1.770 1.777 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.004 1.770 1.777 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.379000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 2.194416E+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 0.052 0.073 90.375 90.376 qs_energies 1 2.0 0.001 0.011 89.356 89.390 ls_scf 1 3.0 0.000 0.002 88.012 88.037 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.138 74.509 multiply_cannon 111 7.7 0.027 0.039 52.578 57.317 ls_scf_main 1 4.0 0.001 0.008 54.470 54.475 multiply_cannon_loop 111 8.7 0.135 0.146 49.809 53.472 density_matrix_trs4 2 5.0 0.002 0.004 48.806 49.019 ls_scf_init_scf 1 4.0 0.000 0.003 29.857 29.859 mp_waitall_1 9105 10.9 20.695 29.199 20.695 29.199 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.643 28.759 multiply_cannon_multrec 1332 9.7 13.287 17.069 22.560 27.445 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.254 26.266 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.446 19.972 make_m2s 222 7.7 0.006 0.007 15.029 15.502 make_images 222 8.7 1.364 1.686 14.998 15.472 dbcsr_mm_accdrv_process 4041 10.4 0.350 0.544 8.872 10.466 dbcsr_mm_accdrv_process_sort 4041 11.4 8.389 9.922 8.389 9.922 make_images_data 222 9.7 0.004 0.005 8.738 9.614 hybrid_alltoall_any 227 10.6 0.543 2.553 8.105 9.537 mp_sum_l 887 5.1 5.063 8.280 5.063 8.280 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.167 7.670 mp_irecv_dv 3311 11.0 3.147 7.616 3.147 7.616 calculate_norms 2376 9.8 6.061 6.639 6.061 6.639 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.839 6.611 multiply_cannon_sync_h2d 1332 9.7 4.884 6.124 4.884 6.124 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.898 5.126 arnoldi_extremal 4 6.8 0.000 0.000 4.609 4.629 arnoldi_normal_ev 4 7.8 0.001 0.005 4.609 4.629 build_subspace 16 8.4 0.014 0.021 4.352 4.355 ls_scf_post 1 4.0 0.000 0.001 3.685 3.716 ls_scf_store_result 1 5.0 0.000 0.000 3.387 3.505 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.137 3.355 dbcsr_matrix_vector_mult_local 304 10.0 2.736 3.214 2.739 3.215 mp_allgather_i34 111 8.7 0.759 3.029 0.759 3.029 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.612 2.703 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.159 2.578 dbcsr_data_new 4174 10.1 2.113 2.395 2.113 2.395 make_images_pack 222 9.7 1.823 2.129 1.825 2.131 dbcsr_sort_data 436 11.2 1.809 2.040 1.809 2.040 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.859 1.864 rebuild_ks_matrix 3 7.3 0.000 0.000 1.846 1.851 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.007 1.846 1.851 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.376000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1787.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 2.899513E+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 0.051 0.079 91.733 91.736 qs_energies 1 2.0 0.000 0.001 91.022 91.026 ls_scf 1 3.0 0.000 0.001 89.597 89.602 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.261 74.565 multiply_cannon 111 7.7 0.038 0.065 51.445 55.958 ls_scf_main 1 4.0 0.000 0.004 55.834 55.839 multiply_cannon_loop 111 8.7 0.116 0.130 48.739 51.820 density_matrix_trs4 2 5.0 0.002 0.004 49.895 50.035 mp_waitall_1 7281 11.0 23.180 32.420 23.180 32.420 ls_scf_init_scf 1 4.0 0.001 0.007 30.108 30.110 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.898 28.993 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.538 26.552 multiply_cannon_multrec 888 9.7 12.642 15.165 21.201 24.628 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 10.606 21.447 make_m2s 222 7.7 0.006 0.008 16.509 17.176 