=== 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: 60e7f91d1367dd062fe2fa4125183e5e3d99f205 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.2, ELPA 2022.11.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.1.0, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.1, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (03.02.2023) # # \ 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 ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta_prefixed_scalapack.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/01 job id: 45184573 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/02 job id: 45184574 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/03 job id: 45184575 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/04 job id: 45184577 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/05 job id: 45184578 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/06 job id: 45184579 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/07 job id: 45184580 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/08 job id: 45184581 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/09 job id: 45184582 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/10 job id: 45184583 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/11 job id: 45184584 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/12 job id: 45184585 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/13 job id: 45184586 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/14 job id: 45184587 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/15 job id: 45184588 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/16 job id: 45184589 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/17 job id: 45184590 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/18 job id: 45184591 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/19 job id: 45184592 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/20 job id: 45184593 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/21 job id: 45184594 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/22 job id: 45184596 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/23 job id: 45184597 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/24 job id: 45184598 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/25 job id: 45184599 --- 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/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/26 job id: 45184601 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 344 9. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.035 136.025 136.026 farming_run 1 2.0 135.474 135.475 135.993 135.996 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.463116E+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 231 1098811. MP_Allreduce 491 2254389. 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 0.010 0.026 116.069 116.069 qs_energies 1 2.0 0.000 0.000 115.836 115.837 mp2_main 1 3.0 0.000 0.000 113.709 113.710 mp2_gpw_main 1 4.0 0.021 0.029 112.663 112.665 mp2_ri_gpw_compute_in 1 5.0 0.174 0.179 93.854 93.965 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 55.426 55.535 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.171 41.736 46.717 get_2c_integrals 1 6.0 0.000 0.004 37.401 38.254 integrate_v_rspace 273 8.0 0.437 0.449 25.055 29.679 pw_transfer 6555 10.6 0.367 0.385 27.459 27.967 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 26.131 26.421 grid_integrate_task_list 273 9.0 20.863 25.927 20.863 25.927 fft_wrap_pw1pw2_100 2178 12.4 1.205 1.447 23.648 23.967 compute_2c_integrals 1 7.0 0.003 0.003 19.743 19.756 compute_2c_integrals_loop_lm 1 8.0 0.006 0.015 18.957 19.468 mp2_eri_2c_integrate_gpw 1 9.0 2.391 2.450 18.951 19.465 rpa_ri_compute_en 1 5.0 0.005 0.008 18.637 18.758 cp_fm_cholesky_decompose 12 8.2 17.672 18.564 17.672 18.564 cholesky_decomp 1 7.0 0.000 0.000 16.504 17.378 fft3d_s 5443 13.4 16.158 16.408 16.181 16.430 ao_to_mo_and_store_B_mult_1 272 7.0 10.856 15.598 10.856 15.598 calculate_wavefunction 272 8.0 5.470 5.535 12.595 13.298 rpa_num_int 1 6.0 0.000 0.002 10.734 10.743 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.712 10.729 calc_mat_Q 8 8.0 0.000 0.000 9.486 9.566 contract_S_to_Q 8 9.0 0.000 0.000 8.910 8.991 calc_potential_gpw 544 9.5 0.006 0.007 8.268 8.658 parallel_gemm_fm 14 9.1 0.000 0.000 8.469 8.568 parallel_gemm_fm_cosma 14 10.1 8.469 8.568 8.469 8.568 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.233 8.563 potential_pw2rs 545 10.0 0.107 0.109 7.704 8.355 collocate_single_gaussian 272 10.0 0.039 0.042 7.505 7.789 create_integ_mat 1 6.0 0.014 0.027 7.690 7.699 array2fm 1 7.0 0.000 0.000 6.595 7.054 pw_scatter_s 2720 13.7 4.443 4.642 4.443 4.642 pw_gather_s 2722 13.2 3.907 4.320 3.907 4.320 array2fm_buffer_send 1 8.0 2.872 3.128 2.872 3.128 pw_poisson_solve 545 10.5 1.129 1.206 2.171 2.347 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=112.663605, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2735.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 344 10. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.027 0.040 398.842 398.844 farming_run 1 2.0 397.997 398.002 398.807 398.809 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.232388E+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 744 385880. MP_Allreduce 1941 22272. MP_Sync 37 MP_Alltoall 77 2372317. 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 0.013 0.042 210.499 210.500 qs_energies 1 2.0 0.000 0.000 210.241 210.252 scf_env_do_scf 1 3.0 0.000 0.000 106.999 106.999 qs_ks_update_qs_env 5 5.0 0.000 0.000 105.870 105.879 rebuild_ks_matrix 4 6.0 0.000 0.000 105.869 105.877 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.065 105.869 105.877 hfx_ks_matrix 4 8.0 0.001 0.001 105.232 105.236 integrate_four_center 4 9.0 0.145 0.458 105.231 105.235 mp2_main 1 3.0 0.000 0.000 102.944 102.955 mp2_gpw_main 1 4.0 0.041 0.067 101.978 101.991 integrate_four_center_main 4 10.0 0.132 0.588 96.746 99.096 integrate_four_center_bin 264 11.0 96.614 99.072 96.614 99.072 init_scf_loop 1 4.0 0.000 0.000 92.428 92.428 mp2_ri_gpw_compute_in 1 5.0 0.065 0.067 75.064 76.186 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 54.508 55.611 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.158 42.165 47.405 integrate_v_rspace 95 8.0 0.398 0.571 28.773 33.836 pw_transfer 2240 10.6 0.142 0.187 30.078 30.515 fft_wrap_pw1pw2 1868 11.4 0.018 0.024 29.082 29.559 grid_integrate_task_list 95 9.0 23.814 29.066 23.814 29.066 mp2_ri_gpw_compute_en 1 5.0 0.062 0.075 26.745 28.440 ao_to_mo_and_store_B_mult_1 91 7.0 10.655 27.357 10.655 27.357 fft_wrap_pw1pw2_100 730 12.4 1.304 1.483 26.663 27.181 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.850 1.907 24.979 24.988 get_2c_integrals 1 6.0 0.000 0.000 20.468 20.509 compute_2c_integrals 1 7.0 0.004 0.007 19.449 19.476 compute_2c_integrals_loop_lm 1 8.0 0.004 0.009 18.926 19.287 mp2_eri_2c_integrate_gpw 1 9.0 1.736 1.901 18.923 19.287 fft3d_s 1823 13.4 18.436 18.794 18.450 18.807 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.566 14.567 calculate_wavefunction 91 8.0 2.024 2.062 9.754 9.996 potential_pw2rs 186 10.0 0.034 0.036 8.917 9.533 mp2_ri_gpw_compute_en_expansio 172 7.0 0.557 0.584 8.778 9.228 local_gemm 172 8.0 8.221 8.671 8.221 8.671 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.282 8.642 mp2_ri_gpw_compute_en_comm 22 7.0 0.500 0.521 7.961 8.374 calc_potential_gpw 182 9.5 0.002 0.003 7.927 8.151 collocate_single_gaussian 91 10.0 0.017 0.021 7.908 8.132 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.426 6.345 6.426 mp_sendrecv_dm3 2068 8.0 5.990 6.411 5.990 6.411 mp_sync 37 10.5 3.058 5.638 3.058 5.638 pw_gather_s 912 13.2 4.902 5.437 4.902 5.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.973548, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1519.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 452.603904E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 272. MP_Sync 530 MP_Alltoall 2083 578641. 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 0.011 0.109 52.560 52.600 qs_mol_dyn_low 1 2.0 0.003 0.003 52.262 52.270 qs_forces 11 3.9 0.002 0.003 52.170 52.171 qs_energies 11 4.9 0.001 0.001 50.613 50.627 scf_env_do_scf 11 5.9 0.000 0.001 43.948 43.948 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 41.782 41.783 dbcsr_multiply_generic 2286 12.5 0.091 0.095 31.720 32.117 qs_scf_new_mos 108 7.5 0.000 0.001 31.507 31.794 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.506 31.793 ot_scf_mini 108 9.5 0.002 0.002 29.984 30.162 multiply_cannon 2286 13.5 0.184 0.196 25.663 27.115 multiply_cannon_loop 2286 14.5 1.443 1.549 24.781 26.243 velocity_verlet 10 3.0 0.003 0.006 24.819 24.820 ot_mini 108 10.5 0.001 0.001 17.605 17.858 qs_ot_get_derivative 108 11.5 0.001 0.001 14.574 14.737 mp_waitall_1 245248 16.5 7.658 13.496 7.658 13.496 multiply_cannon_metrocomm3 54864 15.5 0.071 0.077 5.160 11.945 multiply_cannon_multrec 54864 15.5 4.432 7.003 8.071 11.789 rebuild_ks_matrix 119 8.3 0.000 0.000 8.164 8.296 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 8.163 8.295 qs_ot_get_p 119 10.4 0.001 0.001 8.054 8.294 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.145 7.266 multiply_cannon_sync_h2d 54864 15.5 6.185 6.846 6.185 6.846 init_scf_run 11 5.9 0.000 0.001 5.414 5.414 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.414 5.414 mp_sum_l 7207 12.9 3.597 5.296 3.597 5.296 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.861 5.292 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.207 5.234 dbcsr_mm_accdrv_process 76910 16.1 1.072 1.818 3.559 5.103 sum_up_and_integrate 119 10.3 0.012 0.015 4.768 4.775 integrate_v_rspace 119 11.3 0.002 0.003 4.756 4.763 cp_dbcsr_syevd 50 12.0 0.003 0.006 4.564 4.566 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.456 4.559 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.224 4.297 calculate_rho_elec 119 8.7 0.012 0.017 4.223 4.296 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.280 4.281 cp_fm_redistribute_end 50 14.0 2.177 4.243 2.184 4.247 cp_fm_diag_elpa_base 50 14.0 2.054 4.131 2.059 4.138 calculate_dm_sparse 119 9.5 0.000 0.001 3.428 3.564 jit_kernel_multiply 13 15.8 2.430 3.255 2.430 3.255 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.012 3.216 apply_single 119 13.6 0.000 0.000 3.012 3.216 calculate_first_density_matrix 1 7.0 0.000 0.000 3.087 3.092 rs_pw_transfer 974 11.9 0.012 0.013 2.892 3.058 ot_diis_step 108 11.5 0.006 0.006 2.807 2.809 density_rs2pw 119 9.7 0.004 0.004 2.249 2.443 multiply_cannon_metrocomm1 54864 15.5 0.055 0.060 1.549 2.324 wfi_extrapolate 11 7.9 0.001 0.001 2.266 2.266 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.262 2.265 make_m2s 4572 13.5 0.053 0.056 2.189 2.258 acc_transpose_blocks 54864 15.5 0.218 0.256 1.642 2.177 make_images 4572 14.5 0.133 0.138 2.107 2.177 init_scf_loop 11 6.9 0.000 0.000 2.148 2.148 grid_integrate_task_list 119 12.3 2.026 2.138 2.026 2.138 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 2.075 2.098 potential_pw2rs 119 12.3 0.004 0.004 2.084 2.095 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.026 2.091 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.936 1.985 pw_transfer 1439 11.6 0.053 0.059 1.872 1.943 mp_sum_d 4135 12.0 1.219 1.899 1.219 1.899 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.796 1.870 fft3d_ps 1201 14.6 0.370 0.476 1.566 1.632 mp_alltoall_d11v 2130 13.8 1.425 1.564 1.425 1.564 mp_waitany 12084 13.8 1.258 1.515 1.258 1.515 fft_wrap_pw1pw2_140 487 13.2 0.083 0.096 1.396 1.472 grid_collocate_task_list 119 9.7 1.289 1.352 1.289 1.352 make_images_sizes 4572 15.5 0.004 0.005 0.850 1.128 mp_alltoall_i44 4572 16.5 0.846 1.124 0.846 1.124 dbcsr_dot_sd 1205 11.9 0.051 0.062 0.668 1.107 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.066 1.090 prepare_preconditioner 11 7.9 0.000 0.000 1.045 1.077 make_preconditioner 11 8.9 0.000 0.000 1.045 1.077 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=52.600000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.181818, yerr=1.113404 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.620032E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 54 MP_Alltoall 2060 1030389. 