=== 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: 15b8176f3f1ac853f85861d42f1522ebe17008b3 ################# 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 7.3.2, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Author: Matthias Krack (12.01.2023) # # \ 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=${this_file##*.} libcp2k"; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.2 USE_ELPA := 2022.11.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.1.0 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.1 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.3.2 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.4 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/01 job id: 44216479 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/02 job id: 44216480 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/03 job id: 44216481 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/04 job id: 44216483 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/05 job id: 44216485 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/06 job id: 44216487 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/07 job id: 44216488 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/08 job id: 44216489 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/09 job id: 44216491 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/10 job id: 44216492 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/11 job id: 44216494 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/12 job id: 44216496 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/13 job id: 44216498 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/14 job id: 44216503 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/15 job id: 44216507 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/16 job id: 44216510 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/17 job id: 44216515 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/18 job id: 44216518 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/19 job id: 44216533 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/20 job id: 44216544 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/21 job id: 44216548 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/22 job id: 44216555 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/23 job id: 44216558 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/24 job id: 44216559 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/25 job id: 44216561 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/26 job id: 44216562 --- 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/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 0.034 134.894 134.895 farming_run 1 2.0 134.338 134.340 134.864 134.867 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.452696E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 228 1113141. MP_Allreduce 485 2282278. MP_Sync 27 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 8 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.026 116.204 116.204 qs_energies 1 2.0 0.000 0.000 115.982 115.982 mp2_main 1 3.0 0.000 0.000 114.035 114.035 mp2_gpw_main 1 4.0 0.021 0.027 113.205 113.206 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 94.361 94.806 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 55.490 55.937 mp2_eri_3c_integrate_gpw 272 7.0 0.153 0.172 41.806 47.223 get_2c_integrals 1 6.0 0.000 0.001 38.050 38.698 integrate_v_rspace 273 8.0 0.436 0.454 25.161 30.299 pw_transfer 6555 10.6 0.373 0.389 27.485 28.448 fft_wrap_pw1pw2 5465 11.4 0.044 0.047 26.157 26.992 grid_integrate_task_list 273 9.0 20.961 26.596 20.961 26.596 fft_wrap_pw1pw2_100 2178 12.4 1.229 1.469 23.678 24.516 compute_2c_integrals 1 7.0 0.142 0.227 20.098 20.099 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.953 19.505 mp2_eri_2c_integrate_gpw 1 9.0 2.380 2.463 18.950 19.504 rpa_ri_compute_en 1 5.0 0.004 0.006 18.736 18.954 cp_fm_cholesky_decompose 12 8.2 17.928 18.545 17.928 18.545 cholesky_decomp 1 7.0 0.000 0.000 16.798 17.417 fft3d_s 5443 13.4 16.206 16.797 16.227 16.818 ao_to_mo_and_store_B_mult_1 272 7.0 10.855 15.581 10.855 15.581 calculate_wavefunction 272 8.0 5.426 5.574 12.573 13.172 rpa_num_int 1 6.0 0.000 0.001 10.584 10.584 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.533 10.583 calc_mat_Q 8 8.0 0.000 0.000 9.348 9.463 contract_S_to_Q 8 9.0 0.000 0.000 8.770 8.882 calc_potential_gpw 544 9.5 0.005 0.006 8.231 8.671 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.246 8.625 potential_pw2rs 545 10.0 0.107 0.108 7.748 8.496 parallel_gemm_fm 14 9.1 0.000 0.000 8.359 8.449 parallel_gemm_fm_cosma 14 10.1 8.359 8.449 8.359 8.449 collocate_single_gaussian 272 10.0 0.041 0.043 7.508 7.801 create_integ_mat 1 6.0 0.006 0.008 7.606 7.606 array2fm 1 7.0 0.000 0.000 6.712 7.108 pw_scatter_s 2720 13.7 4.413 4.646 4.413 4.646 pw_gather_s 2722 13.2 3.897 4.263 3.897 4.263 array2fm_buffer_send 1 8.0 3.044 3.171 3.044 3.171 pw_poisson_solve 545 10.5 1.124 1.185 2.159 2.397 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=113.205605, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2724.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 0.026 0.038 395.845 395.845 farming_run 1 2.0 394.994 394.997 395.812 395.814 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.223487E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 703 408373. MP_Allreduce 1821 23730. MP_Sync 38 MP_Alltoall 77 2368424. 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 0.046 209.128 209.128 qs_energies 1 2.0 0.000 0.000 208.879 208.893 scf_env_do_scf 1 3.0 0.000 0.000 105.679 105.679 qs_ks_update_qs_env 5 5.0 0.000 0.000 104.807 104.814 rebuild_ks_matrix 4 6.0 0.000 0.000 104.806 104.813 qs_ks_build_kohn_sham_matrix 4 7.0 0.057 0.063 104.806 104.813 hfx_ks_matrix 4 8.0 0.001 0.001 104.428 104.432 integrate_four_center 4 9.0 0.145 0.458 104.428 104.431 mp2_main 1 3.0 0.000 0.000 102.909 102.923 mp2_gpw_main 1 4.0 0.036 0.059 102.053 102.067 integrate_four_center_main 4 10.0 0.124 0.553 96.606 99.033 integrate_four_center_bin 268 11.0 96.483 99.016 96.483 99.016 init_scf_loop 1 4.0 0.000 0.000 91.531 91.531 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 75.132 76.631 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 54.662 56.160 mp2_eri_3c_integrate_gpw 91 7.0 0.146 0.165 42.395 47.262 integrate_v_rspace 95 8.0 0.398 0.569 28.746 33.452 pw_transfer 2240 10.6 0.145 0.162 29.909 30.405 ao_to_mo_and_store_B_mult_1 91 7.0 10.581 30.077 10.581 30.077 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.920 29.437 grid_integrate_task_list 95 9.0 24.043 28.945 24.043 28.945 mp2_ri_gpw_compute_en 1 5.0 0.061 0.074 26.766 28.675 fft_wrap_pw1pw2_100 730 12.4 1.272 1.480 26.648 27.129 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.843 1.891 24.614 24.625 get_2c_integrals 1 6.0 0.000 0.000 20.381 20.405 compute_2c_integrals 1 7.0 0.003 0.003 19.364 19.366 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.846 19.225 mp2_eri_2c_integrate_gpw 1 9.0 1.744 1.859 18.845 19.224 fft3d_s 1823 13.4 18.429 18.832 18.442 18.845 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 14.146 14.146 calculate_wavefunction 91 8.0 2.024 2.059 9.756 10.001 mp2_ri_gpw_compute_en_expansio 172 7.0 0.554 0.582 8.770 9.609 potential_pw2rs 186 10.0 0.034 0.035 8.620 9.243 local_gemm 172 8.0 8.216 9.048 8.216 9.048 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.222 8.592 collocate_single_gaussian 91 10.0 0.017 0.021 7.878 8.162 calc_potential_gpw 182 9.5 0.002 0.002 7.903 8.142 mp2_ri_gpw_compute_en_comm 22 7.0 0.516 0.539 7.611 7.942 mp2_ri_gpw_compute_en_ener 172 7.0 6.345 6.440 6.345 6.440 mp_sendrecv_dm3 2068 8.0 5.601 5.935 5.601 5.935 pw_gather_s 912 13.2 4.910 5.415 4.910 5.415 mp_sync 38 10.4 2.511 5.104 2.511 5.104 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=102.066133, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1511.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 452.161536E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62385. MP_Allreduce 10249 271. MP_Sync 580 MP_Alltoall 2083 405530. 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 0.031 51.775 51.775 qs_mol_dyn_low 1 2.0 0.003 0.005 51.555 51.563 qs_forces 11 3.9 0.002 0.002 51.492 51.497 qs_energies 11 4.9 0.001 0.001 50.030 50.048 scf_env_do_scf 11 5.9 0.000 0.001 44.290 44.290 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 42.131 42.131 dbcsr_multiply_generic 2286 12.5 0.093 0.097 33.157 33.507 qs_scf_new_mos 108 7.5 0.000 0.001 31.919 32.213 qs_scf_loop_do_ot 108 8.5 0.000 0.001 31.919 32.212 ot_scf_mini 108 9.5 0.002 0.002 30.279 30.451 multiply_cannon 2286 13.5 0.187 0.196 25.695 27.143 multiply_cannon_loop 2286 14.5 1.478 1.558 25.009 26.444 velocity_verlet 10 3.0 0.001 0.001 25.144 25.145 ot_mini 108 10.5 0.001 0.001 19.201 19.461 qs_ot_get_derivative 108 11.5 0.001 0.001 16.250 16.449 mp_waitall_1 245248 16.5 8.383 14.207 8.383 14.207 multiply_cannon_metrocomm3 54864 15.5 0.068 0.073 5.950 12.586 multiply_cannon_multrec 54864 15.5 4.230 6.518 7.474 10.825 rebuild_ks_matrix 119 8.3 0.000 0.000 7.996 8.137 qs_ks_build_kohn_sham_matrix 119 9.3 0.031 0.049 7.996 8.137 multiply_cannon_sync_h2d 54864 15.5 5.907 7.324 5.907 7.324 mp_sum_l 7207 12.9 5.521 7.244 5.521 7.244 qs_ks_update_qs_env 119 7.6 0.001 0.001 7.064 7.196 qs_ot_get_p 119 10.4 0.001 0.001 6.496 6.808 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.873 5.986 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.224 5.688 sum_up_and_integrate 119 10.3 0.012 0.014 4.521 4.528 init_scf_run 11 5.9 0.000 0.001 4.525 4.525 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.524 4.525 integrate_v_rspace 119 11.3 0.002 0.002 4.509 4.517 dbcsr_mm_accdrv_process 76910 16.1 1.162 1.827 3.166 4.438 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.251 4.362 calculate_rho_elec 119 8.7 0.011 0.017 4.250 4.362 qs_ot_p2m_diag 50 11.0 0.004 0.006 3.704 3.772 multiply_cannon_metrocomm1 54864 15.5 0.051 0.057 1.779 3.298 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.859 3.083 apply_single 119 13.6 0.000 0.000 2.859 3.083 calculate_dm_sparse 119 9.5 0.000 0.001 2.856 3.002 rs_pw_transfer 974 11.9 0.012 0.013 2.748 2.906 cp_dbcsr_syevd 50 12.0 0.002 0.003 2.855 2.856 ot_diis_step 108 11.5 0.006 0.006 2.675 2.675 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.651 2.651 cp_fm_redistribute_end 50 14.0 2.413 2.629 2.418 2.630 cp_fm_diag_elpa_base 50 14.0 0.211 2.542 0.211 2.550 jit_kernel_multiply 13 15.8 1.943 2.532 1.943 2.532 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.427 2.504 density_rs2pw 119 9.7 0.004 0.004 2.261 2.456 calculate_first_density_matrix 1 7.0 0.000 0.000 2.405 2.413 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.148 2.150 grid_integrate_task_list 119 12.3 2.031 2.146 2.031 2.146 acc_transpose_blocks 54864 15.5 0.228 0.250 1.712 2.144 init_scf_loop 11 6.9 0.000 0.000 2.140 2.141 wfi_extrapolate 11 7.9 0.001 0.001 2.061 2.061 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.920 1.963 mp_sum_d 4125 12.0 1.310 1.941 1.310 1.941 pw_transfer 1439 11.6 0.053 0.059 1.813 1.893 potential_pw2rs 119 12.3 0.004 0.004 1.869 1.879 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.736 1.821 make_m2s 4572 13.5 0.054 0.056 1.628 1.687 make_images 4572 14.5 0.133 0.139 1.546 1.603 fft3d_ps 1201 14.6 0.371 0.475 1.506 1.593 mp_alltoall_d11v 2130 13.8 1.270 1.448 1.270 1.448 mp_waitany 12084 13.8 1.238 1.433 1.238 1.433 fft_wrap_pw1pw2_140 487 13.2 0.082 0.095 1.342 1.427 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.377 1.426 grid_collocate_task_list 119 9.7 1.288 1.371 1.288 1.371 dbcsr_dot_sd 1205 11.9 0.051 0.061 0.716 1.134 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=51.775000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.454545, yerr=0.782030 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 489.582592E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62664. MP_Allreduce 10226 305. MP_Sync 104 MP_Alltoall 2060 1030452. 