make_images 222 8.7 1.577 1.836 16.471 17.139 make_images_data 222 9.7 0.004 0.004 9.691 10.591 hybrid_alltoall_any 227 10.6 0.640 2.946 9.094 10.443 dbcsr_mm_accdrv_process 3754 10.4 0.304 0.495 8.084 9.345 dbcsr_mm_accdrv_process_sort 3754 11.4 7.649 8.850 7.649 8.850 mp_sum_l 887 5.1 4.624 7.942 4.624 7.942 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.523 7.444 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.483 7.125 mp_irecv_dv 2335 11.1 2.467 7.070 2.467 7.070 multiply_cannon_sync_h2d 888 9.7 6.096 6.911 6.096 6.911 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.484 6.273 arnoldi_extremal 4 6.8 0.000 0.000 5.099 5.120 arnoldi_normal_ev 4 7.8 0.002 0.008 5.099 5.120 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.769 4.969 build_subspace 16 8.4 0.014 0.020 4.789 4.797 calculate_norms 1584 9.8 4.344 4.729 4.344 4.729 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.446 3.791 mp_allgather_i34 111 8.7 0.817 3.746 0.817 3.746 ls_scf_post 1 4.0 0.000 0.001 3.656 3.661 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.594 3.020 3.596 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.483 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.858 2.980 dbcsr_data_new 4116 9.9 2.105 2.450 2.105 2.450 dbcsr_sort_data 325 11.1 1.911 2.175 1.911 2.175 make_images_sizes 222 9.7 0.000 0.000 0.957 2.060 mp_alltoall_i44 222 10.7 0.956 2.060 0.956 2.060 make_images_pack 222 9.7 1.624 1.892 1.627 1.895 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.887 1.889 dbcsr_finalize 304 7.8 0.026 0.032 1.612 1.871 rebuild_ks_matrix 3 7.3 0.000 0.000 1.869 1.871 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.869 1.871 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=91.736000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2268.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 3.333964E+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 0.036 0.048 98.000 98.002 qs_energies 1 2.0 0.000 0.000 97.278 97.283 ls_scf 1 3.0 0.000 0.000 95.407 95.414 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.170 79.378 ls_scf_main 1 4.0 0.000 0.000 59.454 59.455 multiply_cannon 111 7.7 0.045 0.082 51.985 56.694 density_matrix_trs4 2 5.0 0.002 0.003 53.334 53.470 multiply_cannon_loop 111 8.7 0.153 0.169 46.946 50.226 ls_scf_init_scf 1 4.0 0.000 0.000 32.728 32.731 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.523 31.590 mp_waitall_1 6369 11.0 23.187 29.152 23.187 29.152 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.055 29.065 multiply_cannon_multrec 1332 9.7 14.099 17.033 21.987 24.975 make_m2s 222 7.7 0.007 0.008 21.272 22.712 make_images 222 8.7 3.140 3.612 21.221 22.664 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.452 17.584 make_images_data 222 9.7 0.004 0.004 11.886 13.511 hybrid_alltoall_any 227 10.6 0.798 3.744 11.211 12.928 dbcsr_mm_accdrv_process 3641 10.4 0.302 0.480 7.532 9.053 dbcsr_mm_accdrv_process_sort 3641 11.4 7.092 8.563 7.092 8.563 mp_sum_l 887 5.1 4.339 8.273 4.339 8.273 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.369 6.589 multiply_cannon_sync_h2d 1332 9.7 5.447 6.008 5.447 6.008 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.093 5.977 mp_irecv_dv 3229 10.9 2.067 5.920 2.067 5.920 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.682 5.440 arnoldi_extremal 4 6.8 0.000 0.000 5.241 5.259 arnoldi_normal_ev 4 7.8 0.007 0.022 5.241 5.259 build_subspace 16 8.4 0.014 0.021 4.876 4.885 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.528 4.701 calculate_norms 2376 9.8 4.178 4.493 4.178 4.493 mp_allgather_i34 111 