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 0.010 0.028 43.069 43.070 qs_mol_dyn_low 1 2.0 0.003 0.003 42.811 42.844 qs_forces 11 3.9 0.002 0.003 41.811 41.815 qs_energies 11 4.9 0.001 0.001 40.107 40.114 scf_env_do_scf 11 5.9 0.000 0.001 34.797 34.797 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 32.226 32.226 dbcsr_multiply_generic 2286 12.5 0.099 0.102 23.017 23.390 qs_scf_new_mos 108 7.5 0.001 0.001 21.833 22.077 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.832 22.076 ot_scf_mini 108 9.5 0.002 0.003 20.931 21.100 velocity_verlet 10 3.0 0.001 0.002 20.895 20.915 multiply_cannon 2286 13.5 0.207 0.217 17.260 19.086 multiply_cannon_loop 2286 14.5 0.890 0.968 16.053 17.665 ot_mini 108 10.5 0.001 0.001 13.331 13.562 mp_waitall_1 200699 16.5 7.059 12.417 7.059 12.417 qs_ot_get_derivative 108 11.5 0.001 0.001 10.867 11.038 multiply_cannon_metrocomm3 27432 15.5 0.066 0.070 4.659 10.685 multiply_cannon_multrec 27432 15.5 1.981 4.563 6.104 9.478 rebuild_ks_matrix 119 8.3 0.000 0.000 8.086 8.235 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 8.085 8.235 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.242 7.377 dbcsr_mm_accdrv_process 47894 16.0 3.284 5.999 4.054 6.493 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 4.595 5.660 sum_up_and_integrate 119 10.3 0.024 0.027 5.018 5.023 integrate_v_rspace 119 11.3 0.002 0.003 4.994 5.002 qs_ot_get_p 119 10.4 0.001 0.001 4.701 4.927 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.331 4.804 apply_single 119 13.6 0.000 0.000 3.331 4.804 mp_sum_l 7207 12.9 2.367 4.695 2.367 4.695 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.215 4.248 calculate_rho_elec 119 8.7 0.021 0.024 4.214 4.248 init_scf_run 11 5.9 0.000 0.001 4.078 4.079 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.078 4.078 rs_pw_transfer 974 11.9 0.010 0.011 3.451 3.902 make_m2s 4572 13.5 0.052 0.054 2.944 3.166 qs_ot_p2m_diag 50 11.0 0.009 0.012 3.108 3.127 make_images 4572 14.5 0.200 0.236 2.856 3.079 density_rs2pw 119 9.7 0.004 0.004 2.524 2.974 multiply_cannon_sync_h2d 27432 15.5 2.179 2.811 2.179 2.811 potential_pw2rs 119 12.3 0.006 0.006 2.721 2.737 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.673 2.674 init_scf_loop 11 6.9 0.000 0.000 2.548 2.548 calculate_first_density_matrix 1 7.0 0.000 0.000 2.502 2.505 ot_diis_step 108 11.5 0.010 0.011 2.413 2.414 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.307 2.396 pw_transfer 1439 11.6 0.065 0.071 2.307 2.340 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.326 2.327 cp_fm_redistribute_end 50 14.0 1.178 2.296 1.182 2.300 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.215 2.251 cp_fm_diag_elpa_base 50 14.0 1.083 2.190 1.112 2.235 calculate_dm_sparse 119 9.5 0.000 0.001 2.133 2.212 jit_kernel_multiply 9 15.8 0.718 2.155 0.718 2.155 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.000 2.003 fft3d_ps 1201 14.6 0.518 0.572 1.917 1.949 grid_integrate_task_list 119 12.3 1.832 1.910 1.832 1.910 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.805 1.844 make_images_data 4572 15.5 0.044 0.050 1.286 1.811 hybrid_alltoall_any 4725 16.4 0.051 0.112 1.129 1.696 prepare_preconditioner 11 7.9 0.000 0.000 1.604 1.630 make_preconditioner 11 8.9 0.000 0.000 1.603 1.630 fft_wrap_pw1pw2_140 487 13.2 0.079 0.087 1.580 1.619 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.493 1.550 wfi_extrapolate 11 7.9 0.001 0.001 1.521 1.521 mp_alltoall_d11v 2130 13.8 1.318 1.520 1.318 1.520 acc_transpose_blocks 27432 15.5 0.108 0.113 1.191 1.501 mp_allgather_i34 2286 14.5 0.632 1.432 0.632 1.432 grid_collocate_task_list 119 9.7 1.226 1.367 1.226 1.367 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.350 1.358 rs_pw_transfer_PW2RS_50 119 14.3 0.587 0.607 1.136 1.226 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.167 1.214 make_images_sizes 4572 15.5 0.005 0.005 0.843 1.190 mp_alltoall_i44 4572 16.5 0.839 1.186 0.839 1.186 mp_sum_d 4135 12.0 0.584 1.031 0.584 1.031 mp_alltoall_z22v 1201 16.6 0.908 1.025 0.908 1.025 mp_sum_dm 438 4.9 0.975 1.021 0.975 1.021 mp_waitany 5720 13.7 0.548 1.019 0.548 1.019 md_output 10 3.0 0.000 0.000 0.024 0.983 rs_pw_transfer_RS2PW_140 130 11.5 0.138 0.146 0.551 0.978 md_write_output 11 3.9 0.020 0.937 0.022 0.971 update_particle_set 20 4.0 0.000 0.000 0.937 0.958 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 0.943 0.947 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.904 0.917 acc_transpose_blocks_kernels 27432 16.5 0.180 0.270 0.664 0.898 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=43.070000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 521.314304E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63497. MP_Allreduce 10075 307. MP_Sync 54 MP_Alltoall 1821 697512. 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 0.010 0.029 35.809 35.809 qs_mol_dyn_low 1 2.0 0.003 0.003 35.490 35.497 qs_forces 11 3.9 0.002 0.003 35.081 35.082 qs_energies 11 4.9 0.001 0.001 33.440 33.442 scf_env_do_scf 11 5.9 0.000 0.001 28.555 28.555 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 25.929 25.930 dbcsr_multiply_generic 2286 12.5 0.094 0.096 18.131 18.218 velocity_verlet 10 3.0 0.001 0.002 17.242 17.254 qs_scf_new_mos 108 7.5 0.001 0.001 16.984 16.998 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.984 16.998 ot_scf_mini 108 9.5 0.002 0.002 16.254 16.263 multiply_cannon 2286 13.5 0.194 0.202 14.039 14.839 multiply_cannon_loop 2286 14.5 0.629 0.657 13.160 14.052 ot_mini 108 10.5 0.001 0.001 10.100 10.112 qs_ot_get_derivative 108 11.5 0.001 0.001 8.488 8.526 multiply_cannon_multrec 18288 15.5 1.950 2.815 7.127 7.683 rebuild_ks_matrix 119 8.3 0.000 0.000 6.814 6.828 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.814 6.828 dbcsr_mm_accdrv_process 38222 16.0 4.179 5.627 5.093 6.322 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.010 6.024 mp_waitall_1 158411 16.6 3.632 5.182 3.632 5.182 sum_up_and_integrate 119 10.3 0.031 0.032 4.342 4.350 integrate_v_rspace 119 11.3 0.003 0.003 4.311 4.323 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.161 4.085 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.849 3.852 calculate_rho_elec 119 8.7 0.031 0.032 3.848 3.852 qs_ot_get_p 119 10.4 0.001 0.001 3.804 3.840 init_scf_run 11 5.9 0.000 0.001 3.723 3.723 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.723 3.723 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.693 3.202 rs_pw_transfer 974 11.9 0.009 0.010 2.783 3.100 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.419 3.001 apply_single 119 13.6 0.000 0.000 2.419 3.001 density_rs2pw 119 9.7 0.004 0.004 2.360 2.646 init_scf_loop 11 6.9 0.000 0.000 2.609 2.610 calculate_first_density_matrix 1 7.0 0.000 0.000 2.459 2.460 make_m2s 4572 13.5 0.044 0.046 2.228 2.406 make_images 4572 14.5 0.189 0.200 2.143 2.319 jit_kernel_multiply 10 15.9 0.863 2.311 0.863 2.311 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.303 2.309 pw_transfer 1439 11.6 0.065 0.070 2.163 2.180 potential_pw2rs 119 12.3 0.007 0.008 2.083 2.092 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.070 2.091 mp_sum_l 7207 12.9 1.474 2.046 1.474 2.046 cp_dbcsr_syevd 50 12.0 0.003 0.004 2.039 2.041 calculate_dm_sparse 119 9.5 0.000 0.001 1.961 1.973 grid_integrate_task_list 119 12.3 1.796 1.907 1.796 1.907 prepare_preconditioner 11 7.9 0.000 0.000 1.804 1.807 make_preconditioner 11 8.9 0.000 0.000 1.804 1.807 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.788 1.789 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.769 1.777 cp_fm_diag_elpa_base 50 14.0 1.742 1.757 1.766 1.775 fft3d_ps 1201 14.6 0.530 0.549 1.750 1.775 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.657 1.737 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.728 1.735 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.606 1.611 fft_wrap_pw1pw2_140 487 13.2 0.089 0.093 1.567 1.586 ot_diis_step 108 11.5 0.011 0.011 1.560 1.561 multiply_cannon_sync_h2d 18288 15.5 1.405 1.541 1.405 1.541 grid_collocate_task_list 119 9.7 1.207 1.363 1.207 1.363 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.322 1.327 multiply_cannon_metrocomm1 18288 15.5 0.028 0.029 0.649 1.297 acc_transpose_blocks 18288 15.5 0.075 0.076 1.260 1.288 make_images_data 4572 15.5 0.044 0.048 1.020 1.224 wfi_extrapolate 11 7.9 0.001 0.001 1.216 1.216 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.864 1.107 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 0.942 0.943 mp_alltoall_z22v 1201 16.6 0.805 0.907 0.805 0.907 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.881 0.903 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.876 0.877 mp_alltoall_d11v 2130 13.8 0.766 0.862 0.766 0.862 acc_transpose_blocks_kernels 18288 16.5 0.208 0.216 0.825 0.851 mp_waitany 9880 13.7 0.535 0.850 0.535 0.850 rs_pw_transfer_RS2PW_140 130 11.5 0.118 0.120 0.512 0.820 cp_fm_cholesky_invert 11 10.9 0.785 0.789 0.785 0.789 make_images_sizes 4572 15.5 0.005 0.005 0.524 0.780 mp_alltoall_i44 4572 16.5 0.520 0.775 0.520 0.775 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.649 0.742 mp_allgather_i34 2286 14.5 0.339 0.722 0.339 0.722 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=35.809000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=1.437399 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 554.377216E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63496. MP_Allreduce 10074 349. 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 0.046 0.124 40.024 40.026 qs_mol_dyn_low 1 2.0 0.003 0.004 39.618 39.651 qs_forces 11 3.9 0.002 0.003 38.567 38.568 qs_energies 11 4.9 0.004 0.026 36.838 36.845 scf_env_do_scf 11 5.9 0.004 0.028 31.599 31.601 scf_env_do_scf_inner_loop 108 6.5 0.003 0.008 27.954 27.955 velocity_verlet 10 3.0 0.001 0.002 20.485 20.528 dbcsr_multiply_generic 2286 12.5 0.098 0.102 20.110 20.220 qs_scf_new_mos 108 7.5 0.001 0.001 18.726 18.779 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.725 18.778 ot_scf_mini 108 9.5 0.002 0.003 17.735 17.782 multiply_cannon 2286 13.5 0.226 0.267 15.178 16.003 multiply_cannon_loop 2286 14.5 0.929 0.961 14.069 14.953 ot_mini 108 10.5 0.001 0.001 10.897 10.956 multiply_cannon_multrec 27432 15.5 2.353 3.037 8.745 9.159 qs_ot_get_derivative 108 11.5 0.001 0.001 9.021 9.070 dbcsr_mm_accdrv_process 47916 15.9 5.443 6.889 6.300 7.499 rebuild_ks_matrix 119 8.3 0.000 0.000 6.984 7.033 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.984 7.032 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.237 6.279 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.567 4.524 sum_up_and_integrate 119 10.3 0.036 0.038 4.141 4.150 qs_ot_get_p 119 10.4 0.001 0.001 4.049 4.117 integrate_v_rspace 119 11.3 0.003 0.003 4.105 4.115 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.975 4.009 calculate_rho_elec 119 8.7 0.040 0.046 3.974 4.009 init_scf_run 11 5.9 0.000 0.001 3.786 3.786 scf_env_initial_rho_setup 11 6.9 0.002 0.010 3.786 3.786 init_scf_loop 11 6.9 0.001 0.004 3.599 3.600 mp_waitall_1 137007 16.6 2.755 3.264 2.755 3.264 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.632 3.131 apply_single 119 13.6 0.000 0.000 2.632 3.131 make_m2s 4572 13.5 0.055 0.056 2.776 2.895 make_images 4572 14.5 0.269 0.331 2.668 2.785 rs_pw_transfer 974 11.9 0.009 0.010 2.558 2.746 prepare_preconditioner 11 7.9 0.000 0.001 2.705 2.712 make_preconditioner 11 8.9 0.001 0.004 2.705 2.712 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.302 2.634 density_rs2pw 119 9.7 0.004 0.005 2.367 2.551 mp_sum_l 7207 12.9 1.789 2.445 1.789 2.445 calculate_first_density_matrix 1 7.0 0.001 0.004 2.388 2.390 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.329 2.339 pw_transfer 1439 11.6 0.065 0.071 2.243 2.286 calculate_dm_sparse 119 9.5 0.000 0.000 2.171 2.225 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.151 2.198 cp_dbcsr_syevd 50 12.0 0.006 0.013 2.005 2.006 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.959 1.985 jit_kernel_multiply 10 15.7 0.797 1.905 0.797 1.905 grid_integrate_task_list 119 12.3 1.822 1.880 1.822 1.880 potential_pw2rs 119 12.3 0.008 0.010 1.865 1.876 fft3d_ps 1201 14.6 0.564 0.615 1.825 1.864 ot_diis_step 108 11.5 0.012 0.013 1.832 1.834 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.823 1.824 fft_wrap_pw1pw2_140 487 13.2 0.088 0.095 1.742 1.792 cp_fm_diag_elpa 50 13.0 0.000 0.001 1.697 1.708 cp_fm_diag_elpa_base 50 14.0 1.647 1.663 1.687 1.696 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.643 1.655 acc_transpose_blocks 27432 15.5 0.112 0.115 1.465 1.512 grid_collocate_task_list 119 9.7 1.221 1.344 1.221 1.344 wfi_extrapolate 11 7.9 0.001 0.001 1.340 1.340 make_images_data 4572 15.5 0.045 0.048 1.152 1.325 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.298 1.308 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.815 1.297 hybrid_alltoall_any 4725 16.4 0.062 0.150 0.980 1.196 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.172 1.189 cp_fm_upper_to_full 72 14.2 0.819 1.176 0.819 1.176 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.145 1.145 dbcsr_complete_redistribute 329 12.2 0.124 0.153 0.839 1.117 multiply_cannon_sync_h2d 27432 15.5 1.003 1.068 1.003 1.068 mp_sum_dm 438 4.9 1.009 1.060 1.009 1.060 md_write_output 11 3.9 0.043 1.010 0.045 1.046 update_particle_set 20 4.0 0.000 0.000 0.990 1.034 multiply_cannon_metrocomm1 27432 15.5 0.033 0.035 0.599 1.032 md_output 10 3.0 0.000 0.000 0.046 1.026 mp_alltoall_d11v 2130 13.8 0.841 0.952 0.841 0.952 acc_transpose_blocks_kernels 27432 16.5 0.267 0.274 0.846 0.884 mp_alltoall_z22v 1201 16.6 0.844 0.883 0.844 0.883 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.793 0.871 cp_fm_cholesky_invert 11 10.9 0.867 0.870 0.867 0.870 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.593 0.866 make_images_sizes 4572 15.5 0.005 0.005 0.600 0.849 mp_alltoall_i44 4572 16.5 0.595 0.844 0.595 0.844 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.838 0.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=40.026000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=525.909091, yerr=3.527929 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 601.563136E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63495. MP_Allreduce 10074 348. 