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 0.049 0.051 39.677 39.677 qs_mol_dyn_low 1 2.0 0.003 0.004 39.375 39.382 qs_forces 11 3.9 0.005 0.010 38.636 38.643 qs_energies 11 4.9 0.001 0.001 36.956 36.965 scf_env_do_scf 11 5.9 0.000 0.001 31.297 31.297 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 28.862 28.862 dbcsr_multiply_generic 2286 12.5 0.099 0.102 21.655 22.022 qs_scf_new_mos 108 7.5 0.001 0.001 19.702 19.946 qs_scf_loop_do_ot 108 8.5 0.001 0.001 19.701 19.945 ot_scf_mini 108 9.5 0.002 0.003 18.825 18.998 velocity_verlet 10 3.0 0.001 0.001 18.739 18.754 multiply_cannon 2286 13.5 0.207 0.218 16.786 18.329 multiply_cannon_loop 2286 14.5 0.894 0.961 15.701 17.033 ot_mini 108 10.5 0.001 0.001 11.660 11.895 mp_waitall_1 200699 16.5 5.780 11.210 5.780 11.210 multiply_cannon_metrocomm3 27432 15.5 0.068 0.070 4.306 9.862 qs_ot_get_derivative 108 11.5 0.001 0.001 9.213 9.389 multiply_cannon_multrec 27432 15.5 1.982 4.591 6.143 9.137 rebuild_ks_matrix 119 8.3 0.000 0.000 7.177 7.331 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 7.177 7.331 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.341 6.479 dbcsr_mm_accdrv_process 47894 16.0 3.207 5.609 4.092 6.140 qs_ot_get_p 119 10.4 0.001 0.001 4.308 4.537 init_scf_run 11 5.9 0.000 0.001 4.449 4.449 scf_env_initial_rho_setup 11 6.9 0.000 0.000 4.448 4.449 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.586 4.434 sum_up_and_integrate 119 10.3 0.024 0.027 4.174 4.181 integrate_v_rspace 119 11.3 0.002 0.002 4.150 4.158 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.042 4.119 apply_single 119 13.6 0.000 0.000 3.041 4.119 mp_sum_l 7207 12.9 2.006 3.958 2.006 3.958 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.798 3.847 calculate_rho_elec 119 8.7 0.021 0.024 3.798 3.846 calculate_first_density_matrix 1 7.0 0.000 0.000 2.931 2.932 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.896 2.915 rs_pw_transfer 974 11.9 0.010 0.011 2.542 2.885 multiply_cannon_sync_h2d 27432 15.5 2.195 2.881 2.195 2.881 make_m2s 4572 13.5 0.052 0.054 2.413 2.624 make_images 4572 14.5 0.200 0.239 2.324 2.532 density_rs2pw 119 9.7 0.004 0.004 2.126 2.492 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.466 2.466 init_scf_loop 11 6.9 0.000 0.000 2.414 2.414 ot_diis_step 108 11.5 0.010 0.011 2.399 2.400 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.351 2.353 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.100 2.191 calculate_dm_sparse 119 9.5 0.000 0.000 2.110 2.180 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.131 2.179 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.145 2.146 cp_fm_redistribute_end 50 14.0 1.777 2.123 1.780 2.124 jit_kernel_multiply 10 16.0 0.833 2.107 0.833 2.107 cp_fm_diag_elpa_base 50 14.0 0.329 2.017 0.342 2.067 pw_transfer 1439 11.6 0.065 0.072 1.976 2.009 grid_integrate_task_list 119 12.3 1.837 1.930 1.837 1.930 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.884 1.920 potential_pw2rs 119 12.3 0.006 0.006 1.900 1.908 fft3d_ps 1201 14.6 0.512 0.563 1.587 1.614 acc_transpose_blocks 27432 15.5 0.108 0.113 1.252 1.581 prepare_preconditioner 11 7.9 0.000 0.000 1.514 1.542 make_preconditioner 11 8.9 0.000 0.000 1.514 1.542 fft_wrap_pw1pw2_140 487 13.2 0.078 0.084 1.492 1.528 make_images_data 4572 15.5 0.046 0.053 1.119 1.526 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.420 1.477 hybrid_alltoall_any 4725 16.4 0.051 0.111 0.970 1.464 wfi_extrapolate 11 7.9 0.001 0.001 1.462 1.462 grid_collocate_task_list 119 9.7 1.229 1.379 1.229 1.379 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.337 1.345 mp_alltoall_d11v 2130 13.8 1.150 1.331 1.150 1.331 mp_allgather_i34 2286 14.5 0.525 1.274 0.525 1.274 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.148 1.191 mp_sum_d 4125 12.0 0.552 0.994 0.552 0.994 acc_transpose_blocks_kernels 27432 16.5 0.181 0.271 0.723 0.967 qs_energies_init_hamiltonians 11 5.9 0.007 0.008 0.952 0.959 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.893 0.908 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.532 0.876 mp_waitany 5720 13.7 0.512 0.859 0.512 0.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.677000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.636364, yerr=1.366663 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 522.973184E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62660. MP_Allreduce 10225 303. MP_Sync 104 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_comm_split 50 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.088 32.639 32.641 qs_mol_dyn_low 1 2.0 0.003 0.003 32.366 32.373 qs_forces 11 3.9 0.001 0.002 31.627 31.628 qs_energies 11 4.9 0.004 0.025 30.035 30.038 scf_env_do_scf 11 5.9 0.003 0.020 25.306 25.308 scf_env_do_scf_inner_loop 108 6.5 0.019 0.031 22.741 22.742 dbcsr_multiply_generic 2286 12.5 0.092 0.096 16.166 16.256 velocity_verlet 10 3.0 0.001 0.001 15.842 15.864 qs_scf_new_mos 108 7.5 0.001 0.001 14.534 14.563 qs_scf_loop_do_ot 108 8.5 0.001 0.001 14.534 14.563 ot_scf_mini 108 9.5 0.002 0.004 13.821 13.845 multiply_cannon 2286 13.5 0.196 0.202 12.989 13.742 multiply_cannon_loop 2286 14.5 0.632 0.666 12.227 12.998 ot_mini 108 10.5 0.001 0.001 8.568 8.593 qs_ot_get_derivative 108 11.5 0.001 0.001 7.088 7.114 multiply_cannon_multrec 18288 15.5 1.957 2.895 6.765 7.046 rebuild_ks_matrix 119 8.3 0.000 0.000 6.416 6.431 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.018 6.416 6.431 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.652 5.666 dbcsr_mm_accdrv_process 38222 16.0 4.052 5.275 4.724 5.613 sum_up_and_integrate 119 10.3 0.030 0.031 3.971 3.975 integrate_v_rspace 119 11.3 0.003 0.013 3.940 3.948 mp_waitall_1 158411 16.6 2.537 3.709 2.537 3.709 init_scf_run 11 5.9 0.000 0.001 3.522 3.522 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.522 3.522 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.466 3.471 calculate_rho_elec 119 8.7 0.030 0.031 3.465 3.471 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.653 3.268 qs_ot_get_p 119 10.4 0.001 0.001 3.112 3.133 init_scf_loop 11 6.9 0.001 0.006 2.545 2.550 rs_pw_transfer 974 11.9 0.009 0.010 2.236 2.488 multiply_cannon_metrocomm3 18288 15.5 0.044 0.046 1.445 2.367 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.021 2.332 apply_single 119 13.6 0.000 0.000 2.021 2.332 calculate_first_density_matrix 1 7.0 0.000 0.003 2.285 2.286 density_rs2pw 119 9.7 0.004 0.004 1.988 2.250 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.092 2.097 pw_transfer 1439 11.6 0.066 0.072 1.972 1.981 grid_integrate_task_list 119 12.3 1.804 1.904 1.804 1.904 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.878 1.891 make_m2s 4572 13.5 0.044 0.045 1.734 1.876 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.833 1.833 calculate_dm_sparse 119 9.5 0.000 0.000 1.796 1.805 make_images 4572 14.5 0.190 0.203 1.648 1.788 potential_pw2rs 119 12.3 0.007 0.008 1.724 1.730 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.710 1.712 prepare_preconditioner 11 7.9 0.000 0.000 1.708 1.710 make_preconditioner 11 8.9 0.000 0.001 1.708 1.710 multiply_cannon_sync_h2d 18288 15.5 1.407 1.657 1.407 1.657 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.562 1.645 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.619 1.620 cp_fm_redistribute_end 50 14.0 1.208 1.601 1.209 1.601 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.570 1.577 cp_fm_diag_elpa_base 50 14.0 0.376 1.527 0.390 1.571 fft3d_ps 1201 14.6 0.524 0.542 1.557 1.571 fft_wrap_pw1pw2_140 487 13.2 0.091 0.094 1.514 1.527 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.515 1.524 ot_diis_step 108 11.5 0.011 0.011 1.457 1.457 mp_sum_l 7207 12.9 1.103 1.442 1.103 1.442 grid_collocate_task_list 119 9.7 1.213 1.359 1.213 1.359 acc_transpose_blocks 18288 15.5 0.075 0.076 1.239 1.273 jit_kernel_multiply 7 16.1 0.620 1.253 0.620 1.253 wfi_extrapolate 11 7.9 0.001 0.001 1.180 1.180 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.175 1.180 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.953 0.954 make_images_data 4572 15.5 0.046 0.051 0.777 0.938 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.855 0.875 hybrid_alltoall_any 4725 16.4 0.055 0.113 0.667 0.861 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.836 0.837 acc_transpose_blocks_kernels 18288 16.5 0.209 0.219 0.801 0.825 mp_waitany 9880 13.7 0.520 0.794 0.520 0.794 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.514 0.767 mp_alltoall_d11v 2130 13.8 0.655 0.759 0.655 0.759 mp_alltoall_z22v 1201 16.6 0.639 0.740 0.639 0.740 mp_sum_dm 438 4.9 0.716 0.740 0.716 0.740 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.652 0.738 md_write_output 11 3.9 0.023 0.715 0.024 0.727 md_output 10 3.0 0.000 0.000 0.025 0.725 update_particle_set 20 4.0 0.000 0.000 0.701 0.723 cp_fm_cholesky_invert 11 10.9 0.702 0.706 0.702 0.706 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.641000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=497.181818, yerr=1.945540 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/06/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 114.044384E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3805952 0.0% 0.0% 100.0% average stack size 0.0 0.0 38.6 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 558.612480E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62659. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_comm_split 50 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.029 35.403 35.403 qs_mol_dyn_low 1 2.0 0.003 0.007 35.140 35.248 qs_forces 11 3.9 0.020 0.027 34.923 34.930 qs_energies 11 4.9 0.001 0.001 33.179 33.185 scf_env_do_scf 11 5.9 0.000 0.001 28.169 28.171 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 24.669 24.669 dbcsr_multiply_generic 2286 12.5 0.099 0.100 17.931 18.058 velocity_verlet 10 3.0 0.001 0.001 18.021 18.029 qs_scf_new_mos 108 7.5 0.001 0.001 16.079 16.137 qs_scf_loop_do_ot 108 8.5 0.001 0.001 16.079 16.137 ot_scf_mini 108 9.5 0.002 0.003 15.139 15.190 multiply_cannon 2286 13.5 0.223 0.249 14.406 15.189 multiply_cannon_loop 2286 14.5 0.934 0.964 13.454 14.052 ot_mini 108 10.5 0.001 0.001 9.199 9.262 multiply_cannon_multrec 27432 15.5 2.323 2.974 8.548 8.919 qs_ot_get_derivative 108 11.5 0.001 0.001 7.378 7.430 dbcsr_mm_accdrv_process 47916 15.9 5.175 6.905 6.132 7.320 rebuild_ks_matrix 119 8.3 0.000 0.000 6.640 6.692 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.640 6.691 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.899 5.943 sum_up_and_integrate 119 10.3 0.035 0.038 3.859 3.867 integrate_v_rspace 119 11.3 0.002 0.002 3.823 3.832 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.629 3.662 calculate_rho_elec 119 8.7 0.040 0.046 3.629 3.661 init_scf_run 11 5.9 0.000 0.001 3.656 3.656 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.656 3.656 qs_ot_get_p 119 10.4 0.001 0.001 3.412 3.481 init_scf_loop 11 6.9 0.000 0.000 3.476 3.476 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.843 3.167 prepare_preconditioner 11 7.9 0.000 0.000 2.548 2.557 make_preconditioner 11 8.9 0.000 0.000 2.548 2.557 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.162 2.486 mp_waitall_1 137007 16.6 1.888 2.410 1.888 2.410 make_m2s 4572 13.5 0.054 0.056 2.205 2.326 rs_pw_transfer 974 11.9 0.009 0.010 2.106 2.298 calculate_first_density_matrix 1 7.0 0.000 0.000 2.269 2.271 density_rs2pw 119 9.7 0.004 0.004 2.050 2.236 make_images 4572 14.5 0.271 0.331 2.097 2.216 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.977 2.163 apply_single 119 13.6 0.000 0.000 1.976 2.162 pw_transfer 1439 11.6 0.066 0.072 2.073 2.111 calculate_dm_sparse 119 9.5 0.000 0.000 2.031 2.093 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.066 2.075 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.979 2.022 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.931 1.957 grid_integrate_task_list 119 12.3 1.820 1.918 1.820 1.918 jit_kernel_multiply 10 15.8 0.897 1.821 0.897 1.821 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.782 1.783 ot_diis_step 108 11.5 0.012 0.012 1.781 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.750 1.750 fft3d_ps 1201 14.6 0.559 0.606 1.653 1.694 fft_wrap_pw1pw2_140 487 13.2 0.088 0.096 1.609 1.656 potential_pw2rs 119 12.3 0.008 0.009 1.626 1.630 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.606 1.621 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.498 1.499 cp_fm_redistribute_end 50 14.0 0.992 1.473 0.993 1.473 acc_transpose_blocks 27432 15.5 0.112 0.114 1.447 1.468 cp_fm_diag_elpa_base 50 14.0 0.459 1.408 0.478 1.446 multiply_cannon_metrocomm3 27432 15.5 0.038 0.039 0.842 1.395 grid_collocate_task_list 119 9.7 1.221 1.351 1.221 1.351 wfi_extrapolate 11 7.9 0.001 0.001 1.337 1.337 mp_sum_l 7207 12.9 0.896 1.235 0.896 1.235 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.213 1.225 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.100 1.119 cp_fm_upper_to_full 72 13.5 0.792 1.098 0.792 1.098 qs_energies_init_hamiltonians 11 5.9 0.018 0.019 1.089 1.089 multiply_cannon_sync_h2d 27432 15.5 0.995 1.063 0.995 1.063 dbcsr_complete_redistribute 329 12.2 0.120 0.143 0.783 1.052 make_images_data 4572 15.5 0.046 0.050 0.856 1.002 hybrid_alltoall_any 4725 16.4 0.062 0.152 0.731 0.918 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.790 0.868 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.831 0.836 acc_transpose_blocks_kernels 27432 16.5 0.267 0.275 0.824 0.832 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.564 0.828 mp_alltoall_d11v 2130 13.8 0.725 0.806 0.725 0.806 cp_fm_cholesky_invert 11 10.9 0.777 0.780 0.777 0.780 mp_alltoall_z22v 1201 16.6 0.677 0.717 0.677 0.717 mp_alltoall_i22 627 13.8 0.426 0.716 0.426 0.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=35.403000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.909091, yerr=3.315379 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/07/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 117.977176E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 1384136 0.0% 0.0% 100.0% average stack size 0.0 0.0 106.2 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 597.745664E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10224 344. MP_Sync 104 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_comm_split 50 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.035 28.400 28.401 qs_mol_dyn_low 1 2.0 0.003 0.004 28.233 28.240 qs_forces 11 3.9 0.002 0.002 27.520 27.520 qs_energies 11 4.9 0.001 0.001 25.813 25.815 scf_env_do_scf 11 5.9 0.000 0.001 21.050 21.050 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 18.629 18.629 velocity_verlet 10 3.0 0.001 0.001 14.793 14.837 dbcsr_multiply_generic 2286 12.5 0.092 0.095 11.988 12.067 qs_scf_new_mos 108 7.5 0.001 0.001 10.654 10.681 qs_scf_loop_do_ot 108 8.5 0.001 0.001 10.653 10.680 multiply_cannon 2286 13.5 0.232 0.239 9.601 10.016 ot_scf_mini 108 9.5 0.002 0.002 9.987 10.014 multiply_cannon_loop 2286 14.5 0.328 0.337 8.710 8.885 multiply_cannon_multrec 9144 15.5 1.614 1.872 5.945 6.170 rebuild_ks_matrix 119 8.3 0.000 0.000 5.963 5.984 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.962 5.984 ot_mini 108 10.5 0.001 0.001 5.549 5.578 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.311 5.330 dbcsr_mm_accdrv_process 12550 15.8 3.323 4.136 4.230 4.321 qs_ot_get_derivative 108 11.5 0.001 0.001 4.269 4.296 sum_up_and_integrate 119 10.3 0.038 0.041 3.644 3.649 integrate_v_rspace 119 11.3 0.002 0.003 3.606 3.611 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.536 3.543 calculate_rho_elec 119 8.7 0.060 0.061 3.536 3.542 init_scf_run 11 5.9 0.000 0.001 3.344 3.344 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.344 3.344 qs_ot_get_p 119 10.4 0.001 0.001 2.701 2.741 init_scf_loop 11 6.9 0.000 0.000 2.401 2.402 calculate_first_density_matrix 1 7.0 0.000 0.000 2.233 2.234 pw_transfer 1439 11.6 0.066 0.069 2.138 2.150 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 2.044 2.058 density_rs2pw 119 9.7 0.004 0.004 1.859 1.996 grid_integrate_task_list 119 12.3 1.839 1.928 1.839 1.928 mp_waitall_1 115863 16.7 1.255 1.828 1.255 1.828 calculate_dm_sparse 119 9.5 0.000 0.000 1.800 1.822 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.817 1.820 make_m2s 4572 13.5 0.034 0.035 1.655 1.805 jit_kernel_multiply 9 15.7 0.870 1.768 0.870 1.768 rs_pw_transfer 974 11.9 0.008 0.008 1.609 1.749 fft3d_ps 1201 14.6 0.562 0.574 1.702 1.715 fft_wrap_pw1pw2_140 487 13.2 0.087 0.090 1.702 1.714 make_images 4572 14.5 0.268 0.305 1.566 1.714 prepare_preconditioner 11 7.9 0.000 0.000 1.657 1.662 make_preconditioner 11 8.9 0.000 0.000 1.657 1.662 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.628 1.628 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.550 1.575 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.551 1.553 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.538 1.551 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.388 1.397 grid_collocate_task_list 119 9.7 1.271 1.393 1.271 1.393 potential_pw2rs 119 12.3 0.010 0.010 1.382 1.386 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.355 1.355 cp_fm_redistribute_end 50 14.0 0.674 1.332 0.675 1.333 cp_fm_diag_elpa_base 50 14.0 0.612 1.264 0.657 1.320 ot_diis_step 108 11.5 0.012 0.013 1.268 1.268 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.225 1.231 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.215 1.216 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.140 1.158 apply_single 119 13.6 0.000 0.000 1.140 1.158 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.088 1.103 wfi_extrapolate 11 7.9 0.001 0.001 1.063 1.063 hybrid_alltoall_any 4725 16.4 0.063 0.176 0.752 0.977 make_images_data 4572 15.5 0.040 0.044 0.770 0.938 acc_transpose_blocks 9144 15.5 0.038 0.039 0.912 0.923 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.865 0.919 mp_alltoall_d11v 2130 13.8 0.756 0.836 0.756 0.836 cp_fm_cholesky_invert 11 10.9 0.830 0.832 0.830 0.832 multiply_cannon_sync_h2d 9144 15.5 0.710 0.789 0.710 0.789 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.773 0.775 qs_env_update_s_mstruct 11 6.9 0.003 0.004 0.672 0.724 mp_sum_dm 438 4.9 0.668 0.716 0.668 0.716 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.697 0.706 md_write_output 11 3.9 0.043 0.670 0.046 0.703 md_output 10 3.0 0.000 0.000 0.046 0.702 update_particle_set 20 4.0 0.000 0.000 0.655 0.699 acc_transpose_blocks_kernels 9144 16.5 0.117 0.120 0.672 0.679 mp_alltoall_z22v 1201 16.6 0.587 0.645 0.587 0.645 mp_allgather_i34 2286 14.5 0.225 0.615 0.225 0.615 multiply_cannon_metrocomm3 9144 15.5 0.019 0.019 0.304 0.603 yz_to_x 606 15.1 0.266 0.276 0.575 0.594 qs_create_task_list 11 7.9 0.000 0.000 0.542 0.568 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.401000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=567.636364, yerr=3.548949 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 744.927232E+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 0.018 0.044 41.049 41.050 qs_mol_dyn_low 1 2.0 0.003 0.004 40.757 40.812 qs_forces 11 3.9 0.002 0.002 40.692 40.693 qs_energies 11 4.9 0.001 0.001 38.717 38.721 scf_env_do_scf 11 5.9 0.001 0.001 32.872 32.872 scf_env_do_scf_inner_loop 108 6.5 0.004 0.010 25.141 25.142 velocity_verlet 10 3.0 0.001 0.001 23.070 23.075 dbcsr_multiply_generic 2286 12.5 0.099 0.101 17.057 17.232 qs_scf_new_mos 108 7.5 0.001 0.001 15.373 15.475 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.372 15.474 ot_scf_mini 108 9.5 0.002 0.002 14.303 14.405 multiply_cannon 2286 13.5 0.304 0.311 13.441 14.307 multiply_cannon_loop 2286 14.5 0.341 0.347 12.216 13.121 ot_mini 108 10.5 0.001 0.001 8.572 8.691 multiply_cannon_multrec 9144 15.5 3.415 4.937 8.494 8.558 init_scf_loop 11 6.9 0.000 0.000 7.703 7.706 rebuild_ks_matrix 119 8.3 0.000 0.000 7.309 7.452 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.309 7.451 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.611 6.741 prepare_preconditioner 11 7.9 0.000 0.000 6.696 6.709 make_preconditioner 11 8.9 0.000 0.000 6.696 6.709 qs_ot_get_derivative 108 11.5 0.001 0.001 6.576 6.678 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.264 6.588 dbcsr_mm_accdrv_process 12550 15.8 3.935 5.555 4.957 6.298 cp_fm_upper_to_full 72 14.2 3.151 4.523 3.151 4.523 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.293 4.345 calculate_rho_elec 119 8.7 0.118 0.121 4.293 4.345 sum_up_and_integrate 119 10.3 0.065 0.066 4.011 4.017 integrate_v_rspace 119 11.3 0.003 0.003 3.945 3.953 init_scf_run 11 5.9 0.000 0.001 3.823 3.823 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.822 3.822 qs_ot_get_p 119 10.4 0.001 0.001 3.080 3.219 mp_waitall_1 94719 16.7 2.207 3.160 2.207 3.160 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.434 2.824 pw_transfer 1439 11.6 0.069 0.069 2.801 2.805 dbcsr_complete_redistribute 329 12.2 0.286 0.291 1.949 2.754 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.702 2.706 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.648 2.449 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.135 2.435 apply_single 119 