8.7 2.151 4.448 2.151 4.448 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.583 3.880 dbcsr_matrix_vector_mult_local 304 10.0 3.190 3.692 3.192 3.694 dbcsr_sort_data 658 11.4 3.038 3.450 3.038 3.450 dbcsr_special_finalize 555 9.7 0.006 0.007 2.780 3.270 dbcsr_merge_single_wm 555 10.7 0.535 0.670 2.772 3.262 ls_scf_post 1 4.0 0.000 0.000 3.224 3.229 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.014 3.054 ls_scf_store_result 1 5.0 0.000 0.000 2.969 3.028 dbcsr_data_release 10477 10.7 1.588 2.376 1.588 2.376 dbcsr_finalize 304 7.8 0.049 0.061 1.805 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.002000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2698.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 4.740510E+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 0.046 0.061 92.834 92.835 qs_energies 1 2.0 0.000 0.000 92.059 92.063 ls_scf 1 3.0 0.000 0.000 90.109 90.113 dbcsr_multiply_generic 111 6.7 0.018 0.020 71.693 71.957 ls_scf_main 1 4.0 0.000 0.000 57.241 57.241 multiply_cannon 111 7.7 0.086 0.247 53.116 56.104 multiply_cannon_loop 111 8.7 0.088 0.092 50.498 52.559 density_matrix_trs4 2 5.0 0.002 0.003 50.266 50.357 ls_scf_init_scf 1 4.0 0.000 0.000 29.454 29.455 mp_waitall_1 5436 11.0 25.084 29.168 25.084 29.168 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.213 28.251 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.113 26.129 multiply_cannon_multrec 444 9.7 13.748 16.209 20.817 22.167 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.390 16.140 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.031 14.915 make_m2s 222 7.7 0.005 0.005 13.730 14.717 make_images 222 8.7 2.033 2.469 13.662 14.647 hybrid_alltoall_any 227 10.6 0.801 3.814 8.250 10.000 make_images_data 222 9.7 0.003 0.004 8.479 9.788 multiply_cannon_sync_h2d 444 9.7 6.725 8.170 6.725 8.170 dbcsr_mm_accdrv_process 3003 10.4 0.364 0.409 6.761 7.892 dbcsr_mm_accdrv_process_sort 3003 11.4 6.397 7.483 6.397 7.483 arnoldi_extremal 4 6.8 0.000 0.000 5.855 5.872 arnoldi_normal_ev 4 7.8 0.002 0.005 5.854 5.872 build_subspace 16 8.4 0.015 0.020 5.442 5.452 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.438 4.647 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.172 4.381 mp_sum_l 887 5.1 2.846 4.287 2.846 4.287 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.177 3.699 4.179 mp_allgather_i34 111 8.7 1.190 3.823 1.190 3.823 calculate_norms 792 9.8 3.621 3.759 3.621 3.759 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.540 3.631 mp_irecv_dv 1241 11.2 1.527 3.593 1.527 3.593 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.512 3.575 ls_scf_post 1 4.0 0.000 0.000 3.414 3.419 make_images_sizes 222 9.7 0.000 0.000 0.871 3.380 mp_alltoall_i44 222 10.7 0.870 3.380 0.870 3.380 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.032 3.371 ls_scf_store_result 1 5.0 0.000 0.000 3.201 3.237 dbcsr_finalize 304 7.8 0.062 0.077 2.197 2.270 dbcsr_data_new 4608 9.7 1.790 2.220 1.790 2.220 dbcsr_merge_all 275 8.9 0.480 0.525 2.057 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.011 2.012 rebuild_ks_matrix 3 7.3 0.000 0.000 1.979 1.980 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.979 1.980 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 1.934 1.934 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.835000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3689.