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 0.013 0.031 28.532 28.534 qs_mol_dyn_low 1 2.0 0.003 0.003 28.359 28.363 qs_forces 11 3.9 0.002 0.003 28.282 28.283 qs_energies 11 4.9 0.001 0.001 26.561 26.566 scf_env_do_scf 11 5.9 0.000 0.001 21.730 21.731 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.108 19.109 velocity_verlet 10 3.0 0.002 0.003 14.697 14.700 dbcsr_multiply_generic 2286 12.5 0.090 0.096 12.015 12.090 qs_scf_new_mos 108 7.5 0.001 0.001 10.776 10.798 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.775 10.797 ot_scf_mini 108 9.5 0.002 0.002 10.094 10.119 multiply_cannon 2286 13.5 0.229 0.237 9.504 9.972 multiply_cannon_loop 2286 14.5 0.326 0.336 8.599 8.789 rebuild_ks_matrix 119 8.3 0.000 0.000 6.150 6.172 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 6.149 6.172 multiply_cannon_multrec 9144 15.5 1.584 1.840 5.746 5.958 ot_mini 108 10.5 0.001 0.001 5.630 5.661 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.491 5.512 qs_ot_get_derivative 108 11.5 0.001 0.001 4.337 4.364 dbcsr_mm_accdrv_process 12550 15.8 3.227 3.747 4.061 4.127 sum_up_and_integrate 119 10.3 0.038 0.042 3.815 3.821 integrate_v_rspace 119 11.3 0.003 0.003 3.777 3.783 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.724 3.731 calculate_rho_elec 119 8.7 0.060 0.061 3.723 3.731 init_scf_run 11 5.9 0.000 0.001 3.287 3.287 scf_env_initial_rho_setup 11 6.9 0.010 0.025 3.286 3.287 qs_ot_get_p 119 10.4 0.001 0.001 2.696 2.741 init_scf_loop 11 6.9 0.000 0.000 2.591 2.591 pw_transfer 1439 11.6 0.066 0.069 2.277 2.287 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.183 2.195 density_rs2pw 119 9.7 0.004 0.004 2.030 2.174 calculate_first_density_matrix 1 7.0 0.000 0.000 2.137 2.148 mp_waitall_1 115863 16.7 1.508 2.011 1.508 2.011 rs_pw_transfer 974 11.9 0.008 0.008 1.835 1.976 grid_integrate_task_list 119 12.3 1.845 1.947 1.845 1.947 make_m2s 4572 13.5 0.034 0.036 1.748 1.915 fft3d_ps 1201 14.6 0.567 0.579 1.842 1.852 prepare_preconditioner 11 7.9 0.000 0.000 1.846 1.851 make_preconditioner 11 8.9 0.000 0.000 1.846 1.851 make_images 4572 14.5 0.267 0.303 1.660 1.826 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.787 1.790 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.765 1.778 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.734 1.765 calculate_dm_sparse 119 9.5 0.000 0.000 1.738 1.755 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.588 1.588 potential_pw2rs 119 12.3 0.010 0.011 1.543 1.547 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.520 1.521 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.503 1.520 grid_collocate_task_list 119 9.7 1.270 1.388 1.270 1.388 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.355 1.363 jit_kernel_multiply 8 15.8 0.797 1.351 0.797 1.351 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.300 1.308 cp_fm_diag_elpa_base 50 14.0 1.272 1.290 1.298 1.306 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.280 1.280 ot_diis_step 108 11.5 0.013 0.013 1.280 1.280 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.167 1.201 apply_single 119 13.6 0.000 0.000 1.167 1.201 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.154 1.160 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.131 1.144 wfi_extrapolate 11 7.9 0.001 0.001 1.091 1.092 hybrid_alltoall_any 4725 16.4 0.062 0.174 0.803 1.057 make_images_data 4572 15.5 0.038 0.042 0.818 1.023 cp_fm_cholesky_invert 11 10.9 0.961 0.964 0.961 0.964 mp_alltoall_d11v 2130 13.8 0.799 0.924 0.799 0.924 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.869 0.921 acc_transpose_blocks 9144 15.5 0.037 0.038 0.893 0.912 multiply_cannon_sync_h2d 9144 15.5 0.710 0.787 0.710 0.787 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.781 0.785 qs_env_update_s_mstruct 11 6.9 0.001 0.001 0.733 0.784 mp_alltoall_z22v 1201 16.6 0.722 0.774 0.722 0.774 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.719 0.728 acc_transpose_blocks_kernels 9144 16.5 0.116 0.119 0.654 0.669 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.326 0.661 yz_to_x 606 15.1 0.265 0.275 0.645 0.661 mp_allgather_i34 2286 14.5 0.239 0.654 0.239 0.654 x_to_yz 595 16.2 0.276 0.288 0.619 0.626 mp_waitany 5200 13.7 0.482 0.607 0.482 0.607 rs_pw_transfer_RS2PW_140 130 11.5 0.100 0.102 0.438 0.576 qs_create_task_list 11 7.9 0.004 0.005 0.548 0.573 generate_qs_task_list 11 8.9 0.189 0.212 0.544 0.573 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.534000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=571.090909, yerr=3.315379 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 743.235584E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63729. MP_Allreduce 10074 433. MP_Sync 54 MP_Alltoall 1582 7383731. MP_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 0.034 0.074 45.216 45.217 qs_mol_dyn_low 1 2.0 0.003 0.003 44.615 44.619 qs_forces 11 3.9 0.028 0.041 44.058 44.058 qs_energies 11 4.9 0.001 0.001 42.008 42.022 scf_env_do_scf 11 5.9 0.001 0.001 35.635 35.635 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.452 27.453 velocity_verlet 10 3.0 0.002 0.005 24.499 24.565 dbcsr_multiply_generic 2286 12.5 0.098 0.101 18.532 18.605 qs_scf_new_mos 108 7.5 0.001 0.001 16.937 17.034 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.936 17.033 ot_scf_mini 108 9.5 0.002 0.002 15.861 15.959 multiply_cannon 2286 13.5 0.307 0.313 14.150 14.962 multiply_cannon_loop 2286 14.5 0.342 0.348 12.854 13.718 ot_mini 108 10.5 0.001 0.001 9.490 9.603 multiply_cannon_multrec 9144 15.5 3.546 4.992 8.812 8.919 init_scf_loop 11 6.9 0.000 0.000 8.154 8.157 rebuild_ks_matrix 119 8.3 0.000 0.001 7.762 7.900 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.761 7.899 qs_ot_get_derivative 108 11.5 0.001 0.001 7.475 7.572 prepare_preconditioner 11 7.9 0.000 0.000 7.152 7.165 make_preconditioner 11 8.9 0.000 0.000 7.151 7.165 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.038 7.164 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.612 7.043 dbcsr_mm_accdrv_process 12550 15.8 4.110 5.703 5.144 6.525 cp_fm_upper_to_full 72 14.2 3.316 4.809 3.316 4.809 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.627 4.703 calculate_rho_elec 119 8.7 0.118 0.121 4.626 4.703 sum_up_and_integrate 119 10.3 0.065 0.067 4.338 4.354 integrate_v_rspace 119 11.3 0.004 0.004 4.272 4.287 init_scf_run 11 5.9 0.000 0.001 4.133 4.133 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.132 4.132 mp_waitall_1 94719 16.7 2.939 3.870 2.939 3.870 qs_ot_get_p 119 10.4 0.001 0.001 3.590 3.728 pw_transfer 1439 11.6 0.068 0.069 3.050 3.056 dbcsr_complete_redistribute 329 12.2 0.302 0.309 2.111 2.987 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.952 2.957 make_m2s 4572 13.5 0.038 0.039 2.737 2.910 qs_ot_get_derivative_taylor 59 13.0 0.001 0.002 2.438 2.880 make_images 4572 14.5 0.356 0.388 2.616 2.789 calculate_first_density_matrix 1 7.0 0.000 0.000 2.602 2.678 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.779 2.653 fft3d_ps 1201 14.6 0.601 0.614 2.574 2.581 fft_wrap_pw1pw2_140 487 13.2 0.095 0.097 2.465 2.472 density_rs2pw 119 9.7 0.004 0.004 2.444 2.468 mp_alltoall_i22 627 13.8 1.512 2.432 1.512 2.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.199 2.431 apply_single 119 13.6 0.000 0.000 2.199 2.431 calculate_dm_sparse 119 9.5 0.000 0.000 2.344 2.418 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.534 2.398 multiply_cannon_metrocomm3 9144 15.5 0.019 0.020 1.345 2.189 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.112 2.112 grid_integrate_task_list 119 12.3 2.064 2.098 2.064 2.098 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.004 2.052 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.010 2.011 ot_diis_step 108 11.5 0.014 0.015 1.990 1.990 qs_energies_init_hamiltonians 11 5.9 0.039 0.055 1.960 1.970 jit_kernel_multiply 10 15.5 1.007 1.926 1.007 1.926 cp_dbcsr_syevd 50 12.0 0.004 0.004 1.835 1.836 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.748 1.791 mp_sum_l 7207 12.9 1.110 1.766 1.110 1.766 rs_pw_transfer 974 11.9 0.009 0.009 1.689 1.748 make_images_data 4572 15.5 0.043 0.046 1.359 1.663 potential_pw2rs 119 12.3 0.014 0.015 1.650 1.652 hybrid_alltoall_any 4725 16.4 0.087 0.147 1.338 1.642 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.534 1.534 cp_fm_diag_elpa_base 50 14.0 1.369 1.437 1.531 1.531 grid_collocate_task_list 119 9.7 1.492 1.523 1.492 1.523 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.414 1.488 cp_fm_cholesky_invert 11 10.9 1.442 1.445 1.442 1.445 wfi_extrapolate 11 7.9 0.001 0.001 1.388 1.389 mp_alltoall_d11v 2130 13.8 1.248 1.330 1.248 1.330 qs_env_update_s_mstruct 11 6.9 0.020 0.029 1.247 1.307 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.167 1.185 mp_alltoall_z22v 1201 16.6 1.019 1.063 1.019 1.063 yz_to_x 606 15.1 0.461 0.473 1.040 1.048 multiply_cannon_sync_h2d 9144 15.5 1.043 1.047 1.043 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.988 1.023 acc_transpose_blocks 9144 15.5 0.038 0.038 1.001 1.006 qs_create_task_list 11 7.9 0.029 0.029 0.966 0.978 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.936 0.949 generate_qs_task_list 11 8.9 0.372 0.390 0.938 0.949 x_to_yz 595 16.2 0.480 0.490 0.920 0.924 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=45.217000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=702.909091, yerr=7.204223 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 501.604352E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66218. MP_Allreduce 9696 492. MP_Sync 52 MP_Alltoall 1938 1379060. 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 0.040 0.075 84.375 84.376 qs_mol_dyn_low 1 2.0 0.003 0.003 83.902 83.912 qs_forces 11 3.9 0.049 0.112 83.460 83.461 qs_energies 11 4.9 0.001 0.001 80.551 80.616 scf_env_do_scf 11 5.9 0.000 0.001 71.422 71.425 scf_env_do_scf_inner_loop 99 6.5 0.003 0.010 65.694 65.696 dbcsr_multiply_generic 2055 12.4 0.105 0.108 50.501 50.797 qs_scf_new_mos 99 7.5 0.000 0.001 46.826 46.979 qs_scf_loop_do_ot 99 8.5 0.001 0.001 46.825 46.978 ot_scf_mini 99 9.5 0.002 0.002 44.441 44.560 velocity_verlet 10 3.0 0.001 0.002 43.528 43.532 multiply_cannon 2055 13.4 0.180 0.188 42.079 43.029 multiply_cannon_loop 2055 14.4 1.486 1.515 40.920 41.872 ot_mini 99 10.5 0.001 0.001 25.952 26.096 qs_ot_get_derivative 99 11.5 0.001 0.001 19.253 19.378 multiply_cannon_multrec 49320 15.4 12.692 13.380 17.523 18.092 rebuild_ks_matrix 110 8.3 0.000 0.000 15.514 15.698 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 15.513 15.698 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.711 13.872 mp_waitall_1 220248 16.4 10.615 11.517 10.615 11.517 multiply_cannon_sync_h2d 49320 15.4 10.580 11.202 10.580 11.202 qs_ot_get_p 110 10.4 0.001 0.001 10.043 10.161 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.252 7.845 apply_single 110 13.6 0.000 0.001 7.251 7.845 sum_up_and_integrate 110 10.3 0.036 0.043 7.737 7.749 integrate_v_rspace 110 11.3 0.003 0.004 7.700 7.722 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.304 7.433 calculate_rho_elec 110 8.6 0.021 0.026 7.304 7.433 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.679 7.221 multiply_cannon_metrocomm3 49320 15.4 0.081 0.085 5.709 7.106 init_scf_run 11 5.9 0.000 0.001 6.898 6.898 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.898 6.898 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.749 6.768 ot_diis_step 99 11.5 0.006 0.006 6.570 6.570 cp_dbcsr_syevd 48 12.0 0.003 0.003 5.947 5.947 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.846 5.904 init_scf_loop 11 6.9 0.000 0.000 5.697 5.698 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.277 5.300 cp_fm_diag_elpa_base 48 14.0 5.259 5.281 5.274 5.297 dbcsr_mm_accdrv_process 87628 16.1 1.809 1.912 4.705 5.032 rs_pw_transfer 902 11.9 0.012 0.013 4.374 4.855 make_m2s 4110 13.4 0.060 0.066 4.666 4.815 make_images 4110 14.4 0.177 0.190 4.571 4.723 density_rs2pw 110 9.6 0.004 0.005 4.071 4.461 mp_sum_l 6514 12.8 3.302 4.350 3.302 4.350 wfi_extrapolate 11 7.9 0.001 0.001 4.109 4.109 pw_transfer 1331 11.6 0.054 0.065 3.957 4.071 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.870 3.986 calculate_dm_sparse 110 9.5 0.001 0.001 3.762 3.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.621 3.625 multiply_cannon_metrocomm1 49320 15.4 0.062 0.065 2.497 3.556 fft3d_ps 1111 14.6 0.776 0.879 3.398 3.504 prepare_preconditioner 11 7.9 0.000 0.000 3.410 3.428 make_preconditioner 11 8.9 0.000 0.000 3.410 3.428 grid_integrate_task_list 110 12.3 3.218 3.379 3.218 3.379 fft_wrap_pw1pw2_140 451 13.1 0.170 0.189 3.206 3.329 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.253 3.298 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.201 3.251 potential_pw2rs 110 12.3 0.006 0.007 3.158 3.189 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.052 3.097 mp_alltoall_d11v 2046 13.8 2.328 2.831 2.328 2.831 calculate_first_density_matrix 1 7.0 0.000 0.000 2.686 2.704 jit_kernel_multiply 13 15.9 2.628 2.701 2.628 2.701 mp_waitany 14300 13.8 2.006 2.477 2.006 2.477 make_images_data 4110 15.4 0.042 0.044 2.127 2.308 grid_collocate_task_list 110 9.6 2.084 2.308 2.084 2.308 mp_alltoall_z22v 1111 16.6 1.915 2.218 1.915 2.218 hybrid_alltoall_any 4261 16.3 0.082 0.472 1.837 2.122 acc_transpose_blocks 49320 15.4 0.200 0.206 2.016 2.107 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.943 1.965 cp_fm_cholesky_invert 11 10.9 1.836 1.840 1.836 1.840 mp_sum_d 3889 11.9 1.223 1.725 1.223 1.725 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.656 1.688 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=84.376000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.090909, yerr=2.466302 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 590.147584E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66437. MP_Allreduce 9695 570. MP_Sync 52 MP_Alltoall 1717 2462973. 