13.6 0.000 0.000 2.135 2.434 calculate_first_density_matrix 1 7.0 0.000 0.000 2.346 2.401 fft3d_ps 1201 14.6 0.594 0.605 2.326 2.330 fft_wrap_pw1pw2_140 487 13.2 0.095 0.096 2.292 2.299 make_m2s 4572 13.5 0.038 0.038 2.133 2.287 calculate_dm_sparse 119 9.5 0.000 0.000 2.183 2.235 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.426 2.222 mp_alltoall_i22 627 13.8 1.404 2.211 1.404 2.211 density_rs2pw 119 9.7 0.004 0.004 2.176 2.191 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 1.293 2.170 make_images 4572 14.5 0.353 0.385 2.012 2.165 grid_integrate_task_list 119 12.3 2.057 2.066 2.057 2.066 ot_diis_step 108 11.5 0.014 0.014 1.972 1.972 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.850 1.851 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.761 1.762 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.756 1.757 jit_kernel_multiply 10 15.5 0.995 1.719 0.995 1.719 mp_sum_l 7207 12.9 0.945 1.688 0.945 1.688 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.628 1.680 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.583 1.584 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.523 1.557 grid_collocate_task_list 119 9.7 1.487 1.504 1.487 1.504 rs_pw_transfer 974 11.9 0.009 0.009 1.400 1.420 potential_pw2rs 119 12.3 0.014 0.014 1.415 1.418 cp_fm_cholesky_invert 11 10.9 1.383 1.386 1.383 1.386 wfi_extrapolate 11 7.9 0.001 0.001 1.353 1.353 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.259 1.311 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.308 1.308 cp_fm_diag_elpa_base 50 14.0 1.167 1.217 1.307 1.307 hybrid_alltoall_any 4725 16.4 0.087 0.148 1.029 1.212 make_images_data 4572 15.5 0.045 0.047 0.970 1.153 mp_alltoall_d11v 2130 13.8 1.082 1.148 1.082 1.148 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.114 1.133 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.095 1.117 multiply_cannon_sync_h2d 9144 15.5 1.045 1.047 1.045 1.047 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.981 1.019 qs_create_task_list 11 7.9 0.000 0.001 0.936 0.949 generate_qs_task_list 11 8.9 0.371 0.389 0.936 0.949 yz_to_x 606 15.1 0.460 0.473 0.926 0.936 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.910 0.923 acc_transpose_blocks 9144 15.5 0.037 0.038 0.894 0.900 mp_alltoall_z22v 1201 16.6 0.778 0.821 0.778 0.821 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=41.050000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=700.090909, yerr=13.957432 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 502.038528E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65372. MP_Allreduce 9840 486. MP_Sync 100 MP_Alltoall 1938 1379060. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_comm_split 48 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.044 0.056 82.959 82.960 qs_mol_dyn_low 1 2.0 0.003 0.006 82.565 82.575 qs_forces 11 3.9 0.002 0.002 82.500 82.500 qs_energies 11 4.9 0.001 0.002 79.651 79.664 scf_env_do_scf 11 5.9 0.000 0.001 70.785 70.788 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 65.347 65.348 dbcsr_multiply_generic 2055 12.4 0.105 0.110 51.853 52.049 qs_scf_new_mos 99 7.5 0.000 0.001 47.999 48.111 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.999 48.110 ot_scf_mini 99 9.5 0.002 0.002 45.606 45.665 multiply_cannon 2055 13.4 0.180 0.187 42.402 43.165 multiply_cannon_loop 2055 14.4 1.536 1.585 41.454 42.244 velocity_verlet 10 3.0 0.001 0.001 41.876 41.877 ot_mini 99 10.5 0.001 0.001 27.657 27.750 qs_ot_get_derivative 99 11.5 0.001 0.001 20.819 20.903 multiply_cannon_multrec 49320 15.4 12.144 13.044 17.178 17.911 rebuild_ks_matrix 110 8.3 0.000 0.000 14.542 14.638 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.541 14.638 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.761 12.844 mp_waitall_1 220248 16.4 11.306 12.445 11.306 12.445 multiply_cannon_sync_h2d 49320 15.4 9.975 10.558 9.975 10.558 qs_ot_get_p 110 10.4 0.001 0.001 9.477 9.595 multiply_cannon_metrocomm3 49320 15.4 0.079 0.084 6.926 7.834 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.154 7.683 apply_single 110 13.6 0.000 0.001 7.154 7.682 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.078 7.605 sum_up_and_integrate 110 10.3 0.036 0.043 7.071 7.085 integrate_v_rspace 110 11.3 0.002 0.003 7.034 7.057 init_scf_run 11 5.9 0.000 0.001 6.847 6.847 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.847 6.847 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 6.784 6.835 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.596 6.737 calculate_rho_elec 110 8.6 0.021 0.026 6.595 6.736 ot_diis_step 99 11.5 0.005 0.006 6.547 6.547 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.322 6.352 mp_sum_l 6514 12.8 5.224 5.980 5.224 5.980 init_scf_loop 11 6.9 0.000 0.000 5.412 5.412 cp_dbcsr_syevd 48 12.0 0.002 0.002 5.357 5.357 dbcsr_mm_accdrv_process 87628 16.1 2.081 2.185 4.913 5.215 cp_fm_diag_elpa 48 13.0 0.000 0.000 4.908 4.909 cp_fm_redistribute_end 48 14.0 4.286 4.886 4.290 4.887 cp_fm_diag_elpa_base 48 14.0 0.591 4.745 0.595 4.777 rs_pw_transfer 902 11.9 0.012 0.013 3.583 4.184 wfi_extrapolate 11 7.9 0.001 0.001 3.997 3.997 density_rs2pw 110 9.6 0.004 0.005 3.390 3.985 calculate_dm_sparse 110 9.5 0.001 0.001 3.768 3.899 make_m2s 4110 13.4 0.061 0.067 3.767 3.873 make_images 4110 14.4 0.177 0.191 3.672 3.782 multiply_cannon_metrocomm1 49320 15.4 0.060 0.063 2.601 3.623 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.557 3.564 grid_integrate_task_list 110 12.3 3.252 3.403 3.252 3.403 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.243 3.290 pw_transfer 1331 11.6 0.055 0.069 3.223 3.281 prepare_preconditioner 11 7.9 0.000 0.000 3.253 3.268 make_preconditioner 11 8.9 0.000 0.000 3.253 3.268 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.212 3.256 fft_wrap_pw1pw2 1111 12.6 0.007 0.008 3.134 3.195 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.043 3.076 calculate_first_density_matrix 1 7.0 0.000 0.000 2.762 2.768 fft_wrap_pw1pw2_140 451 13.1 0.172 0.192 2.679 2.747 fft3d_ps 1111 14.6 0.778 0.878 2.660 2.712 potential_pw2rs 110 12.3 0.006 0.007 2.590 2.608 jit_kernel_multiply 13 15.9 2.549 2.580 2.549 2.580 mp_waitany 14300 13.8 1.790 2.520 1.790 2.520 mp_alltoall_d11v 2046 13.8 2.057 2.480 2.057 2.480 grid_collocate_task_list 110 9.6 2.083 2.302 2.083 2.302 acc_transpose_blocks 49320 15.4 0.224 0.236 2.166 2.246 mp_sum_d 3879 11.9 1.479 1.998 1.479 1.998 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.941 make_images_data 4110 15.4 0.044 0.047 1.736 1.884 hybrid_alltoall_any 4261 16.3 0.082 0.482 1.510 1.809 cp_fm_cholesky_invert 11 10.9 1.768 1.771 1.768 1.771 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.650 1.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=82.960000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=477.000000, yerr=2.296242 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 587.038720E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65587. MP_Allreduce 9839 562. MP_Sync 100 MP_Alltoall 1717 2464264. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_comm_split 48 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.033 0.048 68.810 68.811 qs_mol_dyn_low 1 2.0 0.003 0.004 68.432 68.442 qs_forces 11 3.9 0.003 0.005 68.366 68.366 qs_energies 11 4.9 0.009 0.054 65.038 65.041 scf_env_do_scf 11 5.9 0.001 0.001 56.194 56.197 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 48.619 48.620 dbcsr_multiply_generic 2055 12.4 0.113 0.118 37.358 37.494 velocity_verlet 10 3.0 0.001 0.001 36.019 36.020 qs_scf_new_mos 99 7.5 0.001 0.001 32.249 32.369 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.248 32.368 multiply_cannon 2055 13.4 0.223 0.244 30.888 32.029 ot_scf_mini 99 9.5 0.003 0.004 30.604 30.711 multiply_cannon_loop 2055 14.4 0.923 0.944 29.612 30.570 ot_mini 99 10.5 0.001 0.001 17.990 18.119 multiply_cannon_multrec 24660 15.4 7.636 9.385 13.816 15.537 rebuild_ks_matrix 110 8.3 0.000 0.001 13.837 13.928 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.023 13.836 13.928 qs_ot_get_derivative 99 11.5 0.001 0.002 12.213 12.329 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.189 12.264 mp_waitall_1 176588 16.5 7.493 10.147 7.493 10.147 multiply_cannon_sync_h2d 24660 15.4 7.025 8.117 7.025 8.117 multiply_cannon_metrocomm3 24660 15.4 0.069 0.070 5.111 7.783 init_scf_loop 11 6.9 0.001 0.008 7.540 7.541 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.493 7.217 apply_single 110 13.6 0.000 0.001 6.492 7.217 sum_up_and_integrate 110 10.3 0.053 0.059 6.645 6.656 integrate_v_rspace 110 11.3 0.002 0.005 6.592 6.605 dbcsr_mm_accdrv_process 52282 16.1 4.663 5.604 6.018 6.313 init_scf_run 11 5.9 0.000 0.001 6.210 6.211 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.210 6.210 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.194 6.202 calculate_rho_elec 110 8.6 0.039 0.047 6.193 6.201 qs_ot_get_p 110 10.4 0.001 0.001 6.009 6.170 ot_diis_step 99 11.5 0.010 0.010 5.731 5.731 prepare_preconditioner 11 7.9 0.000 0.000 5.492 5.510 make_preconditioner 11 8.9 0.001 0.004 5.492 5.510 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.599 5.327 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.063 5.222 make_m2s 4110 13.4 0.057 0.060 4.065 4.544 make_images 4110 14.4 0.399 0.443 3.956 4.432 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.176 4.197 pw_transfer 1331 11.6 0.066 0.073 3.734 3.878 density_rs2pw 110 9.6 0.004 0.005 3.321 3.829 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.627 3.774 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.756 3.756 wfi_extrapolate 11 7.9 0.001 0.001 3.513 3.513 rs_pw_transfer 902 11.9 0.012 0.014 2.951 3.510 grid_integrate_task_list 110 12.3 3.157 3.347 3.157 3.347 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.325 3.327 fft_wrap_pw1pw2_140 451 13.1 0.203 0.220 3.117 3.268 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.247 3.249 cp_fm_redistribute_end 48 14.0 2.430 3.223 2.432 3.224 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.125 3.192 cp_fm_diag_elpa_base 48 14.0 0.759 3.084 0.790 3.169 fft3d_ps 1111 14.6 1.099 1.313 2.978 3.113 calculate_dm_sparse 110 9.5 0.001 0.001 2.942 2.970 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.914 2.957 hybrid_alltoall_any 4261 16.3 0.103 0.444 2.061 2.832 make_images_data 4110 15.4 0.048 0.052 2.294 2.792 calculate_first_density_matrix 1 7.0 0.001 0.004 2.608 2.612 cp_fm_cholesky_invert 11 10.9 2.542 2.549 2.542 2.549 grid_collocate_task_list 110 9.6 2.051 2.480 2.051 2.480 potential_pw2rs 110 12.3 0.008 0.008 2.437 2.450 mp_sum_l 6514 12.8 1.786 2.307 1.786 2.307 mp_alltoall_d11v 2046 13.8 1.743 2.004 1.743 2.004 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.924 1.945 qs_energies_init_hamiltonians 11 5.9 0.002 0.011 1.908 1.910 mp_waitany 10164 13.8 1.236 1.831 1.236 1.831 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.812 1.826 jit_kernel_multiply 9 16.4 1.008 1.747 1.008 1.747 mp_allgather_i34 2055 14.4 0.603 1.600 0.603 1.600 multiply_cannon_metrocomm4 22605 15.4 0.072 0.077 0.781 1.560 acc_transpose_blocks 