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 8.840704E+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 0.079 0.096 104.863 104.865 qs_energies 1 2.0 0.000 0.000 103.482 103.491 ls_scf 1 3.0 0.000 0.000 100.543 100.553 dbcsr_multiply_generic 111 6.7 0.024 0.027 74.211 74.354 ls_scf_main 1 4.0 0.000 0.000 63.207 63.207 density_matrix_trs4 2 5.0 0.002 0.003 54.254 54.309 multiply_cannon 111 7.7 0.122 0.237 48.281 50.305 multiply_cannon_loop 111 8.7 0.097 0.098 45.271 45.811 ls_scf_init_scf 1 4.0 0.000 0.000 33.544 33.545 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.022 32.050 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.317 29.324 mp_waitall_1 4527 11.1 21.421 25.195 21.421 25.195 make_m2s 222 7.7 0.005 0.005 22.332 23.432 make_images 222 8.7 3.573 3.869 22.225 23.324 multiply_cannon_multrec 444 9.7 17.814 18.389 22.463 22.971 hybrid_alltoall_any 227 10.6 1.656 3.613 12.546 15.324 make_images_data 222 9.7 0.003 0.004 12.750 14.886 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.085 10.319 multiply_cannon_sync_h2d 444 9.7 8.799 8.847 8.799 8.847 arnoldi_extremal 4 6.8 0.000 0.000 7.438 7.449 arnoldi_normal_ev 4 7.8 0.003 0.009 7.437 7.449 build_subspace 16 8.4 0.026 0.036 6.853 6.864 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.492 5.628 dbcsr_matrix_vector_mult_local 304 10.0 5.043 5.349 5.045 5.352 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.081 5.172 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.887 5.121 dbcsr_mm_accdrv_process 1814 10.4 0.293 0.368 4.460 4.599 dbcsr_mm_accdrv_process_sort 1814 11.4 4.120 4.252 4.120 4.252 ls_scf_post 1 4.0 0.000 0.000 3.792 3.801 make_images_sizes 222 9.7 0.000 0.000 1.455 3.588 mp_alltoall_i44 222 10.7 1.455 3.587 1.455 3.587 ls_scf_store_result 1 5.0 0.000 0.000 3.505 3.542 mp_allgather_i34 111 8.7 1.129 3.540 1.129 3.540 calculate_norms 792 9.8 3.225 3.272 3.225 3.272 dbcsr_finalize 304 7.8 0.082 0.089 3.085 3.176 dbcsr_merge_all 275 8.9 0.882 0.917 2.863 2.943 qs_energies_init_hamiltonians 1 3.0 0.002 0.004 2.909 2.909 dbcsr_complete_redistribute 5 7.6 1.437 1.478 2.746 2.858 matrix_ls_to_qs 2 6.0 0.000 0.000 2.388 2.512 dbcsr_sort_data 325 11.1 2.444 2.508 2.444 2.508 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.416 2.418 dbcsr_data_new 6591 9.6 1.899 2.406 1.899 2.406 rebuild_ks_matrix 3 7.3 0.000 0.000 2.351 2.353 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.351 2.353 dbcsr_new_transposed 4 7.5 0.242 0.253 2.268 2.278 dbcsr_frobenius_norm 74 6.6 2.054 2.134 2.201 2.240 dbcsr_add_d 103 6.2 0.000 0.000 2.150 2.239 dbcsr_add_anytype 103 7.2 0.859 0.893 2.150 2.238 dbcsr_data_release 12724 10.6 1.981 2.191 1.981 2.191 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=104.865000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7007.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/627b89bbdc49bd82d8905e4d10d261ed7123330b_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 592.969728E+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 907280865. 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 0.052 0.114 369.763 369.766 qs_mol_dyn_low 1 2.0 0.006 0.070 368.371 368.390 qs_forces 5 3.8 0.006 0.061 368.167 368.181 qs_energies 5 4.8 0.007 0.055 363.668 363.833 scf_env_do_scf 5 5.8 0.001 0.010 345.526 345.529 scf_env_do_scf_inner_loop 105 6.6 0.003 0.015 290.876 290.879 qs_scf_new_mos 105 7.6 0.001 0.001 224.861 226.731 qs_scf_loop_do_ot 105 8.6 0.001 0.001 224.861 226.731 ot_scf_mini 105 9.6 0.003 0.004 211.435 212.536 dbcsr_multiply_generic 1445 12.2 0.126 0.134 166.690 176.078 velocity_verlet 4 3.0 0.003 0.011 160.825 160.827 multiply_cannon 1445 13.2 0.282 0.296 128.598 154.214 multiply_cannon_loop 1445 14.2 2.853 3.018 123.821 150.755 mp_waitall_1 488190 16.1 57.029 