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 0.020 0.052 72.916 72.919 qs_mol_dyn_low 1 2.0 0.003 0.004 72.615 72.625 qs_forces 11 3.9 0.003 0.003 72.532 72.533 qs_energies 11 4.9 0.002 0.011 69.040 69.044 scf_env_do_scf 11 5.9 0.001 0.004 60.275 60.279 scf_env_do_scf_inner_loop 99 6.5 0.008 0.021 52.087 52.089 dbcsr_multiply_generic 2055 12.4 0.114 0.118 38.786 38.961 velocity_verlet 10 3.0 0.002 0.002 37.548 37.549 qs_scf_new_mos 99 7.5 0.001 0.001 34.855 34.996 qs_scf_loop_do_ot 99 8.5 0.001 0.006 34.854 34.996 ot_scf_mini 99 9.5 0.004 0.011 33.189 33.320 multiply_cannon 2055 13.4 0.222 0.257 31.670 33.250 multiply_cannon_loop 2055 14.4 0.919 0.940 30.278 31.506 ot_mini 99 10.5 0.001 0.001 19.073 19.209 multiply_cannon_multrec 24660 15.4 7.646 9.179 13.898 15.388 rebuild_ks_matrix 110 8.3 0.000 0.000 14.375 14.479 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 14.374 14.479 qs_ot_get_derivative 99 11.5 0.001 0.001 13.284 13.415 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.607 12.694 mp_waitall_1 176588 16.5 8.594 10.899 8.594 10.899 multiply_cannon_sync_h2d 24660 15.4 7.004 8.445 7.004 8.445 init_scf_loop 11 6.9 0.001 0.004 8.153 8.156 multiply_cannon_metrocomm3 24660 15.4 0.067 0.069 5.596 8.103 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.640 7.917 apply_single 110 13.6 0.000 0.001 6.640 7.916 qs_ot_get_p 110 10.4 0.001 0.001 7.451 7.607 sum_up_and_integrate 110 10.3 0.052 0.060 7.094 7.106 integrate_v_rspace 110 11.3 0.003 0.003 7.041 7.055 dbcsr_mm_accdrv_process 52282 16.1 4.606 5.780 6.091 6.736 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.665 6.690 calculate_rho_elec 110 8.6 0.040 0.048 6.664 6.689 init_scf_run 11 5.9 0.000 0.009 6.308 6.310 scf_env_initial_rho_setup 11 6.9 0.005 0.008 6.308 6.309 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.275 6.108 prepare_preconditioner 11 7.9 0.000 0.000 6.028 6.051 make_preconditioner 11 8.9 0.000 0.002 6.028 6.051 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.570 5.752 ot_diis_step 99 11.5 0.010 0.010 5.741 5.742 qs_ot_p2m_diag 48 11.0 0.029 0.044 5.081 5.101 make_m2s 4110 13.4 0.057 0.059 4.564 5.056 make_images 4110 14.4 0.398 0.444 4.457 4.945 cp_dbcsr_syevd 48 12.0 0.004 0.004 4.651 4.652 density_rs2pw 110 9.6 0.004 0.006 3.769 4.243 pw_transfer 1331 11.6 0.066 0.073 3.971 4.120 rs_pw_transfer 902 11.9 0.012 0.014 3.612 4.119 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.032 4.046 cp_fm_diag_elpa_base 48 14.0 3.956 4.003 4.029 4.042 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.864 4.016 wfi_extrapolate 11 7.9 0.001 0.001 3.615 3.615 fft_wrap_pw1pw2_140 451 13.1 0.203 0.224 3.219 3.373 fft3d_ps 1111 14.6 1.103 1.320 3.213 3.358 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.267 3.331 grid_integrate_task_list 110 12.3 3.147 3.320 3.147 3.320 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.275 3.279 make_images_data 4110 15.4 0.046 0.050 2.552 3.054 calculate_dm_sparse 110 9.5 0.001 0.001 2.987 3.017 hybrid_alltoall_any 4261 16.3 0.102 0.438 2.241 2.992 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.906 2.959 potential_pw2rs 110 12.3 0.008 0.009 2.816 2.839 mp_sum_l 6514 12.8 1.957 2.758 1.957 2.758 calculate_first_density_matrix 1 7.0 0.002 0.021 2.581 2.587 cp_fm_cholesky_invert 11 10.9 2.546 2.553 2.546 2.553 grid_collocate_task_list 110 9.6 2.050 2.478 2.050 2.478 jit_kernel_multiply 13 16.3 1.135 2.472 1.135 2.472 mp_alltoall_d11v 2046 13.8 1.914 2.260 1.914 2.260 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.964 1.983 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.934 1.952 qs_energies_init_hamiltonians 11 5.9 0.003 0.013 1.942 1.943 mp_waitany 10164 13.8 1.337 1.884 1.337 1.884 mp_allgather_i34 2055 14.4 0.722 1.777 0.722 1.777 cp_fm_cholesky_decompose 22 10.9 1.755 1.761 1.755 1.761 multiply_cannon_metrocomm4 22605 15.4 0.076 0.082 0.791 1.721 acc_transpose_blocks 24660 15.4 0.110 0.114 1.516 1.648 mp_irecv_dv 57340 16.2 0.665 1.602 0.665 1.602 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.555 1.567 mp_alltoall_z22v 1111 16.6 1.409 1.526 1.409 1.526 dbcsr_complete_redistribute 325 12.2 0.241 0.302 1.241 1.523 rs_pw_transfer_RS2PW_140 121 11.5 0.205 0.217 0.996 1.515 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.371 1.489 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=72.919000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=557.727273, yerr=7.886593 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 660.709376E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66428. MP_Allreduce 9694 567. 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 0.038 0.065 64.007 64.009 qs_mol_dyn_low 1 2.0 0.003 0.004 63.590 63.599 qs_forces 11 3.9 0.011 0.061 63.187 63.189 qs_energies 11 4.9 0.005 0.024 59.661 59.669 scf_env_do_scf 11 5.9 0.001 0.002 51.429 51.430 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.499 42.501 velocity_verlet 10 3.0 0.005 0.032 33.713 33.724 dbcsr_multiply_generic 2055 12.4 0.107 0.112 29.980 30.230 qs_scf_new_mos 99 7.5 0.001 0.001 27.055 27.158 qs_scf_loop_do_ot 99 8.5 0.003 0.020 27.054 27.158 ot_scf_mini 99 9.5 0.005 0.023 25.808 25.916 multiply_cannon 2055 13.4 0.211 0.221 22.847 24.018 multiply_cannon_loop 2055 14.4 0.613 0.626 21.586 22.876 ot_mini 99 10.5 0.001 0.001 15.130 15.245 rebuild_ks_matrix 110 8.3 0.000 0.000 12.875 12.993 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.875 12.993 multiply_cannon_multrec 16440 15.4 3.982 5.188 9.946 11.487 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.142 11.248 qs_ot_get_derivative 99 11.5 0.001 0.001 10.675 10.784 mp_waitall_1 139946 16.5 7.779 10.691 7.779 10.691 init_scf_loop 11 6.9 0.002 0.007 8.896 8.898 multiply_cannon_metrocomm3 16440 15.4 0.042 0.045 4.468 7.721 prepare_preconditioner 11 7.9 0.000 0.001 7.049 7.065 make_preconditioner 11 8.9 0.001 0.003 7.048 7.065 sum_up_and_integrate 110 10.3 0.060 0.061 6.952 6.968 integrate_v_rspace 110 11.3 0.003 0.004 6.892 6.908 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.147 6.587 dbcsr_mm_accdrv_process 34862 16.1 4.706 5.908 5.819 6.353 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.150 6.158 calculate_rho_elec 110 8.6 0.059 0.059 6.150 6.157 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.293 6.094 apply_single 110 13.6 0.000 0.000 5.293 6.094 qs_ot_get_p 110 10.4 0.001 0.001 5.822 5.934 init_scf_run 11 5.9 0.000 0.001 5.511 5.511 scf_env_initial_rho_setup 11 6.9 0.008 0.016 5.511 5.511 make_m2s 4110 13.4 0.050 0.052 4.597 4.969 make_images 4110 14.4 0.391 0.508 4.483 4.855 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.612 4.549 ot_diis_step 99 11.5 0.010 0.011 4.418 4.419 density_rs2pw 110 9.6 0.005 0.006 3.214 4.365 multiply_cannon_sync_h2d 16440 15.4 3.703 4.165 3.703 4.165 rs_pw_transfer 902 11.9 0.012 0.017 2.971 4.132 qs_ot_p2m_diag 48 11.0 0.042 0.044 4.077 4.080 pw_transfer 1331 11.6 0.065 0.073 3.925 3.937 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.818 3.832 cp_dbcsr_syevd 48 12.0 0.003 0.004 3.740 3.740 grid_integrate_task_list 110 12.3 3.168 3.379 3.168 3.379 fft_wrap_pw1pw2_140 451 13.1 0.214 0.218 3.258 3.270 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.133 3.143 make_images_data 4110 15.4 0.043 0.048 2.638 3.143 cp_fm_diag_elpa_base 48 14.0 3.048 3.096 3.130 3.139 fft3d_ps 1111 14.6 1.091 1.101 3.115 3.126 hybrid_alltoall_any 4261 16.3 0.105 0.372 2.275 3.065 mp_sum_l 6514 12.8 1.948 2.974 1.948 2.974 wfi_extrapolate 11 7.9 0.001 0.001 2.950 2.950 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.831 2.887 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.791 2.792 cp_fm_cholesky_invert 11 10.9 2.684 2.692 2.684 2.692 potential_pw2rs 110 12.3 0.010 0.010 2.581 2.593 calculate_dm_sparse 110 9.5 0.001 0.001 2.538 2.571 mp_waitany 17072 13.8 1.285 2.513 1.285 2.513 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.451 2.493 calculate_first_density_matrix 1 7.0 0.001 0.003 2.461 2.470 grid_collocate_task_list 110 9.6 2.084 2.462 2.084 2.462 mp_alltoall_d11v 2046 13.8 1.981 2.324 1.981 2.324 multiply_cannon_metrocomm4 14385 15.4 0.046 0.049 0.868 2.298 dbcsr_complete_redistribute 325 12.2 0.325 0.358 1.758 2.276 qs_energies_init_hamiltonians 11 5.9 0.001 0.009 2.205 2.209 mp_irecv_dv 48980 15.7 0.798 2.177 0.798 2.177 rs_pw_transfer_RS2PW_140 121 11.5 0.183 0.212 0.983 2.145 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.932 1.947 cp_fm_upper_to_full 70 14.2 1.448 1.923 1.448 1.923 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.240 1.741 mp_allgather_i34 2055 14.4 0.584 1.626 0.584 1.626 cp_fm_cholesky_decompose 22 10.9 1.586 1.609 1.586 1.609 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.473 1.485 jit_kernel_multiply 8 16.4 0.726 1.468 0.726 1.468 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.350 1.449 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 0.891 1.386 qs_env_update_s_mstruct 11 6.9 0.000 0.002 1.299 1.374 rs_gather_matrices 110 12.3 0.234 0.265 1.038 1.351 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.274 1.285 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=64.009000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=627.272727, yerr=7.955454 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 739.831808E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66426. MP_Allreduce 9694 608. 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 0.016 0.030 69.766 69.767 qs_mol_dyn_low 1 2.0 0.003 0.004 69.450 69.458 qs_forces 11 3.9 0.003 0.003 68.976 68.977 qs_energies 11 4.9 0.003 0.012 65.524 65.528 scf_env_do_scf 11 5.9 0.001 0.002 56.860 56.863 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 44.504 44.504 velocity_verlet 10 3.0 0.001 0.002 39.185 39.203 dbcsr_multiply_generic 2055 12.4 0.113 0.119 30.374 30.567 qs_scf_new_mos 99 7.5 0.001 0.001 28.789 28.890 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.789 28.889 ot_scf_mini 99 9.5 0.003 0.003 27.147 27.234 multiply_cannon 2055 13.4 0.240 0.260 22.888 24.118 multiply_cannon_loop 2055 14.4 0.880 0.897 21.411 21.979 ot_mini 99 10.5 0.001 0.001 15.033 15.137 multiply_cannon_multrec 24660 15.4 4.246 6.882 12.680 13.815 rebuild_ks_matrix 110 8.3 0.000 0.000 12.657 12.757 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.656 12.756 init_scf_loop 11 6.9 0.001 0.004 12.313 12.314 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.201 11.290 qs_ot_get_derivative 99 11.5 0.001 0.001 10.726 10.819 prepare_preconditioner 11 7.9 0.000 0.000 10.493 10.506 make_preconditioner 11 8.9 0.001 0.002 10.493 10.506 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.685 10.146 dbcsr_mm_accdrv_process 52304 16.0 6.813 7.962 8.288 9.167 qs_ot_get_p 110 10.4 0.001 0.001 7.017 7.130 mp_waitall_1 121746 16.5 4.986 6.944 4.986 6.944 sum_up_and_integrate 110 10.3 0.068 0.071 6.901 6.914 integrate_v_rspace 110 11.3 0.003 0.003 6.833 6.847 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.450 6.462 calculate_rho_elec 110 8.6 0.077 0.081 6.450 6.462 make_m2s 4110 13.4 0.060 0.062 5.699 5.924 make_images 4110 14.4 0.573 0.692 5.557 5.778 init_scf_run 11 5.9 0.000 0.002 5.770 5.771 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.770 5.771 qs_ot_p2m_diag 48 11.0 0.055 0.063 4.899 4.913 cp_fm_upper_to_full 70 14.2 3.372 4.887 3.372 4.887 cp_dbcsr_syevd 48 12.0 0.003 0.004 4.415 4.417 ot_diis_step 99 11.5 0.011 0.011 4.244 4.244 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.111 4.214 apply_single 110 13.6 0.000 0.000 4.111 4.214 pw_transfer 1331 11.6 0.065 0.074 4.130 4.165 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.024 4.064 dbcsr_complete_redistribute 325 12.2 0.420 0.467 2.856 4.006 density_rs2pw 110 9.6 0.004 0.004 3.339 3.845 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.749 3.760 cp_fm_diag_elpa_base 48 14.0 3.585 3.652 3.746 3.758 multiply_cannon_metrocomm3 24660 15.4 0.036 0.037 1.666 3.674 fft_wrap_pw1pw2_140 451 13.1 0.202 0.210 3.404 3.449 grid_integrate_task_list 110 12.3 3.255 3.442 3.255 3.442 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.395 3.438 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.205 3.349 fft3d_ps 1111 14.6 1.096 1.135 3.316 3.342 multiply_cannon_sync_h2d 24660 15.4 3.166 3.307 3.166 3.307 rs_pw_transfer 902 11.9 0.010 0.011 2.702 3.218 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.152 3.203 make_images_data 4110 15.4 0.046 0.050 2.887 3.142 hybrid_alltoall_any 4261 16.3 0.120 0.460 2.479 3.066 wfi_extrapolate 11 7.9 0.001 0.001 3.037 3.037 calculate_dm_sparse 110 9.5 0.001 0.001 2.962 2.997 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.798 2.925 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.881 2.883 cp_fm_cholesky_invert 11 10.9 2.860 2.871 2.860 2.871 mp_alltoall_i22 605 13.7 1.669 2.834 1.669 2.834 calculate_first_density_matrix 1 7.0 0.000 0.002 2.631 2.634 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.471 2.507 potential_pw2rs 110 12.3 0.012 0.013 2.488 2.501 grid_collocate_task_list 110 9.6 2.182 2.460 2.182 2.460 mp_alltoall_d11v 2046 13.8 1.991 2.322 1.991 2.322 qs_energies_init_hamiltonians 11 5.9 0.002 0.012 2.298 2.302 cp_fm_cholesky_decompose 22 10.9 2.039 2.100 2.039 2.100 jit_kernel_multiply 10 15.6 1.145 1.870 1.145 1.870 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.743 1.773 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.701 1.712 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.600 1.702 mp_waitany 13376 13.8 1.140 1.646 1.140 1.646 mp_allgather_i34 2055 14.4 0.577 1.644 0.577 1.644 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.583 1.603 mp_sum_l 6514 12.8 1.105 1.599 1.105 1.599 acc_transpose_blocks 24660 15.4 0.107 0.109 1.496 1.517 multiply_cannon_metrocomm4 20550 15.4 0.062 0.065 0.849 1.469 mp_alltoall_z22v 1111 16.6 1.357 1.465 1.357 1.465 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=69.767000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=699.454545, yerr=11.873298 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 827.318272E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66424. MP_Allreduce 9694 649. 