24660 15.4 0.109 0.111 1.534 1.557 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.529 1.539 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.219 0.963 1.506 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.364 1.470 mp_irecv_dv 57340 16.2 0.659 1.442 0.659 1.442 dbcsr_complete_redistribute 325 12.2 0.308 0.375 1.175 1.437 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.811000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=556.000000, yerr=6.888726 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 659.939328E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65578. MP_Allreduce 9838 559. MP_Sync 100 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_comm_split 48 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.067 0.119 60.769 60.770 qs_mol_dyn_low 1 2.0 0.003 0.004 59.955 59.964 qs_forces 11 3.9 0.024 0.177 59.659 59.662 qs_energies 11 4.9 0.001 0.003 56.301 56.305 scf_env_do_scf 11 5.9 0.001 0.001 48.213 48.213 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 39.740 39.740 velocity_verlet 10 3.0 0.001 0.001 32.562 32.571 dbcsr_multiply_generic 2055 12.4 0.107 0.110 28.107 28.407 qs_scf_new_mos 99 7.5 0.001 0.001 24.613 24.700 qs_scf_loop_do_ot 99 8.5 0.001 0.001 24.612 24.700 ot_scf_mini 99 9.5 0.002 0.003 23.401 23.499 multiply_cannon 2055 13.4 0.211 0.225 21.946 23.022 multiply_cannon_loop 2055 14.4 0.614 0.633 20.801 21.987 ot_mini 99 10.5 0.001 0.001 13.324 13.417 rebuild_ks_matrix 110 8.3 0.000 0.000 12.454 12.582 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.453 12.582 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.973 11.095 multiply_cannon_multrec 16440 15.4 4.012 5.138 9.796 10.849 mp_waitall_1 139946 16.5 6.638 9.225 6.638 9.225 qs_ot_get_derivative 99 11.5 0.001 0.001 8.932 9.027 init_scf_loop 11 6.9 0.001 0.003 8.440 8.440 multiply_cannon_metrocomm3 16440 15.4 0.043 0.045 4.063 6.772 prepare_preconditioner 11 7.9 0.000 0.000 6.657 6.675 make_preconditioner 11 8.9 0.000 0.001 6.657 6.675 sum_up_and_integrate 110 10.3 0.060 0.062 6.583 6.598 integrate_v_rspace 110 11.3 0.002 0.003 6.522 6.538 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.018 6.375 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.007 6.016 calculate_rho_elec 110 8.6 0.058 0.059 6.006 6.015 dbcsr_mm_accdrv_process 34862 16.1 4.770 5.318 5.639 5.770 init_scf_run 11 5.9 0.000 0.001 5.700 5.700 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.700 5.700 qs_ot_get_p 110 10.4 0.001 0.001 5.337 5.461 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.940 5.321 apply_single 110 13.6 0.000 0.000 4.940 5.321 make_m2s 4110 13.4 0.050 0.052 4.098 4.432 multiply_cannon_sync_h2d 16440 15.4 3.749 4.425 3.749 4.425 ot_diis_step 99 11.5 0.010 0.011 4.364 4.365 density_rs2pw 110 9.6 0.004 0.005 3.120 4.364 make_images 4110 14.4 0.392 0.508 3.982 4.317 rs_pw_transfer 902 11.9 0.010 0.011 2.641 3.874 pw_transfer 1331 11.6 0.066 0.073 3.749 3.756 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.748 3.751 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.031 3.682 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.642 3.651 grid_integrate_task_list 110 12.3 3.192 3.419 3.192 3.419 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.412 3.412 fft_wrap_pw1pw2_140 451 13.1 0.213 0.218 3.144 3.154 fft3d_ps 1111 14.6 1.088 1.103 2.940 2.950 wfi_extrapolate 11 7.9 0.001 0.001 2.934 2.934 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.904 2.905 cp_fm_redistribute_end 48 14.0 1.812 2.879 1.814 2.879 cp_fm_diag_elpa_base 48 14.0 1.003 2.728 1.061 2.836 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 2.818 2.820 make_images_data 4110 15.4 0.045 0.050 2.366 2.814 hybrid_alltoall_any 4261 16.3 0.106 0.374 2.080 2.759 calculate_first_density_matrix 1 7.0 0.000 0.003 2.681 2.682 cp_fm_cholesky_invert 11 10.9 2.585 2.591 2.585 2.591 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.509 2.572 calculate_dm_sparse 110 9.5 0.001 0.001 2.482 2.517 mp_waitany 17072 13.8 1.183 2.483 1.183 2.483 grid_collocate_task_list 110 9.6 2.081 2.475 2.081 2.475 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.300 2.362 multiply_cannon_metrocomm4 14385 15.4 0.045 0.048 0.860 2.331 potential_pw2rs 110 12.3 0.011 0.011 2.310 2.320 mp_alltoall_d11v 2046 13.8 1.694 2.216 1.694 2.216 mp_irecv_dv 48980 15.7 0.790 2.211 0.790 2.211 rs_pw_transfer_RS2PW_140 121 11.5 0.175 0.178 0.895 2.123 mp_sum_l 6514 12.8 1.493 2.061 1.493 2.061 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.960 1.962 dbcsr_complete_redistribute 325 12.2 0.326 0.363 1.403 1.867 cp_fm_upper_to_full 70 13.6 1.402 1.856 1.402 1.856 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.680 1.691 cp_fm_cholesky_decompose 22 10.9 1.549 1.572 1.549 1.572 mp_allgather_i34 2055 14.4 0.470 1.477 0.470 1.477 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.450 1.461 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.354 1.451 copy_fm_to_dbcsr 174 11.2 0.001 0.001 0.919 1.371 rs_gather_matrices 110 12.3 0.234 0.261 0.942 1.351 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.235 1.243 acc_transpose_blocks 16440 15.4 0.071 0.073 1.221 1.234 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.770000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=625.363636, yerr=8.282212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 736.362496E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65576. MP_Allreduce 9838 600. MP_Sync 100 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_comm_split 48 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.033 64.719 64.721 qs_mol_dyn_low 1 2.0 0.003 0.004 64.484 64.492 qs_forces 11 3.9 0.002 0.002 64.399 64.399 qs_energies 11 4.9 0.001 0.002 60.987 60.991 scf_env_do_scf 11 5.9 0.000 0.001 52.699 52.702 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 41.183 41.184 velocity_verlet 10 3.0 0.001 0.001 36.838 36.844 dbcsr_multiply_generic 2055 12.4 0.114 0.116 29.254 29.441 qs_scf_new_mos 99 7.5 0.001 0.001 25.976 26.102 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.975 26.102 ot_scf_mini 99 9.5 0.003 0.003 24.354 24.457 multiply_cannon 2055 13.4 0.244 0.265 22.463 23.472 multiply_cannon_loop 2055 14.4 0.881 0.900 20.917 21.523 ot_mini 99 10.5 0.001 0.001 13.917 14.047 multiply_cannon_multrec 24660 15.4 4.205 6.825 12.603 13.841 rebuild_ks_matrix 110 8.3 0.000 0.000 12.302 12.411 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.302 12.411 init_scf_loop 11 6.9 0.000 0.000 11.476 11.477 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.870 10.966 qs_ot_get_derivative 99 11.5 0.001 0.002 9.765 9.872 prepare_preconditioner 11 7.9 0.000 0.000 9.729 9.745 make_preconditioner 11 8.9 0.000 0.000 9.729 9.745 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.965 9.422 dbcsr_mm_accdrv_process 52304 16.0 6.609 8.007 8.251 9.148 sum_up_and_integrate 110 10.3 0.067 0.071 6.566 6.580 integrate_v_rspace 110 11.3 0.002 0.003 6.498 6.512 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.193 6.204 calculate_rho_elec 110 8.6 0.077 0.081 6.193 6.204 mp_waitall_1 121746 16.5 4.087 6.163 4.087 6.163 make_m2s 4110 13.4 0.060 0.061 5.217 5.691 qs_ot_get_p 110 10.4 0.001 0.001 5.462 5.593 make_images 4110 14.4 0.576 0.694 5.076 5.548 init_scf_run 11 5.9 0.000 0.001 5.546 5.546 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.545 5.546 cp_fm_upper_to_full 70 13.8 3.256 4.640 3.256 4.640 ot_diis_step 99 11.5 0.011 0.011 4.117 4.117 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.017 4.077 apply_single 110 13.6 0.000 0.000 4.017 4.077 pw_transfer 1331 11.6 0.065 0.074 3.910 3.940 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.804 3.838 dbcsr_complete_redistribute 325 12.2 0.482 0.532 2.699 3.828 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.696 3.710 density_rs2pw 110 9.6 0.004 0.004 3.132 3.648 grid_integrate_task_list 110 12.3 3.264 3.482 3.264 3.482 multiply_cannon_sync_h2d 24660 15.4 3.185 3.340 3.185 3.340 fft_wrap_pw1pw2_140 451 13.1 0.202 0.213 3.283 3.321 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.170 3.287 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.232 3.279 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.220 3.220 fft3d_ps 1111 14.6 1.090 1.127 3.095 3.117 multiply_cannon_metrocomm3 24660 15.4 0.035 0.036 1.421 3.115 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.991 3.052 make_images_data 4110 15.4 0.048 0.052 2.569 3.025 hybrid_alltoall_any 4261 16.3 0.120 0.459 2.229 2.990 wfi_extrapolate 11 7.9 0.001 0.001 2.967 2.967 calculate_dm_sparse 110 9.5 0.001 0.001 2.903 2.934 rs_pw_transfer 902 11.9 0.010 0.011 2.400 2.923 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.875 2.877 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.755 2.857 mp_alltoall_i22 605 13.7 1.639 2.805 1.639 2.805 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.698 2.700 cp_fm_redistribute_end 48 14.0 1.351 2.673 1.352 2.674 cp_fm_diag_elpa_base 48 14.0 1.241 2.539 1.319 2.648 cp_fm_cholesky_invert 11 10.9 2.617 2.625 2.617 2.625 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.460 2.492 calculate_first_density_matrix 1 7.0 0.000 0.000 2.486 2.490 grid_collocate_task_list 110 9.6 2.181 2.428 2.181 2.428 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.242 2.243 potential_pw2rs 110 12.3 0.012 0.012 2.212 2.220 jit_kernel_multiply 12 15.7 1.314 2.195 1.314 2.195 mp_alltoall_d11v 2046 13.8 1.703 2.032 1.703 2.032 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.683 1.716 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 1.592 1.