100.551 57.029 100.551 qs_ot_get_p 112 10.4 0.001 0.001 97.194 97.966 multiply_cannon_metrocomm3 69360 15.2 0.202 0.213 34.826 83.793 ot_mini 105 10.6 0.001 0.003 78.233 78.496 qs_ot_p2m_diag 40 11.0 0.020 0.032 78.144 78.462 cp_dbcsr_syevd 40 12.0 0.002 0.002 73.616 73.620 cp_fm_syevd 40 13.0 0.000 0.000 66.879 67.056 cp_fm_redistribute_end 40 14.0 30.567 60.880 30.583 60.886 cp_fm_syevd_base 40 14.0 30.281 60.639 30.281 60.639 qs_ot_get_derivative 55 11.6 0.001 0.003 55.094 55.589 init_scf_loop 7 6.6 0.000 0.007 54.500 54.505 multiply_cannon_metrocomm1 69360 15.2 0.100 0.108 10.087 49.881 prepare_preconditioner 7 7.6 0.000 0.001 47.418 47.469 make_preconditioner 7 8.6 0.000 0.007 47.418 47.469 multiply_cannon_multrec 69360 15.2 28.031 33.700 37.869 43.389 rebuild_ks_matrix 110 8.4 0.000 0.000 40.714 42.436 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.013 40.713 42.436 qs_ks_update_qs_env 112 7.6 0.001 0.001 37.355 38.928 make_full_inverse_cholesky 7 9.6 0.000 0.001 36.021 36.174 qs_rho_update_rho_low 110 7.6 0.001 0.002 33.321 34.437 calculate_rho_elec 110 8.6 0.030 0.033 33.321 34.436 multiply_cannon_sync_h2d 69360 15.2 26.897 32.495 26.897 32.495 mp_sum_l 4764 12.2 23.936 29.617 23.936 29.617 cp_fm_cholesky_invert 7 10.6 25.353 25.400 25.353 25.400 density_rs2pw 110 9.6 0.005 0.007 24.606 25.093 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.640 23.918 apply_single 62 13.6 0.000 0.000 23.640 23.918 ot_new_cg_direction 55 11.6 0.001 0.003 21.673 21.687 qs_ot_get_orbitals 105 10.6 0.001 0.001 19.649 21.412 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 20.354 20.660 transfer_rs2pw 445 10.6 0.007 0.008 18.902 19.515 pw_transfer 1645 12.4 0.079 0.102 17.344 18.091 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 17.206 17.958 sum_up_and_integrate 60 10.3 0.028 0.038 16.924 16.936 integrate_v_rspace 60 11.3 0.002 0.002 16.896 16.909 calculate_dm_sparse 110 9.5 0.000 0.001 15.261 16.142 make_m2s 2890 13.2 0.077 0.083 13.710 15.229 make_images 2890 14.2 0.241 0.260 13.605 15.120 qs_ot_get_derivative_diag 18 12.0 0.000 0.000 14.789 14.978 fft_wrap_pw1pw2_240 915 15.0 1.131 1.226 14.006 14.757 init_scf_run 5 5.8 0.000 0.001 14.742 14.743 scf_env_initial_rho_setup 5 6.8 0.001 0.005 14.742 14.743 fft3d_pb 915 16.0 2.344 2.612 12.245 13.073 qs_vxc_create 110 10.4 0.003 0.028 12.946 13.008 mp_sendrecv_dv 168740 12.6 11.866 12.254 11.866 12.254 transfer_rs2pw_30 110 11.6 1.280 1.346 10.546 12.147 check_diag 80 13.5 8.594 8.888 11.794 11.968 potential_pw2rs 60 12.3 0.003 0.003 11.420 11.540 mp_alltoall_z22v 2340 17.7 9.542 11.393 9.542 11.393 mp_alltoall_d11v 1300 13.8 9.352 11.067 9.352 11.067 make_full_single_inverse 7 9.6 0.001 0.002 10.825 10.877 dbcsr_mm_accdrv_process 154766 15.8 6.154 6.456 9.716 10.589 cp_fm_cholesky_decompose 14 10.2 9.876 9.886 9.876 9.886 transfer_pw2rs 245 13.2 0.003 0.004 9.709 9.845 xc_rho_set_and_dset_create 110 12.4 0.077 0.098 8.788 9.118 xc_vxc_pw_create 60 11.3 0.038 0.049 8.838 8.897 make_images_sizes 2890 15.2 0.004 0.004 7.319 8.870 mp_alltoall_i44 2890 16.2 7.315 8.866 7.315 8.866 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.003 8.554 8.573 acc_transpose_blocks 69360 15.2 0.360 0.379 8.084 8.485 xc_pw_derive 510 13.4 0.005 0.007 8.374 8.462 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.357 7.898 calculate_first_density_matrix 1 7.0 0.000 0.003 7.755 7.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=369.766000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 627b89bbdc49bd82d8905e4d10d261ed7123330b Summary: empty Status: OK