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 0.037 0.064 58.875 58.876 qs_mol_dyn_low 1 2.0 0.003 0.003 58.506 58.514 qs_forces 11 3.9 0.003 0.003 58.361 58.367 qs_energies 11 4.9 0.001 0.001 54.623 54.632 scf_env_do_scf 11 5.9 0.001 0.001 45.928 45.929 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 37.818 37.819 velocity_verlet 10 3.0 0.002 0.002 32.525 32.533 dbcsr_multiply_generic 2055 12.4 0.104 0.108 23.927 24.073 qs_scf_new_mos 99 7.5 0.001 0.001 21.818 21.884 qs_scf_loop_do_ot 99 8.5 0.001 0.001 21.817 21.884 ot_scf_mini 99 9.5 0.002 0.002 20.543 20.600 multiply_cannon 2055 13.4 0.246 0.257 18.199 19.608 multiply_cannon_loop 2055 14.4 0.318 0.331 16.795 17.077 rebuild_ks_matrix 110 8.3 0.000 0.000 12.593 12.630 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.592 12.629 ot_mini 99 10.5 0.001 0.001 11.225 11.276 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.221 11.256 multiply_cannon_multrec 8220 15.4 3.522 4.964 7.928 9.281 mp_waitall_1 103326 16.6 6.725 8.709 6.725 8.709 init_scf_loop 11 6.9 0.000 0.000 8.062 8.064 qs_ot_get_derivative 99 11.5 0.001 0.001 7.345 7.401 sum_up_and_integrate 110 10.3 0.079 0.081 7.006 7.019 integrate_v_rspace 110 11.3 0.003 0.003 6.926 6.939 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.909 6.920 calculate_rho_elec 110 8.6 0.116 0.116 6.909 6.919 prepare_preconditioner 11 7.9 0.000 0.000 6.251 6.260 make_preconditioner 11 8.9 0.000 0.000 6.251 6.260 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.844 5.914 dbcsr_mm_accdrv_process 17442 15.9 2.885 4.190 4.275 5.698 init_scf_run 11 5.9 0.000 0.002 5.545 5.545 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.545 5.545 qs_ot_get_p 110 10.4 0.001 0.001 5.323 5.407 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.238 4.939 make_m2s 4110 13.4 0.039 0.040 4.394 4.687 make_images 4110 14.4 0.635 0.690 4.265 4.557 pw_transfer 1331 11.6 0.066 0.073 4.491 4.523 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.383 4.418 density_rs2pw 110 9.6 0.004 0.004 3.547 3.920 ot_diis_step 99 11.5 0.012 0.012 3.858 3.858 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.769 3.813 apply_single 110 13.6 0.000 0.000 3.768 3.813 fft_wrap_pw1pw2_140 451 13.1 0.215 0.218 3.745 3.798 qs_ot_p2m_diag 48 11.0 0.081 0.085 3.789 3.793 fft3d_ps 1111 14.6 1.155 1.205 3.627 3.654 grid_integrate_task_list 110 12.3 3.343 3.523 3.343 3.523 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.467 3.468 multiply_cannon_sync_h2d 8220 15.4 2.893 3.062 2.893 3.062 cp_fm_cholesky_invert 11 10.9 3.050 3.054 3.050 3.054 wfi_extrapolate 11 7.9 0.001 0.001 3.049 3.049 rs_pw_transfer 902 11.9 0.010 0.010 2.657 3.042 make_images_data 4110 15.4 0.038 0.043 2.541 3.006 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.901 2.907 cp_fm_diag_elpa_base 48 14.0 2.841 2.866 2.899 2.905 hybrid_alltoall_any 4261 16.3 0.199 0.861 2.457 2.853 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 2.694 2.700 calculate_dm_sparse 110 9.5 0.001 0.001 2.563 2.602 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.569 2.571 grid_collocate_task_list 110 9.6 2.278 2.536 2.278 2.536 potential_pw2rs 110 12.3 0.015 0.016 2.478 2.487 calculate_first_density_matrix 1 7.0 0.000 0.000 2.343 2.348 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.197 2.214 mp_alltoall_d11v 2046 13.8 1.879 2.113 1.879 2.113 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.966 1.996 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.957 1.995 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.774 1.984 jit_kernel_multiply 10 15.8 1.083 1.861 1.083 1.861 mp_allgather_i34 2055 14.4 0.582 1.769 0.582 1.769 cp_fm_cholesky_decompose 22 10.9 1.739 1.767 1.739 1.767 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.709 1.714 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.528 1.664 mp_waitany 9240 13.8 1.190 1.599 1.190 1.599 dbcsr_complete_redistribute 325 12.2 0.556 0.584 1.489 1.583 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.490 1.501 mp_alltoall_z22v 1111 16.6 1.362 1.394 1.362 1.394 multiply_cannon_metrocomm4 6165 15.4 0.017 0.019 0.476 1.349 qs_create_task_list 11 7.9 0.000 0.001 1.226 1.331 generate_qs_task_list 11 8.9 0.381 0.449 1.225 1.331 mp_irecv_dv 24056 15.7 0.453 1.306 0.453 1.306 multiply_cannon_metrocomm1 8220 15.4 0.021 0.022 0.945 1.293 yz_to_x 561 15.1 0.524 0.537 1.255 1.267 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.165 0.842 1.243 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.216 1.238 x_to_yz 550 16.1 0.572 0.583 1.203 1.217 rs_gather_matrices 110 12.3 0.327 0.373 1.014 1.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=58.876000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=781.636364, yerr=11.467525 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 1.357451E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67104. MP_Allreduce 9672 819. MP_Sync 52 MP_Alltoall 1474 16505187. MP_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 0.056 90.983 90.984 qs_mol_dyn_low 1 2.0 0.003 0.003 90.632 90.642 qs_forces 11 3.9 0.004 0.007 90.547 90.551 qs_energies 11 4.9 0.001 0.001 86.346 86.350 scf_env_do_scf 11 5.9 0.001 0.001 76.200 76.200 velocity_verlet 10 3.0 0.002 0.002 57.663 57.669 scf_env_do_scf_inner_loop 99 6.5 0.004 0.009 46.747 46.749 dbcsr_multiply_generic 2055 12.4 0.118 0.123 29.320 29.457 init_scf_loop 11 6.9 0.000 0.000 29.375 29.379 qs_scf_new_mos 99 7.5 0.001 0.001 27.353 27.431 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.352 27.430 prepare_preconditioner 11 7.9 0.000 0.000 27.280 27.293 make_preconditioner 11 8.9 0.000 0.000 27.280 27.292 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.356 26.735 ot_scf_mini 99 9.5 0.002 0.002 25.561 25.627 multiply_cannon 2055 13.4 0.343 0.371 21.877 22.712 multiply_cannon_loop 2055 14.4 0.340 0.341 20.004 20.322 cp_fm_upper_to_full 70 14.2 12.974 18.601 12.974 18.601 rebuild_ks_matrix 110 8.3 0.001 0.001 14.813 14.903 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 14.812 14.903 ot_mini 99 10.5 0.001 0.001 14.174 14.243 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.460 13.543 dbcsr_complete_redistribute 325 12.2 1.022 1.047 7.533 10.835 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.480 9.776 multiply_cannon_multrec 8220 15.4 4.387 4.625 9.537 9.652 qs_ot_get_derivative 99 11.5 0.001 0.001 9.422 9.491 mp_waitall_1 84994 16.7 8.125 9.209 8.125 9.209 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.908 9.179 mp_alltoall_i22 605 13.7 5.536 8.807 5.536 8.807 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.225 8.256 calculate_rho_elec 110 8.6 0.227 0.228 8.225 8.255 sum_up_and_integrate 110 10.3 0.150 0.151 7.825 7.842 integrate_v_rspace 110 11.3 0.004 0.004 7.675 7.693 qs_ot_get_p 110 10.4 0.001 0.001 6.112 6.202 make_m2s 4110 13.4 0.043 0.043 5.470 6.031 make_images 4110 14.4 0.878 0.928 5.282 5.842 cp_fm_cholesky_invert 11 10.9 5.835 5.840 5.835 5.840 init_scf_run 11 5.9 0.000 0.001 5.797 5.797 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.797 5.797 pw_transfer 1331 11.6 0.075 0.075 5.634 5.642 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.517 5.526 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.888 5.479 apply_single 110 13.6 0.000 0.000 4.888 5.479 multiply_cannon_metrocomm3 8220 15.4 0.018 0.018 5.079 5.430 dbcsr_mm_accdrv_process 11614 15.7 3.261 3.895 5.008 5.251 fft_wrap_pw1pw2_140 451 13.1 0.221 0.223 4.753 4.763 ot_diis_step 99 11.5 0.016 0.016 4.728 4.728 fft3d_ps 1111 14.6 1.298 1.313 4.710 4.719 qs_ot_p2m_diag 48 11.0 0.151 0.155 4.426 4.433 density_rs2pw 110 9.6 0.004 0.004 4.168 4.223 cp_dbcsr_syevd 48 12.0 0.004 0.004 3.979 3.980 multiply_cannon_sync_h2d 8220 15.4 3.953 3.962 3.953 3.962 grid_integrate_task_list 110 12.3 3.725 3.886 3.725 3.886 hybrid_alltoall_any 4261 16.3 0.259 0.558 3.003 3.772 make_images_data 4110 15.4 0.041 0.044 3.030 3.749 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.737 3.741 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.983 3.525 wfi_extrapolate 11 7.9 0.001 0.001 3.394 3.394 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.348 3.349 cp_fm_diag_elpa_base 48 14.0 2.798 2.997 3.346 3.347 calculate_dm_sparse 110 9.5 0.001 0.001 3.163 3.189 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.924 2.927 grid_collocate_task_list 110 9.6 2.678 2.765 2.678 2.765 potential_pw2rs 110 12.3 0.021 0.022 2.742 2.749 rs_pw_transfer 902 11.9 0.011 0.011 2.599 2.658 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.545 2.579 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.401 2.428 qs_env_update_s_mstruct 11 6.9 0.003 0.004 2.261 2.316 mp_alltoall_d11v 2046 13.8 2.183 2.304 2.183 2.304 calculate_first_density_matrix 1 7.0 0.000 0.000 2.277 2.279 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.090 2.167 cp_fm_cholesky_decompose 22 10.9 2.019 2.041 2.019 2.041 qs_create_task_list 11 7.9 0.011 0.012 1.926 1.974 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.958 1.966 generate_qs_task_list 11 8.9 0.738 0.793 1.915 1.963 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.831 1.863 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=90.984000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1222.363636, yerr=57.906308 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 629.829632E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57910. MP_Allreduce 10979 803. MP_Sync 87 MP_Alltoall 2226 2523700. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.014 0.032 206.410 206.411 qs_mol_dyn_low 1 2.0 0.003 0.003 205.940 205.954 qs_forces 11 3.9 0.007 0.017 205.828 205.830 qs_energies 11 4.9 0.001 0.001 200.180 200.192 scf_env_do_scf 11 5.9 0.001 0.001 183.351 183.355 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 161.790 161.792 velocity_verlet 10 3.0 0.003 0.004 124.756 124.757 dbcsr_multiply_generic 2507 12.6 0.176 0.181 121.298 121.959 qs_scf_new_mos 117 7.6 0.001 0.001 120.637 120.900 qs_scf_loop_do_ot 117 8.6 0.001 0.001 120.636 120.899 ot_scf_mini 117 9.6 0.003 0.003 114.106 114.338 multiply_cannon 2507 13.6 0.243 0.252 100.588 102.404 multiply_cannon_loop 2507 14.6 1.968 2.012 98.261 100.067 ot_mini 117 10.6 0.001 0.001 63.476 63.761 multiply_cannon_multrec 60168 15.6 34.193 36.574 42.003 44.492 qs_ot_get_derivative 117 11.6 0.001 0.001 38.730 38.974 rebuild_ks_matrix 128 8.3 0.001 0.001 34.480 34.791 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 34.480 34.790 multiply_cannon_sync_h2d 60168 15.6 28.819 31.347 28.819 31.347 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.993 31.293 mp_waitall_1 267128 16.5 26.498 30.145 26.498 30.145 qs_ot_get_p 128 10.4 0.001 0.001 29.076 29.496 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.465 25.163 apply_single 128 13.6 0.001 0.001 24.464 25.162 ot_diis_step 117 11.6 0.008 0.008 24.578 24.580 qs_ot_p2m_diag 83 11.4 0.079 0.091 22.447 22.484 init_scf_loop 11 6.9 0.000 0.000 21.482 21.483 cp_dbcsr_syevd 83 12.4 0.005 0.006 20.482 20.484 cp_fm_diag_elpa 83 13.4 0.000 0.000 17.247 17.290 cp_fm_diag_elpa_base 83 14.4 17.170 17.210 17.243 17.283 multiply_cannon_metrocomm3 60168 15.6 0.118 0.123 14.327 16.884 prepare_preconditioner 11 7.9 0.000 0.000 16.840 16.878 make_preconditioner 11 8.9 0.000 0.000 16.840 16.878 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 16.527 16.694 make_full_inverse_cholesky 11 9.9 0.000 0.000 16.066 16.249 sum_up_and_integrate 128 10.3 0.090 0.109 14.792 14.808 integrate_v_rspace 128 11.3 0.004 0.005 14.702 14.721 make_m2s 5014 13.6 0.103 0.112 14.109 14.454 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.202 14.298 calculate_rho_elec 128 8.7 0.046 0.065 14.202 14.297 make_images 5014 14.6 0.398 0.416 13.929 14.290 init_scf_run 11 5.9 0.000 0.002 12.551 12.552 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.551 12.552 density_rs2pw 128 9.7 0.006 0.007 7.577 10.803 rs_pw_transfer 1046 11.9 0.017 0.019 6.343 9.610 cp_fm_cholesky_invert 11 10.9 9.443 9.450 9.443 9.450 wfi_extrapolate 11 7.9 0.001 0.001 9.309 9.309 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.703 8.917 pw_transfer 1547 11.6 0.076 0.102 8.255 8.494 calculate_dm_sparse 128 9.5 0.001 0.001 8.373 8.489 fft_wrap_pw1pw2 1291 12.7 0.010 0.013 8.049 8.292 make_images_data 5014 15.6 0.065 0.071 6.931 7.887 dbcsr_mm_accdrv_process 124484 16.2 2.770 2.869 7.333 7.857 grid_integrate_task_list 128 12.3 7.037 7.656 7.037 7.656 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.364 7.442 hybrid_alltoall_any 5200 16.5 0.292 2.263 6.049 7.339 fft_wrap_pw1pw2_140 523 13.2 0.443 0.512 6.959 7.239 multiply_cannon_metrocomm1 60168 15.6 0.091 0.096 5.216 7.227 fft3d_ps 1291 14.7 2.105 2.885 6.835 7.025 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 6.707 6.714 mp_sum_l 7870 13.0 5.327 6.651 5.327 6.651 mp_waitany 16020 13.9 2.728 6.027 2.728 6.027 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.843 5.919 grid_collocate_task_list 128 9.7 4.580 5.728 4.580 5.728 mp_alltoall_d11v 2415 14.1 4.304 5.587 4.304 5.587 rs_pw_transfer_RS2PW_140 139 11.5 0.279 0.300 2.118 5.368 cp_fm_cholesky_decompose 22 10.9 5.208 5.223 5.208 5.223 potential_pw2rs 128 12.3 0.009 0.011 5.123 5.158 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=206.411000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=6.918654 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 828.370944E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453842923456 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58204. MP_Allreduce 11006 967. MP_Sync 86 MP_Alltoall 1955 4525997. 