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.677 1.689 cp_fm_cholesky_decompose 22 10.9 1.633 1.685 1.633 1.685 mp_allgather_i34 2055 14.4 0.638 1.567 0.638 1.567 mp_waitany 13376 13.8 1.053 1.547 1.053 1.547 acc_transpose_blocks 24660 15.4 0.105 0.108 1.506 1.540 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.516 1.535 mp_sum_l 6514 12.8 0.887 1.523 0.887 1.523 multiply_cannon_metrocomm4 20550 15.4 0.057 0.061 0.854 1.523 mp_irecv_dv 62702 16.1 0.755 1.446 0.755 1.446 rs_pw_transfer_RS2PW_140 121 11.5 0.169 0.179 0.806 1.333 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.195 1.304 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.721000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.909091, yerr=10.908333 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 845.111296E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3521 65574. MP_Allreduce 9838 640. MP_Sync 100 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_comm_split 48 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.034 56.402 56.404 qs_mol_dyn_low 1 2.0 0.003 0.004 56.037 56.046 qs_forces 11 3.9 0.003 0.003 55.423 55.425 qs_energies 11 4.9 0.001 0.002 51.725 51.733 scf_env_do_scf 11 5.9 0.000 0.001 43.260 43.260 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 35.582 35.583 velocity_verlet 10 3.0 0.001 0.001 31.665 31.702 dbcsr_multiply_generic 2055 12.4 0.104 0.106 22.769 22.916 qs_scf_new_mos 99 7.5 0.001 0.001 20.295 20.334 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.295 20.333 ot_scf_mini 99 9.5 0.002 0.002 19.065 19.087 multiply_cannon 2055 13.4 0.246 0.258 17.327 18.682 multiply_cannon_loop 2055 14.4 0.320 0.334 15.968 16.253 rebuild_ks_matrix 110 8.3 0.000 0.000 12.085 12.101 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 12.085 12.101 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.731 10.745 ot_mini 99 10.5 0.001 0.001 10.354 10.372 multiply_cannon_multrec 8220 15.4 3.237 4.457 7.502 8.340 mp_waitall_1 103326 16.6 5.998 7.739 5.998 7.739 init_scf_loop 11 6.9 0.000 0.000 7.632 7.633 sum_up_and_integrate 110 10.3 0.080 0.081 6.603 6.615 qs_ot_get_derivative 99 11.5 0.001 0.001 6.563 6.585 integrate_v_rspace 110 11.3 0.003 0.003 6.523 6.536 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.387 6.399 calculate_rho_elec 110 8.6 0.116 0.116 6.386 6.399 prepare_preconditioner 11 7.9 0.000 0.000 5.955 5.959 make_preconditioner 11 8.9 0.000 0.000 5.955 5.959 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.536 5.609 init_scf_run 11 5.9 0.000 0.001 5.341 5.341 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.341 5.341 dbcsr_mm_accdrv_process 17442 15.9 2.801 3.936 4.137 5.055 qs_ot_get_p 110 10.4 0.001 0.001 4.801 4.823 multiply_cannon_metrocomm3 8220 15.4 0.017 0.018 3.038 4.525 make_m2s 4110 13.4 0.038 0.040 4.183 4.457 make_images 4110 14.4 0.637 0.690 4.053 4.327 pw_transfer 1331 11.6 0.066 0.072 4.165 4.175 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.058 4.071 ot_diis_step 99 11.5 0.012 0.012 3.770 3.770 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.691 3.721 apply_single 110 13.6 0.000 0.000 3.690 3.721 grid_integrate_task_list 110 12.3 3.357 3.566 3.357 3.566 fft_wrap_pw1pw2_140 451 13.1 0.215 0.219 3.516 3.531 density_rs2pw 110 9.6 0.004 0.004 3.133 3.508 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.493 3.497 fft3d_ps 1111 14.6 1.142 1.173 3.300 3.306 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.188 3.188 multiply_cannon_sync_h2d 8220 15.4 2.924 3.020 2.924 3.020 cp_fm_cholesky_invert 11 10.9 2.867 2.870 2.867 2.870 make_images_data 4110 15.4 0.040 0.045 2.395 2.817 wfi_extrapolate 11 7.9 0.001 0.001 2.737 2.737 hybrid_alltoall_any 4261 16.3 0.201 0.865 2.308 2.736 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.679 2.681 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.669 2.670 cp_fm_redistribute_end 48 14.0 0.681 2.645 0.686 2.646 cp_fm_diag_elpa_base 48 14.0 1.777 2.454 1.952 2.609 rs_pw_transfer 902 11.9 0.010 0.011 2.192 2.605 grid_collocate_task_list 110 9.6 2.277 2.539 2.277 2.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.530 2.531 calculate_first_density_matrix 1 7.0 0.000 0.000 2.508 2.509 calculate_dm_sparse 110 9.5 0.001 0.001 2.450 2.493 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.183 2.197 potential_pw2rs 110 12.3 0.015 0.015 2.146 2.149 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.765 1.980 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.872 1.890 mp_alltoall_d11v 2046 13.8 1.601 1.885 1.601 1.885 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.776 1.788 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.688 1.690 cp_fm_cholesky_decompose 22 10.9 1.659 1.668 1.659 1.668 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.519 1.645 mp_allgather_i34 2055 14.4 0.541 1.634 0.541 1.634 jit_kernel_multiply 9 15.9 1.028 1.616 1.028 1.616 mp_waitany 9240 13.8 1.082 1.526 1.082 1.526 dbcsr_complete_redistribute 325 12.2 0.564 0.582 1.425 1.507 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.420 1.433 qs_create_task_list 11 7.9 0.001 0.001 1.222 1.326 generate_qs_task_list 11 8.9 0.380 0.447 1.222 1.326 rs_pw_transfer_RS2PW_140 121 11.5 0.161 0.166 0.803 1.219 rs_gather_matrices 110 12.3 0.324 0.361 0.929 1.183 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.112 1.139 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=56.404000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=792.000000, yerr=15.039342 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 1.366172E+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 0.020 0.054 87.444 87.458 qs_mol_dyn_low 1 2.0 0.003 0.004 86.881 86.889 qs_forces 11 3.9 0.002 0.002 86.813 86.814 qs_energies 11 4.9 0.001 0.002 82.608 82.609 scf_env_do_scf 11 5.9 0.001 0.001 72.545 72.545 velocity_verlet 10 3.0 0.001 0.001 55.573 55.580 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 44.412 44.414 dbcsr_multiply_generic 2055 12.4 0.118 0.121 28.460 28.624 init_scf_loop 11 6.9 0.000 0.000 28.062 28.067 qs_scf_new_mos 99 7.5 0.001 0.001 26.023 26.076 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.022 26.075 prepare_preconditioner 11 7.9 0.000 0.000 26.025 26.031 make_preconditioner 11 8.9 0.000 0.000 26.025 26.031 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.293 25.501 ot_scf_mini 99 9.5 0.002 0.002 24.271 24.308 multiply_cannon 2055 13.4 0.348 0.375 21.634 22.315 multiply_cannon_loop 2055 14.4 0.341 0.344 19.875 20.189 cp_fm_upper_to_full 70 14.2 12.630 18.077 12.630 18.077 rebuild_ks_matrix 110 8.3 0.000 0.001 14.219 14.266 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 14.218 14.265 ot_mini 99 10.5 0.001 0.001 13.693 13.728 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.882 12.924 dbcsr_complete_redistribute 325 12.2 1.021 1.042 7.278 10.479 multiply_cannon_multrec 8220 15.4 4.464 4.608 9.647 9.938 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.269 9.469 qs_ot_get_derivative 99 11.5 0.001 0.001 9.166 9.204 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.717 8.897 mp_alltoall_i22 605 13.7 5.345 8.528 5.345 8.528 mp_waitall_1 84994 16.7 7.518 8.485 7.518 8.485 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.777 7.817 calculate_rho_elec 110 8.6 0.227 0.227 7.777 7.817 sum_up_and_integrate 110 10.3 0.150 0.151 7.383 7.398 integrate_v_rspace 110 11.3 0.004 0.004 7.233 7.247 init_scf_run 11 5.9 0.000 0.001 5.672 5.672 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.672 5.672 make_m2s 4110 13.4 0.043 0.044 5.108 5.598 qs_ot_get_p 110 10.4 0.001 0.001 5.422 5.472 make_images 4110 14.4 0.882 0.936 4.920 5.408 pw_transfer 1331 11.6 0.076 0.076 5.326 5.333 dbcsr_mm_accdrv_process 11614 15.7 3.358 3.669 5.042 5.286 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 5.209 5.216 multiply_cannon_metrocomm3 8220 15.4 0.018 0.019 4.883 5.195 cp_fm_cholesky_invert 11 10.9 5.187 5.192 5.187 5.192 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.640 5.105 apply_single 110 13.6 0.000 0.000 4.640 5.104 fft_wrap_pw1pw2_140 451 13.1 0.222 0.223 4.542 4.547 ot_diis_step 99 11.5 0.015 0.015 4.506 4.506 fft3d_ps 1111 14.6 1.294 1.303 4.402 4.409 multiply_cannon_sync_h2d 8220 15.4 3.948 3.954 3.948 3.954 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.916 3.923 density_rs2pw 110 9.6 0.004 0.004 3.808 3.844 qs_energies_init_hamiltonians 11 5.9 0.068 0.096 3.770 3.774 grid_integrate_task_list 110 12.3 3.663 3.718 3.663 3.718 hybrid_alltoall_any 4261 16.3 0.257 0.555 2.841 3.485 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.472 3.472 make_images_data 4110 15.4 0.044 0.047 2.754 3.375 wfi_extrapolate 11 7.9 0.001 0.001 3.284 3.284 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.817 3.257 calculate_dm_sparse 110 9.5 0.001 0.001 3.104 3.125 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.886 2.886 cp_fm_diag_elpa_base 48 14.0 2.360 2.547 2.884 2.884 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.827 2.831 grid_collocate_task_list 110 9.6 2.630 2.662 2.630 2.662 potential_pw2rs 110 12.3 0.021 0.021 2.517 2.519 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.373 2.397 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.338 2.350 rs_pw_transfer 902 11.9 0.011 0.011 2.251 2.275 calculate_first_density_matrix 1 7.0 0.000 0.000 2.265 2.266 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.209 2.264 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.096 2.202 mp_alltoall_d11v 2046 13.8 1.908 1.991 1.908 1.991 cp_fm_cholesky_decompose 22 10.9 1.950 1.969 1.950 1.969 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.941 1.946 qs_create_task_list 11 7.9 0.001 0.001 1.893 1.945 generate_qs_task_list 11 8.9 0.735 0.788 1.893 1.944 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.740 1.762 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=87.458000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1228.181818, yerr=64.155802 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 629.190656E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175954870160 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 56898. MP_Allreduce 11228 786. MP_Sync 170 MP_Alltoall 2226 2520958. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_comm_split 83 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.033 205.644 205.645 qs_mol_dyn_low 1 2.0 0.003 0.004 205.215 205.228 qs_forces 11 3.9 0.003 0.003 205.130 205.131 qs_energies 11 4.9 0.001 0.002 199.466 199.492 scf_env_do_scf 11 5.9 0.001 0.001 181.932 181.936 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 161.509 161.511 dbcsr_multiply_generic 2507 12.6 0.178 0.185 125.601 126.087 velocity_verlet 10 3.0 0.001 0.001 123.448 123.449 qs_scf_new_mos 117 7.6 0.001 0.001 121.269 121.484 qs_scf_loop_do_ot 117 8.6 0.001 0.001 121.268 121.483 ot_scf_mini 117 9.6 0.003 0.003 114.652 114.876 multiply_cannon 2507 13.6 0.236 0.245 102.089 104.210 multiply_cannon_loop 2507 14.6 2.095 2.177 99.857 101.467 ot_mini 117 10.6 0.001 0.001 66.199 66.451 multiply_cannon_multrec 60168 15.6 33.176 35.204 42.278 43.860 qs_ot_get_derivative 117 11.6 0.001 0.001 41.463 41.691 rebuild_ks_matrix 128 8.3 0.001 0.001 33.791 34.140 qs_ks_build_kohn_sham_matrix 128 9.3 0.015 0.017 33.790 34.139 mp_waitall_1 267128 16.5 28.416 30.966 28.416 30.966 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.365 30.690 multiply_cannon_sync_h2d 60168 15.6 27.155 28.809 27.155 28.809 qs_ot_get_p 128 10.4 0.001 0.001 26.659 27.101 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.226 24.844 apply_single 128 13.6 0.001 0.001 24.226 24.844 ot_diis_step 117 11.6 0.007 0.008 24.336 24.337 init_scf_loop 11 6.9 0.000 0.000 20.343 20.345 qs_ot_p2m_diag 83 11.4 0.077 0.091 20.023 20.121 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.501 19.706 multiply_cannon_metrocomm3 60168 15.6 0.116 0.121 15.890 17.687 cp_dbcsr_syevd 83 12.4 0.004 0.005 17.418 17.419 prepare_preconditioner 11 7.9 0.000 0.000 15.715 15.773 make_preconditioner 11 8.9 0.000 0.000 15.715 15.773 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.948 15.100 cp_fm_diag_elpa 83 13.4 0.000 0.000 14.581 14.585 cp_fm_redistribute_end 83 14.4 11.529 14.509 11.540 14.511 sum_up_and_integrate 128 10.3 0.089 0.107 14.331 14.346 integrate_v_rspace 128 11.3 0.003 0.004 14.242 14.260 cp_fm_diag_elpa_base 83 14.4 2.930 14.141 2.960 14.247 make_m2s 5014 13.6 0.105 0.113 13.687 14.060 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.881 13.978 calculate_rho_elec 128 8.7 0.045 0.064 13.880 13.977 make_images 5014 14.6 0.401 0.421 13.505 13.887 init_scf_run 11 5.9 0.000 0.001 13.382 13.383 scf_env_initial_rho_setup 11 6.9 0.000 0.001 13.382 13.383 density_rs2pw 128 9.7 0.006 0.007 7.240 10.378 mp_sum_l 7870 13.0 8.557 9.783 8.557 9.783 dbcsr_mm_accdrv_process 124484 16.2 3.278 3.503 8.669 9.706 wfi_extrapolate 11 7.9 0.001 0.001 9.062 9.062 rs_pw_transfer 1046 11.9 0.016 0.019 5.734 8.920 cp_fm_cholesky_invert 11 10.9 8.908 8.916 8.908 8.916 calculate_dm_sparse 128 9.5 0.001 0.001 8.509 8.605 pw_transfer 1547 11.6 0.077 0.103 8.083 8.329 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 7.877 8.132 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.949 8.105 qs_ot_get_orbitals 117 10.6 0.001 0.001 7.778 7.885 make_images_data 5014 15.6 0.070 0.074 6.732 7.794 cp_dbcsr_sm_fm_multiply 37 9.5 0.003 0.003 7.661 7.669 multiply_cannon_metrocomm1 60168 15.6 0.090 0.093 6.056 7.629 grid_integrate_task_list 128 12.3 7.081 7.554 7.081 7.554 hybrid_alltoall_any 5200 16.5 0.291 2.253 5.901 7.178 fft_wrap_pw1pw2_140 523 13.2 0.444 0.510 6.857 7.165 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 6.823 6.919 fft3d_ps 1291 14.7 2.103 2.888 6.666 6.854 mp_alltoall_d11v 2415 14.1 4.195 5.952 4.195 5.952 grid_collocate_task_list 128 9.7 4.559 5.797 4.559 5.797 mp_waitany 16020 13.9 2.613 5.766 2.613 5.766 rs_pw_transfer_RS2PW_140 139 11.5 0.278 0.296 2.097 5.269 potential_pw2rs 128 12.3 0.009 0.011 4.721 4.742 cp_fm_cholesky_decompose 22 10.9 4.614 4.627 4.614 4.627 mp_sum_d 4455 12.2 3.641 4.409 3.641 4.409 calculate_first_density_matrix 1 7.0 0.000 0.000 4.158 4.