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 0.022 0.049 202.374 202.385 qs_mol_dyn_low 1 2.0 0.003 0.004 201.855 201.869 qs_forces 11 3.9 0.005 0.005 201.710 201.713 qs_energies 11 4.9 0.002 0.005 194.590 194.603 scf_env_do_scf 11 5.9 0.001 0.002 176.512 176.522 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 141.057 141.060 velocity_verlet 10 3.0 0.003 0.007 126.781 126.783 dbcsr_multiply_generic 2485 12.5 0.185 0.190 101.533 103.147 qs_scf_new_mos 116 7.6 0.001 0.001 99.870 100.381 qs_scf_loop_do_ot 116 8.6 0.001 0.001 99.869 100.380 ot_scf_mini 116 9.6 0.004 0.004 95.045 95.532 multiply_cannon 2485 13.5 0.473 0.530 79.305 83.212 multiply_cannon_loop 2485 14.5 1.233 1.275 75.656 79.006 ot_mini 116 10.6 0.001 0.001 53.369 53.875 mp_waitall_1 212858 16.6 28.126 42.053 28.126 42.053 multiply_cannon_multrec 29820 15.5 21.361 24.897 31.365 35.405 init_scf_loop 11 6.9 0.001 0.004 35.365 35.367 rebuild_ks_matrix 127 8.3 0.001 0.001 34.550 35.178 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 34.549 35.178 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.183 31.746 qs_ot_get_derivative 116 11.6 0.001 0.002 30.642 31.165 multiply_cannon_metrocomm3 29820 15.5 0.090 0.097 17.653 31.161 prepare_preconditioner 11 7.9 0.000 0.000 30.485 30.567 make_preconditioner 11 8.9 0.000 0.002 30.485 30.567 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.877 29.548 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 22.697 24.122 apply_single 127 13.6 0.001 0.001 22.696 24.122 qs_ot_get_p 127 10.4 0.001 0.001 23.402 24.043 ot_diis_step 116 11.6 0.014 0.015 22.441 22.443 multiply_cannon_sync_h2d 29820 15.5 18.720 20.821 18.720 20.821 qs_ot_p2m_diag 82 11.4 0.186 0.214 17.981 18.026 cp_fm_cholesky_invert 11 10.9 17.598 17.610 17.598 17.610 make_m2s 4970 13.5 0.087 0.093 15.426 16.810 cp_dbcsr_syevd 82 12.4 0.006 0.006 16.757 16.759 make_images 4970 14.5 1.141 1.325 15.222 16.609 sum_up_and_integrate 127 10.3 0.116 0.131 15.689 15.717 integrate_v_rspace 127 11.3 0.004 0.004 15.573 15.605 qs_rho_update_rho_low 127 7.7 0.001 0.001 14.405 14.449 calculate_rho_elec 127 8.7 0.088 0.105 14.405 14.448 cp_fm_diag_elpa 82 13.4 0.000 0.000 13.456 13.480 cp_fm_diag_elpa_base 82 14.4 13.195 13.307 13.450 13.475 init_scf_run 11 5.9 0.000 0.002 13.001 13.003 scf_env_initial_rho_setup 11 6.9 0.001 0.002 13.001 13.003 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 12.211 12.544 make_images_data 4970 15.5 0.063 0.070 9.453 11.321 multiply_cannon_metrocomm4 27335 15.5 0.099 0.115 3.830 10.731 mp_irecv_dv 68888 16.3 3.635 10.345 3.635 10.345 hybrid_alltoall_any 5155 16.4 0.340 1.474 8.018 10.233 dbcsr_mm_accdrv_process 61726 16.2 4.611 5.890 9.475 10.208 density_rs2pw 127 9.7 0.006 0.006 7.767 10.128 wfi_extrapolate 11 7.9 0.001 0.001 9.731 9.731 pw_transfer 1535 11.6 0.086 0.107 9.394 9.484 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.171 9.257 rs_pw_transfer 1038 11.9 0.014 0.016 6.440 8.818 fft_wrap_pw1pw2_140 519 13.2 0.470 0.522 7.964 8.059 cp_fm_cholesky_decompose 22 10.9 7.653 7.743 7.653 7.743 fft3d_ps 1281 14.7 2.735 2.912 7.547 7.601 grid_integrate_task_list 127 12.3 7.094 7.526 7.094 7.526 mp_sum_l 7804 13.0 5.116 7.459 5.116 7.459 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.235 7.032 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.940 6.947 calculate_dm_sparse 127 9.5 0.001 0.001 6.583 6.710 potential_pw2rs 127 12.3 0.014 0.016 5.843 5.869 mp_allgather_i34 2485 14.5 2.240 5.848 2.240 5.848 grid_collocate_task_list 127 9.7 4.664 5.793 4.664 5.793 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.471 5.727 mp_alltoall_d11v 2401 14.1 4.926 5.518 4.926 5.518 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.361 5.429 mp_waitany 11660 13.9 2.583 4.984 2.583 4.984 dbcsr_complete_redistribute 393 12.7 0.771 0.857 4.025 4.971 mp_sum_d 4452 12.1 3.075 4.647 3.075 4.647 rs_pw_transfer_RS2PW_140 138 11.5 0.348 0.376 2.098 4.458 dbcsr_dot_sd 1305 12.0 0.765 0.842 2.814 4.171 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=202.385000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=788.727273, yerr=3.193485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 3984192 0.0% 0.0% 100.0% average stack size 0.0 0.0 1695.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 934.211584E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1042912 MPI messages size (bytes): total size 2.716210E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.604448E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 281856 36943429632 131072 < size <= 4194304 660064 996105256960 4194304 < size <= 16777216 65632 931531083728 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 10977 1007. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.210 189.671 189.709 qs_mol_dyn_low 1 2.0 0.003 0.006 189.113 189.127 qs_forces 11 3.9 0.004 0.005 188.942 188.946 qs_energies 11 4.9 0.001 0.004 181.986 181.999 scf_env_do_scf 11 5.9 0.002 0.012 165.498 165.499 scf_env_do_scf_inner_loop 117 6.6 0.004 0.013 127.449 127.450 velocity_verlet 10 3.0 0.005 0.015 121.962 121.966 qs_scf_new_mos 117 7.6 0.001 0.001 87.442 87.828 qs_scf_loop_do_ot 117 8.6 0.001 0.002 87.442 87.827 dbcsr_multiply_generic 2507 12.6 0.179 0.187 85.109 86.288 ot_scf_mini 117 9.6 0.004 0.004 83.173 83.614 multiply_cannon 2507 13.6 0.500 0.524 63.307 67.286 multiply_cannon_loop 2507 14.6 0.851 0.877 59.850 62.607 ot_mini 117 10.6 0.001 0.001 45.504 45.949 init_scf_loop 11 6.9 0.001 0.004 37.919 37.921 mp_waitall_1 170520 16.6 27.387 36.799 27.387 36.799 prepare_preconditioner 11 7.9 0.000 0.000 33.788 33.844 make_preconditioner 11 8.9 0.000 0.002 33.788 33.844 make_full_inverse_cholesky 11 9.9 0.000 0.000 31.327 32.769 rebuild_ks_matrix 128 8.3 0.001 0.001 31.796 32.280 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.017 31.796 32.280 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.472 28.922 multiply_cannon_metrocomm3 20056 15.6 0.058 0.063 16.549 26.282 qs_ot_get_derivative 117 11.6 0.002 0.002 25.306 25.742 multiply_cannon_multrec 20056 15.6 13.296 16.344 21.912 24.958 qs_ot_get_p 128 10.4 0.001 0.001 22.278 22.799 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 20.125 21.280 apply_single 128 13.6 0.001 0.001 20.125 21.280 ot_diis_step 117 11.6 0.018 0.019 20.074 20.075 qs_ot_p2m_diag 83 11.4 0.266 0.273 17.071 17.083 make_m2s 5014 13.6 0.081 0.086 15.900 16.858 make_images 5014 14.6 1.168 1.246 15.669 16.629 cp_fm_cholesky_invert 11 10.9 16.270 16.281 16.270 16.281 multiply_cannon_sync_h2d 20056 15.6 14.165 16.191 14.165 16.191 cp_dbcsr_syevd 83 12.4 0.006 0.008 16.029 16.032 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.946 15.995 calculate_rho_elec 128 8.7 0.132 0.147 15.946 15.995 sum_up_and_integrate 128 10.3 0.134 0.145 15.249 15.274 integrate_v_rspace 128 11.3 0.004 0.005 15.115 15.143 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.657 12.681 cp_fm_diag_elpa_base 83 14.4 12.182 12.354 12.649 12.671 make_images_data 5014 15.6 0.060 0.067 9.831 11.374 init_scf_run 11 5.9 0.000 0.001 11.209 11.209 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.208 11.209 density_rs2pw 128 9.7 0.007 0.014 8.683 10.663 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.027 10.363 hybrid_alltoall_any 5200 16.5 0.434 1.979 8.397 10.286 pw_transfer 1547 11.6 0.086 0.107 10.114 10.276 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.889 10.056 multiply_cannon_metrocomm4 17549 15.6 0.064 0.074 3.507 9.515 mp_irecv_dv 50230 16.2 3.384 9.265 3.384 9.265 fft_wrap_pw1pw2_140 523 13.2 0.477 0.521 8.570 8.738 fft3d_ps 1291 14.7 2.670 2.891 8.149 8.271 cp_fm_cholesky_decompose 22 10.9 8.204 8.240 8.204 8.240 dbcsr_mm_accdrv_process 41502 16.2 4.432 5.286 8.078 8.234 rs_pw_transfer 1046 11.9 0.014 0.015 6.208 8.228 grid_integrate_task_list 128 12.3 7.285 7.771 7.285 7.771 wfi_extrapolate 11 7.9 0.001 0.001 7.589 7.589 cp_fm_upper_to_full 105 14.8 5.835 7.438 5.835 7.438 dbcsr_complete_redistribute 395 12.7 1.166 1.194 4.860 6.639 calculate_dm_sparse 128 9.5 0.001 0.001 5.872 5.978 grid_collocate_task_list 128 9.7 4.894 5.818 4.894 5.818 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.067 5.779 mp_sum_l 7870 13.0 4.163 5.539 4.163 5.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.518 5.523 mp_alltoall_d11v 2415 14.1 4.775 5.401 4.775 5.401 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.581 5.363 mp_allgather_i34 2507 14.6 1.984 5.328 1.984 5.328 potential_pw2rs 128 12.3 0.020 0.023 5.282 5.300 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.001 4.573 4.704 mp_waitany 11748 13.9 2.517 4.577 2.517 4.577 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.444 4.487 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.441 4.208 mp_alltoall_i22 716 14.1 2.078 4.021 2.078 4.021 rs_pw_transfer_RS2PW_140 139 11.5 0.329 0.351 2.013 4.014 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.952 3.954 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.844 3.900 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=189.709000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=884.818182, yerr=6.561848 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/18/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 4.320339E+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 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 1.130840E+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 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58281. MP_Allreduce 10990 1091. MP_Sync 86 MP_Alltoall 1700 12496381. 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 0.098 0.283 198.563 198.564 qs_mol_dyn_low 1 2.0 0.003 0.004 197.465 197.477 qs_forces 11 3.9 0.026 0.035 197.316 197.323 qs_energies 11 4.9 0.004 0.019 190.050 190.078 scf_env_do_scf 11 5.9 0.001 0.002 172.395 172.402 velocity_verlet 10 3.0 0.002 0.002 130.333 130.337 scf_env_do_scf_inner_loop 116 6.6 0.003 0.010 123.705 123.706 qs_scf_new_mos 116 7.6 0.001 0.001 84.559 84.834 qs_scf_loop_do_ot 116 8.6 0.001 0.001 84.559 84.833 dbcsr_multiply_generic 2485 12.5 0.186 0.197 82.092 82.809 ot_scf_mini 116 9.6 0.003 0.005 79.929 80.254 multiply_cannon 2485 13.5 0.550 0.581 55.925 59.528 multiply_cannon_loop 2485 14.5 1.166 1.199 51.777 53.648 init_scf_loop 11 6.9 0.001 0.006 48.566 48.567 ot_mini 116 10.6 0.001 0.001 44.835 45.141 prepare_preconditioner 11 7.9 0.000 0.001 44.399 44.420 make_preconditioner 11 8.9 0.000 0.002 44.399 44.420 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.988 42.989 rebuild_ks_matrix 127 8.3 0.001 0.001 30.886 31.180 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.019 30.886 31.179 multiply_cannon_multrec 29820 15.5 13.821 19.456 25.628 30.747 mp_waitall_1 146592 16.7 19.728 29.793 19.728 29.793 qs_ks_update_qs_env 127 7.6 0.001 0.001 27.870 28.142 qs_ot_get_derivative 116 11.6 0.001 0.002 24.851 25.173 make_m2s 4970 13.5 0.096 0.100 21.271 22.477 make_images 4970 14.5 1.924 2.232 20.968 22.175 qs_ot_get_p 127 10.4 0.001 0.001 20.822 21.144 ot_diis_step 116 11.6 0.017 0.018 19.861 19.863 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.220 19.711 apply_single 127 13.6 0.001 0.001 19.220 19.710 cp_fm_cholesky_invert 11 10.9 18.012 18.021 18.012 18.021 cp_fm_upper_to_full 104 14.8 11.375 16.771 11.375 16.771 multiply_cannon_metrocomm3 29820 15.5 0.047 0.049 7.374 16.227 qs_ot_p2m_diag 82 11.4 0.339 0.386 16.009 16.061 sum_up_and_integrate 127 10.3 0.140 0.151 15.534 15.562 qs_rho_update_rho_low 127 7.7 0.001 0.001 15.407 15.453 calculate_rho_elec 127 8.7 0.174 0.189 15.406 15.452 integrate_v_rspace 127 11.3 0.004 0.004 15.394 15.427 cp_dbcsr_syevd 82 12.4 0.006 0.006 14.731 14.739 make_images_data 4970 15.5 0.063 0.068 11.623 13.807 dbcsr_complete_redistribute 393 12.7 1.514 1.664 9.295 13.027 multiply_cannon_sync_h2d 29820 15.5 11.610 12.965 11.610 12.965 hybrid_alltoall_any 5155 16.4 0.520 2.193 10.343 12.534 dbcsr_mm_accdrv_process 61748 16.2 7.197 8.203 11.386 11.844 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.856 11.577 cp_fm_diag_elpa 82 13.4 0.000 0.000 11.430 11.446 cp_fm_diag_elpa_base 82 14.4 10.415 10.746 11.423 11.435 init_scf_run 11 5.9 0.000 0.002 11.262 11.264 scf_env_initial_rho_setup 11 6.9 0.001 0.004 11.262 11.263 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 10.529 10.773 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.388 10.052 pw_transfer 1535 11.6 0.087 0.106 9.764 9.829 fft_wrap_pw1pw2 1281 12.7 0.010 0.012 9.538 9.611 mp_alltoall_i22 712 14.1 5.636 9.333 5.636 9.333 density_rs2pw 127 9.7 0.006 0.007 7.582 8.822 fft_wrap_pw1pw2_140 519 13.2 0.480 0.493 8.356 8.443 cp_fm_cholesky_decompose 22 10.9 7.880 7.981 7.880 7.981 grid_integrate_task_list 127 12.3 7.432 7.952 7.432 7.952 wfi_extrapolate 11 7.9 0.001 0.001 7.880 7.880 fft3d_ps 1281 14.7 2.753 2.842 7.774 7.834 multiply_cannon_metrocomm4 24850 15.5 0.083 0.095 2.821 7.304 rs_pw_transfer 1038 11.9 0.014 0.018 5.721 7.039 mp_irecv_dv 75445 16.2 2.673 7.031 2.673 7.031 mp_alltoall_d11v 2401 14.1 5.719 6.845 5.719 6.845 calculate_dm_sparse 127 9.5 0.001 0.001 6.297 6.393 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.691 5.743 grid_collocate_task_list 127 9.7 5.042 5.733 5.042 5.733 potential_pw2rs 127 12.3 0.022 0.023 5.263 5.277 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.039 5.145 qs_energies_init_hamiltonians 11 5.9 0.003 0.019 4.866 4.887 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.002 4.470 4.567 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.494 4.551 mp_allgather_i34 2485 14.5 2.271 4.328 2.271 4.328 mp_sum_l 7804 13.0 2.982 4.170 2.982 4.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=198.564000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1068.272727, yerr=11.249977 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.461793E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108638608 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58193. MP_Allreduce 11005 1175. MP_Sync 86 MP_Alltoall 1700 18828160. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.022 0.128 182.459 182.503 qs_mol_dyn_low 1 2.0 0.003 0.003 182.030 182.037 qs_forces 11 3.9 0.009 0.011 181.814 181.821 qs_energies 11 4.9 0.001 0.001 174.053 174.062 scf_env_do_scf 11 5.9 0.001 0.001 155.091 155.100 velocity_verlet 10 3.0 0.002 0.002 118.898 118.905 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 118.066 118.067 qs_scf_new_mos 116 7.6 0.001 0.001 77.042 77.167 qs_scf_loop_do_ot 116 8.6 0.001 0.001 77.041 77.166 dbcsr_multiply_generic 2485 12.5 0.178 0.189 75.662 76.109 ot_scf_mini 116 9.6 0.003 0.004 72.673 72.780 multiply_cannon 2485 13.5 0.578 0.602 54.841 59.398 multiply_cannon_loop 2485 14.5 0.438 0.448 50.272 51.162 ot_mini 116 10.6 0.001 0.001 40.301 40.423 init_scf_loop 11 6.9 0.000 0.000 36.875 36.876 mp_waitall_1 124680 16.7 27.996 35.656 27.996 35.656 prepare_preconditioner 11 7.9 0.000 0.000 32.863 32.888 make_preconditioner 11 8.9 0.000 0.000 32.863 32.888 rebuild_ks_matrix 127 8.3 0.001 0.001 31.798 31.900 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 31.797 31.899 make_full_inverse_cholesky 11 9.9 0.000 0.000 30.546 30.803 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.946 29.042 multiply_cannon_multrec 9940 15.5 10.314 14.534 17.541 21.925 qs_ot_get_derivative 116 11.6 0.001 0.002 20.470 20.582 multiply_cannon_metrocomm3 9940 15.5 0.022 0.024 12.996 20.322 ot_diis_step 116 11.6 0.020 0.021 19.753 19.754 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.253 19.627 apply_single 127 13.6 0.001 0.001 19.253 19.627 qs_ot_get_p 127 10.4 0.001 0.001 19.318 19.481 make_m2s 4970 13.5 0.066 0.070 17.015 18.966 cp_fm_cholesky_invert 11 10.9 18.960 18.965 18.960 18.965 make_images 4970 14.5 2.245 2.602 16.714 18.667 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.335 17.387 calculate_rho_elec 127 8.7 0.257 0.267 17.334 17.387 sum_up_and_integrate 127 10.3 0.179 0.189 16.428 16.473 integrate_v_rspace 127 11.3 0.004 0.005 16.249 16.303 qs_ot_p2m_diag 82 11.4 0.489 0.495 14.845 14.862 cp_dbcsr_syevd 82 12.4 0.006 0.006 13.716 13.719 make_images_data 4970 15.5 0.051 0.059 10.510 13.065 hybrid_alltoall_any 5155 16.4 0.808 3.776 10.217 12.665 multiply_cannon_sync_h2d 9940 15.5 11.433 12.055 11.433 12.055 init_scf_run 11 5.9 0.000 0.001 11.966 11.966 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.965 11.966 cp_fm_diag_elpa 82 13.4 0.000 0.000 10.615 10.626 cp_fm_diag_elpa_base 82 14.4 10.319 10.452 10.604 10.614 pw_transfer 1535 11.6 0.085 0.093 10.373 10.431 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.152 10.216 density_rs2pw 127 9.7 0.005 0.006 8.059 8.929 fft_wrap_pw1pw2_140 519 13.2 0.492 0.511 8.838 8.900 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 8.536 8.615 wfi_extrapolate 11 7.9 0.001 0.001 8.367 8.367 fft3d_ps 1281 14.7 2.691 2.764 8.326 8.361 cp_fm_cholesky_decompose 22 10.9 8.046 8.136 8.046 8.136 grid_integrate_task_list 127 12.3 7.687 8.073 7.687 8.073 multiply_cannon_metrocomm1 9940 15.5 0.028 0.029 4.697 7.816 mp_allgather_i34 2485 14.5 2.782 7.762 2.782 7.762 mp_alltoall_d11v 2401 14.1 6.781 7.577 6.781 7.577 dbcsr_mm_accdrv_process 20590 16.1 2.622 3.415 6.862 7.495 rs_pw_transfer 1038 11.9 0.013 0.013 5.530 6.425 calculate_dm_sparse 127 9.5 0.001 0.001 6.114 6.236 grid_collocate_task_list 127 9.7 5.352 6.055 5.352 6.055 dbcsr_complete_redistribute 393 12.7 2.114 2.168 5.549 5.933 potential_pw2rs 127 12.3 0.026 0.027 5.443 5.481 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.369 5.390 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 5.323 5.326 multiply_cannon_metrocomm4 7455 15.5 0.024 0.027 1.857 5.274 mp_irecv_dv 28618 15.9 1.823 5.186 1.823 5.186 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.188 4.217 rs_scatter_matrices 138 9.7 1.497 1.861 3.901 4.024 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.642 3.973 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.568 3.869 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.713 3.814 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.751 3.763 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.720 3.743 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=182.503000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1369.909091, yerr=37.166589 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430454546432 0.0% 0.0% 100.0% flops 32 x 32 x 32 1975684956160 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992006770688 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613065416704 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239182565376 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239182565376 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.245913E+12 0.0% 0.0% 100.0% flops max/rank 11.787674E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806580192 0.0% 0.0% 100.0% number of processed stacks 1982496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3433.3 marketing flops 145.663816E+12 ------------------------------------------------------------------------------- # multiplications 2535 max memory usage/rank 2.970423E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101400 MPI messages size (bytes): total size 1.145171E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.293599E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45888 35634806784 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592540656 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4041 58592. MP_Allreduce 11099 1511. MP_Sync 88 MP_Alltoall 1724 36993632. MP_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 0.028 0.087 313.311 313.319 qs_mol_dyn_low 1 2.0 0.003 0.003 312.716 312.744 qs_forces 11 3.9 0.006 0.006 312.514 312.519 qs_energies 11 4.9 0.001 0.001 303.078 303.082 scf_env_do_scf 11 5.9 0.001 0.001 279.267 279.280 velocity_verlet 10 3.0 0.002 0.002 224.432 224.449 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 150.110 150.112 init_scf_loop 11 6.9 0.000 0.000 128.875 128.878 prepare_preconditioner 11 7.9 0.000 0.000 123.763 123.801 make_preconditioner 11 8.9 0.000 0.000 123.763 123.801 make_full_inverse_cholesky 11 9.9 0.000 0.000 99.217 120.695 qs_scf_new_mos 118 7.6 0.001 0.001 97.246 97.400 qs_scf_loop_do_ot 118 8.6 0.001 0.001 97.245 97.399 ot_scf_mini 118 9.6 0.004 0.004 92.187 92.274 dbcsr_multiply_generic 2535 12.6 0.214 0.227 88.043 88.877 cp_fm_upper_to_full 106 14.8 53.375 76.685 53.375 76.685 multiply_cannon 2535 13.6 0.715 0.756 62.047 63.021 multiply_cannon_loop 2535 14.6 0.474 0.484 58.009 60.130 ot_mini 118 10.6 0.001 0.001 47.661 47.783 dbcsr_complete_redistribute 397 12.7 3.973 4.019 30.307 43.614 copy_fm_to_dbcsr 210 11.7 0.002 0.002 26.852 40.147 rebuild_ks_matrix 129 8.3 0.001 0.001 39.668 39.810 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 39.667 39.810 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 24.501 37.574 qs_ks_update_qs_env 129 7.6 0.001 0.001 36.567 36.695 mp_waitall_1 104820 16.8 31.255 36.231 31.255 36.231 cp_fm_cholesky_invert 11 10.9 35.747 35.753 35.747 35.753 mp_alltoall_i22 720 14.1 22.156 35.701 22.156 35.701 qs_ot_get_p 129 10.4 0.001 0.001 28.617 28.743 qs_ot_get_derivative 118 11.6 0.002 0.002 25.907 25.996 qs_ot_p2m_diag 84 11.4 0.889 0.894 24.421 24.453 multiply_cannon_metrocomm3 10140 15.6 0.023 0.024 20.914 22.732 cp_dbcsr_syevd 84 12.4 0.007 0.007 22.605 22.609 make_m2s 5070 13.6 0.078 0.080 20.856 22.058 ot_diis_step 118 11.6 0.022 0.023 21.691 21.693 make_images 5070 14.6 3.769 3.926 20.376 21.583 qs_rho_update_rho_low 129 7.7 0.001 0.001 21.526 21.547 calculate_rho_elec 129 8.7 0.486 0.487 21.526 21.546 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.692 21.123 apply_single 129 13.6 0.001 0.001 20.691 21.123 sum_up_and_integrate 129 10.3 0.325 0.327 20.453 20.535 integrate_v_rspace 129 11.3 0.004 0.004 20.128 20.208 cp_fm_diag_elpa 84 13.4 0.000 0.000 19.306 19.306 cp_fm_diag_elpa_base 84 14.4 14.830 16.493 19.302 19.302 multiply_cannon_multrec 10140 15.6 10.569 12.377 18.630 18.886 multiply_cannon_sync_h2d 10140 15.6 15.784 15.789 15.784 15.789 make_images_data 5070 15.6 0.060 0.065 11.374 13.440 hybrid_alltoall_any 5257 16.5 1.312 3.048 11.382 13.413 init_scf_run 11 5.9 0.004 0.030 13.193 13.193 scf_env_initial_rho_setup 11 6.9 0.001 0.001 13.189 13.193 pw_transfer 1559 11.6 0.094 0.095 12.811 12.817 fft_wrap_pw1pw2 1301 12.7 0.012 0.012 12.575 12.580 fft_wrap_pw1pw2_140 527 13.2 0.548 0.551 11.132 11.140 fft3d_ps 1301 14.7 2.773 2.780 10.588 10.594 mp_alltoall_d11v 2429 14.1 9.308 10.362 9.308 10.362 qs_ot_get_derivative_diag 78 12.4 0.002 0.003 10.227 10.288 cp_fm_cholesky_decompose 22 10.9 9.924 9.940 9.924 9.940 dbcsr_mm_accdrv_process 20958 16.1 3.827 5.778 7.820 9.724 wfi_extrapolate 11 7.9 0.001 0.001 9.295 9.295 density_rs2pw 129 9.7 0.005 0.006 8.730 8.853 grid_integrate_task_list 129 12.3 8.634 8.825 8.634 8.825 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.344 8.345 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 7.074 7.121 calculate_dm_sparse 129 9.5 0.001 0.001 6.949 7.043 rs_scatter_matrices 140 9.7 3.634 4.573 6.362 6.584 grid_collocate_task_list 129 9.7 6.413 6.461 6.413 6.461 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.177 6.281 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=313.319000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2633.727273, yerr=177.362541 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 1.262121E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255669. MP_Allreduce 3059 6274. MP_Sync 4 MP_Alltoall 54 6805335. 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 0.036 0.064 86.013 86.015 qs_energies 1 2.0 0.000 0.000 85.474 85.481 ls_scf 1 3.0 0.000 0.000 84.554 84.561 dbcsr_multiply_generic 111 6.7 0.014 0.015 73.361 73.562 multiply_cannon 111 7.7 0.017 0.020 56.111 57.078 multiply_cannon_loop 111 8.7 0.199 0.217 52.632 53.922 ls_scf_main 1 4.0 0.000 0.000 52.827 52.827 density_matrix_trs4 2 5.0 0.002 0.004 47.222 47.296 ls_scf_init_scf 1 4.0 0.000 0.000 28.647 28.648 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.538 27.585 mp_waitall_1 11031 10.9 23.081 25.504 23.081 25.504 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.444 25.459 multiply_cannon_multrec 2664 9.7 8.148 8.899 15.451 17.212 multiply_cannon_sync_h2d 2664 9.7 13.563 15.695 13.563 15.695 make_m2s 222 7.7 0.008 0.010 13.448 13.801 make_images 222 8.7 0.098 0.108 13.426 13.780 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.882 12.419 make_images_data 222 9.7 0.004 0.005 7.966 8.498 hybrid_alltoall_any 227 10.6 0.215 1.835 6.848 8.255 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.563 7.992 dbcsr_mm_accdrv_process 4760 10.4 0.512 0.632 6.924 7.904 dbcsr_mm_accdrv_process_sort 4760 11.4 6.212 7.108 6.212 7.108 calculate_norms 4752 9.8 5.522 6.343 5.522 6.343 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.035 5.188 mp_sum_l 807 5.4 3.262 4.640 3.262 4.640 make_images_sizes 222 9.7 0.000 0.000 0.746 3.783 mp_alltoall_i44 222 10.7 0.746 3.782 0.746 3.782 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.456 3.676 arnoldi_extremal 4 6.8 0.000 0.000 3.366 3.390 arnoldi_normal_ev 4 7.8 0.001 0.002 3.366 3.390 build_subspace 16 8.4 0.009 0.012 3.267 3.269 multiply_cannon_metrocomm4 2442 9.7 0.012 0.014 2.080 3.145 mp_irecv_dv 6231 10.9 2.063 3.129 2.063 3.129 ls_scf_post 1 4.0 0.000 0.000 3.080 3.087 ls_scf_store_result 1 5.0 0.000 0.000 2.899 2.937 dbcsr_special_finalize 555 9.7 0.005 0.006 2.385 2.757 dbcsr_merge_single_wm 555 10.7 0.487 0.636 2.376 2.748 make_images_pack 222 9.7 2.206 2.633 2.207 2.634 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.344 2.624 dbcsr_matrix_vector_mult_local 304 10.0 2.062 2.445 2.065 2.447 dbcsr_sort_data 658 11.4 2.148 2.444 2.148 2.444 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.364 2.438 buffer_matrices_ensure_size 222 8.7 1.821 2.194 1.821 2.194 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.793 1.795 rebuild_ks_matrix 3 7.3 0.000 0.000 1.784 1.785 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.784 1.785 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=86.015000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1142.