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.645000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.636364, yerr=6.554287 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 828.497920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4075 57341. MP_Allreduce 11227 947. MP_Sync 170 MP_Alltoall 1969 5618928. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_comm_split 83 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.136 191.377 191.382 qs_mol_dyn_low 1 2.0 0.003 0.004 190.671 190.684 qs_forces 11 3.9 0.003 0.003 190.119 190.119 qs_energies 11 4.9 0.112 0.246 183.328 183.339 scf_env_do_scf 11 5.9 0.001 0.002 165.679 165.689 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 133.015 133.018 velocity_verlet 10 3.0 0.001 0.001 119.642 119.653 dbcsr_multiply_generic 2507 12.6 0.186 0.194 96.672 97.788 qs_scf_new_mos 117 7.6 0.001 0.001 92.896 93.346 qs_scf_loop_do_ot 117 8.6 0.001 0.001 92.895 93.345 ot_scf_mini 117 9.6 0.004 0.004 88.126 88.642 multiply_cannon 2507 13.6 0.477 0.529 77.495 81.329 multiply_cannon_loop 2507 14.6 1.253 1.292 74.321 76.569 ot_mini 117 10.6 0.001 0.001 49.132 49.613 multiply_cannon_multrec 30084 15.6 22.323 26.721 32.016 36.635 mp_waitall_1 214728 16.6 23.381 36.459 23.381 36.459 rebuild_ks_matrix 128 8.3 0.001 0.001 33.397 33.801 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 33.396 33.801 init_scf_loop 11 6.9 0.001 0.004 32.577 32.578 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.070 30.440 prepare_preconditioner 11 7.9 0.000 0.000 28.173 28.256 make_preconditioner 11 8.9 0.000 0.001 28.173 28.256 qs_ot_get_derivative 117 11.6 0.001 0.002 27.309 27.804 multiply_cannon_metrocomm3 30084 15.6 0.093 0.098 15.196 27.557 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.879 27.426 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 21.874 23.009 apply_single 128 13.6 0.001 0.001 21.874 23.009 multiply_cannon_sync_h2d 30084 15.6 19.659 22.405 19.659 22.405 ot_diis_step 117 11.6 0.014 0.015 21.655 21.657 qs_ot_get_p 128 10.4 0.001 0.001 20.835 21.449 cp_fm_cholesky_invert 11 10.9 16.434 16.446 16.434 16.446 qs_ot_p2m_diag 83 11.4 0.188 0.216 16.224 16.259 make_m2s 5014 13.6 0.090 0.097 13.641 15.138 cp_dbcsr_syevd 83 12.4 0.005 0.006 15.085 15.086 make_images 5014 14.6 1.153 1.346 13.430 14.925 sum_up_and_integrate 128 10.3 0.116 0.133 14.731 14.763 integrate_v_rspace 128 11.3 0.003 0.003 14.614 14.650 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.049 14.082 calculate_rho_elec 128 8.7 0.088 0.105 14.048 14.081 cp_fm_diag_elpa 83 13.4 0.000 0.001 12.013 12.022 cp_fm_redistribute_end 83 14.4 7.016 11.952 7.028 11.955 init_scf_run 11 5.9 0.000 0.001 11.831 11.832 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.831 11.832 cp_fm_diag_elpa_base 83 14.4 4.690 11.444 4.909 11.811 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 10.728 11.087 multiply_cannon_metrocomm4 27577 15.6 0.094 0.108 3.669 10.416 make_images_data 5014 15.6 0.067 0.075 8.145 10.075 mp_irecv_dv 69486 16.3 3.478 10.031 3.478 10.031 density_rs2pw 128 9.7 0.006 0.007 7.493 9.883 dbcsr_mm_accdrv_process 62242 16.2 4.497 5.320 9.144 9.764 hybrid_alltoall_any 5200 16.5 0.342 1.491 7.015 9.357 pw_transfer 1547 11.6 0.088 0.105 8.935 9.019 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.708 8.796 wfi_extrapolate 11 7.9 0.001 0.001 8.348 8.348 rs_pw_transfer 1046 11.9 0.014 0.016 5.669 8.071 fft_wrap_pw1pw2_140 523 13.2 0.472 0.525 7.696 7.797 grid_integrate_task_list 128 12.3 7.183 7.597 7.183 7.597 fft3d_ps 1291 14.7 2.751 2.919 7.072 7.109 cp_fm_cholesky_decompose 22 10.9 6.804 6.884 6.804 6.884 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.090 6.841 calculate_dm_sparse 128 9.5 0.001 0.001 6.441 6.598 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.226 6.234 grid_collocate_task_list 128 9.7 4.700 5.861 4.700 5.861 mp_sum_l 7870 13.0 3.850 5.714 3.850 5.714 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.430 5.557 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.303 5.373 potential_pw2rs 128 12.3 0.015 0.017 5.023 5.048 mp_waitany 11748 13.9 2.451 4.979 2.451 4.979 mp_allgather_i34 2507 14.6 1.749 4.563 1.749 4.563 mp_alltoall_d11v 2415 14.1 3.940 4.534 3.940 4.534 rs_pw_transfer_RS2PW_140 139 11.5 0.353 0.383 2.072 4.442 mp_sum_d 4455 12.2 2.812 4.327 2.812 4.327 qs_energies_init_hamiltonians 11 5.9 0.212 0.540 3.834 4.208 dbcsr_complete_redistribute 395 12.7 0.767 0.846 3.103 3.957 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=191.382000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=789.090909, yerr=3.175318 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 2.928533E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 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 934.174720E+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 931531265168 16777216 < size 28672 751619276800 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56951. MP_Allreduce 11296 984. MP_Sync 170 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_comm_split 83 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 0.036 0.083 178.808 178.810 qs_mol_dyn_low 1 2.0 0.003 0.004 177.357 177.370 qs_forces 11 3.9 0.003 0.007 177.258 177.263 qs_energies 11 4.9 0.053 0.054 170.593 170.606 scf_env_do_scf 11 5.9 0.001 0.002 153.738 153.739 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 118.489 118.490 velocity_verlet 10 3.0 0.001 0.001 113.395 113.397 dbcsr_multiply_generic 2507 12.6 0.181 0.185 80.728 81.858 qs_scf_new_mos 117 7.6 0.001 0.001 80.266 80.584 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.265 80.583 ot_scf_mini 117 9.6 0.004 0.005 76.186 76.550 multiply_cannon 2507 13.6 0.502 0.527 61.312 65.552 multiply_cannon_loop 2507 14.6 0.854 0.887 58.135 60.663 ot_mini 117 10.6 0.001 0.001 42.191 42.531 init_scf_loop 11 6.9 0.001 0.004 35.150 35.153 mp_waitall_1 170520 16.6 23.950 33.903 23.950 33.903 rebuild_ks_matrix 128 8.3 0.001 0.001 30.753 31.271 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.019 30.753 31.270 prepare_preconditioner 11 7.9 0.000 0.000 31.103 31.153 make_preconditioner 11 8.9 0.000 0.002 31.103 31.153 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.758 30.191 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.657 28.133 multiply_cannon_multrec 20056 15.6 13.658 16.775 22.411 25.553 multiply_cannon_metrocomm3 20056 15.6 0.060 0.064 14.545 24.536 qs_ot_get_derivative 117 11.6 0.001 0.002 22.451 22.811 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.867 20.775 apply_single 128 13.6 0.001 0.001 19.867 20.775 qs_ot_get_p 128 10.4 0.001 0.002 19.565 20.004 ot_diis_step 117 11.6 0.018 0.018 19.638 19.639 multiply_cannon_sync_h2d 20056 15.6 14.299 15.850 14.299 15.850 qs_ot_p2m_diag 83 11.4 0.266 0.274 15.372 15.381 make_m2s 5014 13.6 0.081 0.085 14.372 15.159 make_images 5014 14.6 1.168 1.259 14.139 14.918 sum_up_and_integrate 128 10.3 0.133 0.144 14.772 14.805 integrate_v_rspace 128 11.3 0.003 0.005 14.639 14.673 cp_fm_cholesky_invert 11 10.9 14.546 14.555 14.546 14.555 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.383 14.422 calculate_rho_elec 128 8.7 0.132 0.147 14.383 14.421 cp_dbcsr_syevd 83 12.4 0.005 0.005 14.363 14.364 cp_fm_diag_elpa 83 13.4 0.000 0.001 11.190 11.191 cp_fm_redistribute_end 83 14.4 4.230 11.129 4.243 11.131 cp_fm_diag_elpa_base 83 14.4 6.461 10.535 6.868 11.013 init_scf_run 11 5.9 0.000 0.001 10.792 10.793 scf_env_initial_rho_setup 11 6.9 0.000 0.001 10.792 10.792 make_images_data 5014 15.6 0.064 0.071 8.809 9.874 pw_transfer 1547 11.6 0.087 0.109 9.177 9.296 density_rs2pw 128 9.7 0.006 0.006 7.284 9.213 hybrid_alltoall_any 5200 16.5 0.436 2.004 7.668 9.211 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.951 9.075 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.795 9.064 multiply_cannon_metrocomm4 17549 15.6 0.062 0.073 3.409 9.063 mp_irecv_dv 50230 16.2 3.287 8.810 3.287 8.810 dbcsr_mm_accdrv_process 41502 16.2 4.294 5.044 8.203 8.348 fft_wrap_pw1pw2_140 523 13.2 0.478 0.527 7.909 8.041 grid_integrate_task_list 128 12.3 7.273 7.718 7.273 7.718 cp_fm_upper_to_full 105 14.5 5.803 7.484 5.803 7.484 cp_fm_cholesky_decompose 22 10.9 7.382 7.411 7.382 7.411 wfi_extrapolate 11 7.9 0.001 0.001 7.365 7.365 fft3d_ps 1291 14.7 2.671 2.891 7.209 7.294 rs_pw_transfer 1046 11.9 0.014 0.015 5.185 7.081 dbcsr_complete_redistribute 395 12.7 1.178 1.206 4.553 6.336 calculate_dm_sparse 128 9.5 0.001 0.001 5.680 5.792 grid_collocate_task_list 128 9.7 4.907 5.776 4.907 5.776 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.485 5.491 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.509 5.174 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.379 5.160 mp_alltoall_d11v 2415 14.1 4.203 4.971 4.203 4.971 potential_pw2rs 128 12.3 0.020 0.022 4.909 4.931 qs_energies_init_hamiltonians 11 5.9 0.127 0.209 4.722 4.725 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.595 4.720 mp_sum_l 7870 13.0 3.232 4.624 3.232 4.624 mp_allgather_i34 2507 14.6 1.699 4.460 1.699 4.460 mp_waitany 11748 13.9 2.341 4.284 2.341 4.284 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 2.326 4.081 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.013 4.048 mp_alltoall_i22 716 14.1 1.968 3.881 1.968 3.881 mp_sum_d 4467 12.2 2.786 3.834 2.786 3.834 rs_pw_transfer_RS2PW_140 139 11.5 0.333 0.351 1.932 3.801 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.609 