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 2.102481E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 10339. MP_Sync 4 MP_Alltoall 47 15335933. MP_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 0.066 0.087 94.167 94.168 qs_energies 1 2.0 0.000 0.000 93.592 93.599 ls_scf 1 3.0 0.000 0.000 92.273 92.281 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.332 77.712 multiply_cannon 111 7.7 0.028 0.042 54.280 58.104 ls_scf_main 1 4.0 0.000 0.000 55.934 55.939 multiply_cannon_loop 111 8.7 0.115 0.122 50.980 54.354 density_matrix_trs4 2 5.0 0.002 0.004 50.054 50.254 ls_scf_init_scf 1 4.0 0.000 0.000 32.463 32.464 mp_waitall_1 9105 10.9 22.723 32.163 22.723 32.163 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.256 31.355 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.024 28.041 multiply_cannon_multrec 1332 9.7 13.003 16.602 22.344 26.869 multiply_cannon_metrocomm3 1332 9.7 0.006 0.008 12.723 22.189 make_m2s 222 7.7 0.006 0.007 15.766 16.388 make_images 222 8.7 1.572 1.963 15.736 16.360 dbcsr_mm_accdrv_process 4041 10.4 0.329 1.120 8.940 10.464 make_images_data 222 9.7 0.004 0.004 9.220 10.142 dbcsr_mm_accdrv_process_sort 4041 11.4 8.353 9.877 8.353 9.877 hybrid_alltoall_any 227 10.6 0.521 2.471 8.649 9.437 mp_sum_l 807 5.4 6.047 9.141 6.047 9.141 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.215 7.729 mp_irecv_dv 3311 11.0 3.196 7.682 3.196 7.682 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.371 7.154 calculate_norms 2376 9.8 5.981 6.811 5.981 6.811 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.806 6.053 multiply_cannon_sync_h2d 1332 9.7 4.732 6.013 4.732 6.013 arnoldi_extremal 4 6.8 0.000 0.000 5.293 5.323 arnoldi_normal_ev 4 7.8 0.001 0.005 5.293 5.323 build_subspace 16 8.4 0.014 0.020 4.941 4.943 ls_scf_post 1 4.0 0.000 0.000 3.877 3.885 ls_scf_store_result 1 5.0 0.000 0.000 3.525 3.642 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.172 3.428 dbcsr_matrix_vector_mult_local 304 10.0 2.733 3.208 2.735 3.209 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.408 2.883 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.725 2.835 make_images_pack 222 9.7 2.019 2.436 2.021 2.438 mp_allgather_i34 111 8.7 0.995 2.362 0.995 2.362 dbcsr_sort_data 436 11.2 1.826 2.042 1.826 2.042 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.897 1.899 rebuild_ks_matrix 3 7.3 0.000 0.000 1.884 1.886 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.884 1.886 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=94.168000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1748.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 2.704806E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265470. MP_Allreduce 3058 11181. MP_Sync 4 MP_Alltoall 47 23526250. MP_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 0.047 0.072 96.009 96.011 qs_energies 1 2.0 0.000 0.000 95.358 95.367 ls_scf 1 3.0 0.000 0.001 93.927 93.936 dbcsr_multiply_generic 111 6.7 0.016 0.017 78.329 78.568 ls_scf_main 1 4.0 0.001 0.028 58.437 58.441 multiply_cannon 111 7.7 0.048 0.109 53.740 57.948 multiply_cannon_loop 111 8.7 0.099 0.105 50.144 54.319 density_matrix_trs4 2 5.0 0.004 0.034 52.369 52.556 mp_waitall_1 7281 11.0 25.366 35.381 25.366 35.381 ls_scf_init_scf 1 4.0 0.000 0.001 31.506 31.508 ls_scf_init_matrix_S 1 5.0 0.000 0.003 30.075 30.131 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.619 27.642 multiply_cannon_multrec 888 9.7 12.506 15.091 21.028 24.207 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.660 23.669 make_m2s 222 7.7 0.006 0.007 17.538 18.797 make_images 222 8.7 1.966 2.280 17.500 18.758 make_images_data 222 9.7 0.003 0.004 10.258 11.278 hybrid_alltoall_any 227 10.6 0.621 2.858 9.877 11.121 dbcsr_mm_accdrv_process 3754 10.4 0.261 0.475 8.040 9.171 mp_sum_l 807 5.4 5.498 8.965 5.498 8.965 dbcsr_mm_accdrv_process_sort 3754 11.4 7.639 8.759 7.639 8.759 multiply_cannon_sync_h2d 888 9.7 6.051 7.530 6.051 7.530 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.459 7.502 mp_irecv_dv 2335 11.1 2.443 7.442 2.443 7.442 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 4.071 7.389 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.260 7.191 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.464 5.634 arnoldi_extremal 4 6.8 0.000 0.000 5.094 5.127 arnoldi_normal_ev 4 7.8 0.001 0.004 5.094 5.127 build_subspace 16 8.4 0.014 0.020 4.779 4.787 calculate_norms 1584 9.8 4.259 4.709 4.259 4.709 ls_scf_post 1 4.0 0.001 0.012 3.984 3.993 mp_allgather_i34 111 8.7 1.432 3.904 1.432 3.904 ls_scf_store_result 1 5.0 0.000 0.000 3.721 3.810 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.434 3.751 dbcsr_matrix_vector_mult_local 304 10.0 3.006 3.576 3.008 3.578 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.914 3.019 make_images_sizes 222 9.7 0.000 0.000 0.969 2.208 mp_alltoall_i44 222 10.7 0.969 2.208 0.969 2.208 make_images_pack 222 9.7 1.810 2.128 1.813 2.131 dbcsr_sort_data 325 11.1 1.894 2.106 1.894 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.029 2.030 rebuild_ks_matrix 3 7.3 0.000 0.000 2.008 2.009 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.008 2.009 dbcsr_data_release 9322 10.9 1.312 1.990 1.312 1.990 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=96.011000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2215.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 3.319828E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266696. MP_Allreduce 3058 13371. MP_Sync 4 MP_Alltoall 47 30278988. MP_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 0.035 0.046 98.929 98.931 qs_energies 1 2.0 0.000 0.000 98.267 98.281 ls_scf 1 3.0 0.000 0.001 96.548 96.566 dbcsr_multiply_generic 111 6.7 0.017 0.019 79.499 79.741 ls_scf_main 1 4.0 0.000 0.002 59.922 59.923 multiply_cannon 111 7.7 0.049 0.089 52.253 56.672 density_matrix_trs4 2 5.0 0.003 0.006 53.617 53.738 multiply_cannon_loop 111 8.7 0.114 0.126 47.144 49.632 ls_scf_init_scf 1 4.0 0.000 0.001 33.346 33.348 ls_scf_init_matrix_S 1 5.0 0.000 0.001 32.114 32.180 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.635 29.653 mp_waitall_1 6369 11.0 23.565 29.526 23.565 29.526 multiply_cannon_multrec 1332 9.7 14.193 17.090 22.015 25.072 make_m2s 222 7.7 0.007 0.008 21.340 22.760 make_images 222 8.7 3.126 3.596 21.290 22.712 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.533 17.498 make_images_data 222 9.7 0.004 0.004 12.011 13.607 hybrid_alltoall_any 227 10.6 0.799 3.818 11.302 13.068 dbcsr_mm_accdrv_process 3641 10.4 0.229 0.414 7.451 8.950 dbcsr_mm_accdrv_process_sort 3641 11.4 7.076 8.531 7.076 8.531 mp_sum_l 807 5.4 4.314 8.363 4.314 8.363 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.326 6.585 multiply_cannon_sync_h2d 1332 9.7 5.550 6.296 5.550 6.296 arnoldi_extremal 4 6.8 0.000 0.000 5.899 5.919 arnoldi_normal_ev 4 7.8 0.001 0.004 5.899 5.919 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.078 5.916 mp_irecv_dv 3229 10.9 2.056 5.843 2.056 5.843 build_subspace 16 8.4 0.014 0.020 5.553 5.559 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.844 5.520 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.609 4.796 mp_allgather_i34 111 8.7 2.223 4.599 2.223 4.599 calculate_norms 2376 9.8 4.185 4.534 4.185 4.534 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.689 4.022 dbcsr_matrix_vector_mult_local 304 10.0 3.167 3.662 3.169 3.664 dbcsr_sort_data 658 11.4 3.038 3.477 3.038 3.477 ls_scf_post 1 4.0 0.000 0.002 3.281 3.298 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.153 3.217 dbcsr_special_finalize 555 9.7 0.006 0.009 2.780 3.177 dbcsr_merge_single_wm 555 10.7 0.536 0.662 2.772 3.168 ls_scf_store_result 1 5.0 0.000 0.000 3.013 3.079 dbcsr_data_release 10477 10.7 1.600 2.476 1.600 2.476 dbcsr_finalize 304 7.8 0.050 0.061 1.804 2.062 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.991 1.993 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.931000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 4.627395E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265558. MP_Allreduce 3049 15663. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.040 0.059 100.617 100.621 qs_energies 1 2.0 0.000 0.000 99.826 99.832 ls_scf 1 3.0 0.000 0.000 97.844 97.854 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.688 78.896 ls_scf_main 1 4.0 0.000 0.000 63.004 63.005 multiply_cannon 111 7.7 0.098 0.216 55.778 61.258 density_matrix_trs4 2 5.0 0.002 0.003 55.417 55.518 multiply_cannon_loop 111 8.7 0.070 0.087 51.219 52.792 mp_waitall_1 5436 11.0 27.103 32.854 27.103 32.854 ls_scf_init_scf 1 4.0 0.000 0.000 31.215 31.220 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.810 29.853 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.511 27.522 multiply_cannon_multrec 444 9.7 14.111 16.245 21.132 23.133 make_m2s 222 7.7 0.004 0.005 17.972 20.523 make_images 222 8.7 3.709 4.398 17.910 20.463 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.422 16.668 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.289 15.263 make_images_data 222 9.7 0.003 0.004 10.173 12.680 hybrid_alltoall_any 227 10.6 0.790 3.767 9.964 12.331 dbcsr_mm_accdrv_process 3003 10.4 0.190 0.344 6.717 7.862 multiply_cannon_sync_h2d 444 9.7 6.502 7.702 6.502 7.702 dbcsr_mm_accdrv_process_sort 3003 11.4 6.381 7.484 6.381 7.484 mp_allgather_i34 111 8.7 2.796 7.105 2.796 7.105 arnoldi_extremal 4 6.8 0.000 0.000 6.017 6.040 arnoldi_normal_ev 4 7.8 0.001 0.005 6.017 6.039 build_subspace 16 8.4 0.015 0.020 5.574 5.582 mp_sum_l 807 5.4 2.979 5.434 2.979 5.434 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.853 4.993 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.267 4.424 dbcsr_matrix_vector_mult_local 304 10.0 3.656 4.127 3.658 4.130 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.970 4.014 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.861 3.962 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.649 3.768 mp_irecv_dv 1241 11.2 1.630 3.749 1.630 3.749 make_images_sizes 222 9.7 0.000 0.000 1.149 3.743 mp_alltoall_i44 222 10.7 1.149 3.743 1.149 3.743 calculate_norms 792 9.8 3.588 3.715 3.588 3.715 ls_scf_post 1 4.0 0.000 0.000 3.625 3.630 ls_scf_store_result 1 5.0 0.000 0.000 3.406 3.446 dbcsr_finalize 304 7.8 0.062 0.078 2.199 2.258 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.201 2.202 rebuild_ks_matrix 3 7.3 0.000 0.000 2.169 2.171 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.169 2.171 dbcsr_merge_all 275 8.9 0.473 0.527 2.044 2.096 dbcsr_data_release 10123 10.8 1.347 2.038 1.347 2.038 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=100.621000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3641.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/60e7f91d1367dd062fe2fa4125183e5e3d99f205_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 8.728240E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284111. MP_Allreduce 3043 21950. MP_Sync 4 MP_Alltoall 47 88727262. MP_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 0.098 0.136 110.369 110.383 qs_energies 1 2.0 0.000 0.000 108.899 108.907 ls_scf 1 3.0 0.000 0.000 105.950 105.958 dbcsr_multiply_generic 111 6.7 0.024 0.027 78.819 78.909 ls_scf_main 1 4.0 0.000 0.000 67.166 67.166 density_matrix_trs4 2 5.0 0.002 0.005 57.767 57.813 multiply_cannon 111 7.7 0.122 0.216 50.620 52.299 multiply_cannon_loop 111 8.7 0.068 0.070 47.239 48.254 ls_scf_init_scf 1 4.0 0.000 0.000 35.091 35.092 ls_scf_init_matrix_S 1 5.0 0.000 0.000 33.369 33.385 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 30.528 30.536 mp_waitall_1 4527 11.1 23.663 26.955 23.663 26.955 make_m2s 222 7.7 0.005 0.005 24.512 25.376 make_images 222 8.7 4.577 4.969 24.405 25.267 multiply_cannon_multrec 444 9.7 17.803 18.558 22.385 23.206 hybrid_alltoall_any 227 10.6 1.660 3.621 13.383 16.302 make_images_data 222 9.7 0.003 0.003 13.679 16.018 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 11.352 12.155 multiply_cannon_sync_h2d 444 9.7 8.839 8.915 8.839 8.915 arnoldi_extremal 4 6.8 0.000 0.000 7.713 7.726 arnoldi_normal_ev 4 7.8 0.002 0.008 7.713 7.726 build_subspace 16 8.4 0.026 0.037 7.150 7.161 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.724 5.883 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.351 5.444 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.040 5.291 dbcsr_matrix_vector_mult_local 304 10.0 4.964 5.267 4.966 5.270 dbcsr_mm_accdrv_process 1814 10.4 0.209 0.322 4.409 4.528 dbcsr_mm_accdrv_process_sort 1814 11.4 4.109 4.241 4.109 4.241 ls_scf_post 1 4.0 0.000 0.000 3.693 3.701 make_images_sizes 222 9.7 0.000 0.000 1.500 3.688 mp_alltoall_i44 222 10.7 1.499 3.687 1.499 3.687 ls_scf_store_result 1 5.0 0.000 0.000 3.432 3.445 calculate_norms 792 9.8 3.254 3.294 3.254 3.294 dbcsr_finalize 304 7.8 0.082 0.089 3.086 3.169 dbcsr_merge_all 275 8.9 0.892 0.922 2.873 2.950 dbcsr_complete_redistribute 5 7.6 1.434 1.464 2.837 2.942 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.918 2.918 dbcsr_data_release 12724 10.6 2.339 2.868 2.339 2.868 mp_allgather_i34 111 8.7 0.991 2.848 0.991 2.848 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.646 2.648 matrix_ls_to_qs 2 6.0 0.000 0.000 2.497 2.611 rebuild_ks_matrix 3 7.3 0.000 0.000 2.580 2.582 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.580 2.582 dbcsr_sort_data 325 11.1 2.439 2.499 2.439 2.499 dbcsr_new_transposed 4 7.5 0.246 0.265 2.315 2.325 dbcsr_frobenius_norm 74 6.6 2.055 2.131 2.206 2.246 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=110.383000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6852.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 60e7f91d1367dd062fe2fa4125183e5e3d99f205 Summary: empty Status: OK