3.651 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=178.810000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=883.545455, yerr=9.856826 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.136255E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4103 56948. MP_Allreduce 11296 1066. MP_Sync 170 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_comm_split 83 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.032 0.048 186.968 186.971 qs_mol_dyn_low 1 2.0 0.003 0.004 186.546 186.559 qs_forces 11 3.9 0.003 0.009 186.429 186.435 qs_energies 11 4.9 0.002 0.004 179.246 179.259 scf_env_do_scf 11 5.9 0.003 0.008 161.633 161.640 velocity_verlet 10 3.0 0.001 0.001 123.400 123.404 scf_env_do_scf_inner_loop 117 6.6 0.004 0.009 115.648 115.650 dbcsr_multiply_generic 2507 12.6 0.190 0.194 78.595 79.327 qs_scf_new_mos 117 7.6 0.001 0.001 78.186 78.512 qs_scf_loop_do_ot 117 8.6 0.001 0.001 78.185 78.512 ot_scf_mini 117 9.6 0.004 0.005 73.663 73.977 multiply_cannon 2507 13.6 0.558 0.594 54.484 57.243 multiply_cannon_loop 2507 14.6 1.181 1.205 50.847 52.431 init_scf_loop 11 6.9 0.025 0.154 45.860 45.860 ot_mini 117 10.6 0.001 0.001 41.935 42.275 prepare_preconditioner 11 7.9 0.000 0.000 41.698 41.762 make_preconditioner 11 8.9 0.011 0.046 41.698 41.762 make_full_inverse_cholesky 11 9.9 0.000 0.000 35.420 40.400 multiply_cannon_multrec 30084 15.6 14.099 19.804 25.982 31.189 rebuild_ks_matrix 128 8.3 0.001 0.001 29.929 30.206 qs_ks_build_kohn_sham_matrix 128 9.3 0.030 0.126 29.929 30.206 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.973 27.222 mp_waitall_1 147882 16.7 16.766 26.233 16.766 26.233 qs_ot_get_derivative 117 11.6 0.001 0.002 22.348 22.661 make_m2s 5014 13.6 0.096 0.102 19.772 20.825 make_images 5014 14.6 1.960 2.255 19.464 20.516 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 19.066 19.603 apply_single 128 13.6 0.001 0.001 19.065 19.603 ot_diis_step 117 11.6 0.017 0.018 19.462 19.463 qs_ot_get_p 128 10.4 0.001 0.001 18.077 18.452 cp_fm_upper_to_full 105 14.7 11.019 16.262 11.019 16.262 cp_fm_cholesky_invert 11 10.9 15.745 15.754 15.745 15.754 sum_up_and_integrate 128 10.3 0.140 0.151 14.741 14.768 multiply_cannon_metrocomm3 30084 15.6 0.046 0.049 6.348 14.751 integrate_v_rspace 128 11.3 0.003 0.005 14.601 14.633 qs_rho_update_rho_low 128 7.7 0.001 0.001 14.537 14.577 calculate_rho_elec 128 8.7 0.175 0.190 14.536 14.576 qs_ot_p2m_diag 83 11.4 0.343 0.391 14.085 14.137 cp_dbcsr_syevd 83 12.4 0.005 0.005 12.831 12.832 multiply_cannon_sync_h2d 30084 15.6 11.759 12.766 11.759 12.766 dbcsr_complete_redistribute 395 12.7 1.518 1.639 8.750 12.434 make_images_data 5014 15.6 0.066 0.070 10.528 12.235 dbcsr_mm_accdrv_process 62264 16.2 7.272 8.374 11.458 11.955 init_scf_run 11 5.9 0.000 0.001 11.714 11.716 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.714 11.715 copy_fm_to_dbcsr 209 11.7 0.002 0.002 7.428 11.119 hybrid_alltoall_any 5200 16.5 0.526 2.206 9.494 11.075 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.244 9.855 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.811 9.813 cp_fm_redistribute_end 83 14.4 1.698 9.751 1.712 9.757 cp_fm_diag_elpa_base 83 14.4 7.438 9.141 8.015 9.633 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.214 9.440 pw_transfer 1547 11.6 0.088 0.106 9.356 9.417 fft_wrap_pw1pw2 1291 12.7 0.010 0.012 9.128 9.200 mp_alltoall_i22 716 14.1 5.495 9.181 5.495 9.181 density_rs2pw 128 9.7 0.006 0.006 7.040 8.635 fft_wrap_pw1pw2_140 523 13.2 0.483 0.494 8.131 8.228 grid_integrate_task_list 128 12.3 7.509 7.826 7.509 7.826 wfi_extrapolate 11 7.9 0.001 0.001 7.534 7.534 cp_fm_cholesky_decompose 22 10.9 7.386 7.488 7.386 7.488 fft3d_ps 1291 14.7 2.775 2.859 7.344 7.405 multiply_cannon_metrocomm4 25070 15.6 0.075 0.085 2.802 7.216 mp_irecv_dv 76098 16.2 2.657 6.952 2.657 6.952 rs_pw_transfer 1046 11.9 0.013 0.015 4.626 6.281 calculate_dm_sparse 128 9.5 0.001 0.001 6.175 6.256 grid_collocate_task_list 128 9.7 5.068 5.903 5.068 5.903 mp_alltoall_d11v 2415 14.1 4.633 5.390 4.633 5.390 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.317 5.361 potential_pw2rs 128 12.3 0.022 0.023 4.621 4.632 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 4.432 4.436 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.314 4.402 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.293 4.380 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.149 4.200 calculate_first_density_matrix 1 7.0 0.001 0.005 3.979 3.983 mp_waitany 14952 13.9 2.245 3.822 2.245 3.822 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=186.971000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1068.181818, yerr=22.290716 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.503367E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4010 58095. MP_Allreduce 11022 1174. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.021 0.053 173.412 173.419 qs_mol_dyn_low 1 2.0 0.003 0.004 172.974 172.987 qs_forces 11 3.9 0.003 0.003 172.882 172.886 qs_energies 11 4.9 0.001 0.002 165.323 165.331 scf_env_do_scf 11 5.9 0.001 0.001 147.377 147.382 velocity_verlet 10 3.0 0.001 0.001 113.217 113.221 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 112.004 112.005 dbcsr_multiply_generic 2507 12.6 0.182 0.187 74.618 75.068 qs_scf_new_mos 117 7.6 0.001 0.001 73.131 73.220 qs_scf_loop_do_ot 117 8.6 0.001 0.001 73.130 73.219 ot_scf_mini 117 9.6 0.004 0.004 68.719 68.792 multiply_cannon 2507 13.6 0.583 0.620 55.103 59.420 multiply_cannon_loop 2507 14.6 0.446 0.470 50.597 51.411 ot_mini 117 10.6 0.001 0.001 39.382 39.454 init_scf_loop 11 6.9 0.000 0.000 35.229 35.230 mp_waitall_1 125778 16.7 25.687 32.812 25.687 32.812 prepare_preconditioner 11 7.9 0.000 0.000 31.245 31.276 make_preconditioner 11 8.9 0.000 0.000 31.245 31.276 rebuild_ks_matrix 128 8.3 0.001 0.001 30.100 30.170 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 30.099 30.169 make_full_inverse_cholesky 11 9.9 0.000 0.000 29.116 29.398 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.341 27.401 multiply_cannon_multrec 10028 15.6 10.426 14.272 18.592 21.783 ot_diis_step 117 11.6 0.020 0.021 19.789 19.789 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.336 19.610 apply_single 128 13.6 0.001 0.001 19.335 19.609 qs_ot_get_derivative 117 11.6 0.002 0.002 19.524 19.609 multiply_cannon_metrocomm3 10028 15.6 0.023 0.025 12.262 19.149 make_m2s 5014 13.6 0.066 0.071 15.977 18.217 make_images 5014 14.6 2.273 2.723 15.670 17.905 cp_fm_cholesky_invert 11 10.9 17.761 17.766 17.761 17.766 qs_ot_get_p 128 10.4 0.001 0.001 16.546 16.628 qs_rho_update_rho_low 128 7.7 0.001 0.001 15.471 15.503 calculate_rho_elec 128 8.7 0.259 0.268 15.470 15.502 sum_up_and_integrate 128 10.3 0.181 0.191 15.141 15.184 integrate_v_rspace 128 11.3 0.004 0.004 14.960 15.012 qs_ot_p2m_diag 83 11.4 0.496 0.503 12.882 12.900 make_images_data 5014 15.6 0.054 0.063 9.789 12.341 multiply_cannon_sync_h2d 10028 15.6 11.614 12.065 11.614 12.065 hybrid_alltoall_any 5200 16.5 0.839 3.797 9.643 11.845 cp_dbcsr_syevd 83 12.4 0.005 0.005 11.766 11.768 init_scf_run 11 5.9 0.000 0.001 11.189 11.190 scf_env_initial_rho_setup 11 6.9 0.000 0.000 11.189 11.189 pw_transfer 1547 11.6 0.085 0.093 9.909 9.945 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 9.685 9.725 cp_fm_diag_elpa 83 13.4 0.000 0.000 8.803 8.812 cp_fm_diag_elpa_base 83 14.4 8.564 8.639 8.796 8.805 fft_wrap_pw1pw2_140 523 13.2 0.495 0.514 8.566 8.618 dbcsr_mm_accdrv_process 20762 16.1 2.805 3.583 7.793 8.437 grid_integrate_task_list 128 12.3 7.760 8.150 7.760 8.150 cp_fm_cholesky_decompose 22 10.9 7.815 7.917 7.815 7.917 density_rs2pw 128 9.7 0.005 0.006 7.030 7.916 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.833 7.889 fft3d_ps 1291 14.7 2.708 2.801 7.848 7.867 wfi_extrapolate 11 7.9 0.001 0.001 7.399 7.399 multiply_cannon_metrocomm1 10028 15.6 0.028 0.029 4.424 7.379 mp_allgather_i34 2507 14.6 2.712 7.016 2.712 7.016 calculate_dm_sparse 128 9.5 0.001 0.001 6.870 6.955 grid_collocate_task_list 128 9.7 5.396 6.053 5.396 6.053 mp_alltoall_d11v 2415 14.1 5.019 5.673 5.019 5.673 dbcsr_complete_redistribute 395 12.7 2.121 2.201 5.142 5.495 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.268 5.269 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.177 5.192 rs_pw_transfer 1046 11.9 0.012 0.013 4.048 4.962 potential_pw2rs 128 12.3 0.026 0.026 4.555 4.567 multiply_cannon_metrocomm4 7521 15.6 0.024 0.027 1.852 4.289 mp_irecv_dv 28860 15.9 1.815 4.214 1.815 4.214 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.122 4.148 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.554 3.863 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.663 3.703 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.664 3.674 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.366 3.673 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.505 3.589 calculate_first_density_matrix 1 7.0 0.000 0.000 3.577 3.578 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.465 3.490 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=173.419000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1392.909091, yerr=52.199712 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 11.785520E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 1979552 0.0% 0.0% 100.0% average stack size 0.0 0.0 3438.3 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 2.929943E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101080 MPI messages size (bytes): total size 1.144901E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.326685E+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 45568 35366371328 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726591885696 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 58831. MP_Allreduce 11057 1512. MP_Sync 87 MP_Alltoall 1724 36993716. 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 0.024 0.052 298.122 298.127 qs_mol_dyn_low 1 2.0 0.003 0.005 297.508 297.523 qs_forces 11 3.9 0.003 0.003 297.398 297.401 qs_energies 11 4.9 0.002 0.002 288.003 288.016 scf_env_do_scf 11 5.9 0.001 0.001 264.992 265.002 velocity_verlet 10 3.0 0.001 0.002 214.289 214.300 scf_env_do_scf_inner_loop 118 6.6 0.004 0.009 141.346 141.347 init_scf_loop 11 6.9 0.000 0.000 123.386 123.388 prepare_preconditioner 11 7.9 0.000 0.000 118.313 118.341 make_preconditioner 11 8.9 0.000 0.000 118.313 118.341 make_full_inverse_cholesky 11 9.9 0.000 0.000 94.711 115.490 qs_scf_new_mos 118 7.6 0.001 0.001 89.892 90.035 qs_scf_loop_do_ot 118 8.6 0.001 0.001 89.891 90.035 ot_scf_mini 118 9.6 0.004 0.004 85.091 85.114 dbcsr_multiply_generic 2527 12.6 0.214 0.221 84.130 84.626 cp_fm_upper_to_full 105 14.8 52.559 75.149 52.559 75.149 multiply_cannon 2527 13.6 0.711 0.770 59.973 60.479 multiply_cannon_loop 2527 14.6 0.475 0.495 56.407 57.585 ot_mini 118 10.6 0.001 0.001 44.492 44.519 dbcsr_complete_redistribute 395 12.7 3.977 4.022 29.219 41.940 rebuild_ks_matrix 129 8.3 0.001 0.001 38.740 38.782 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.018 38.739 38.781 copy_fm_to_dbcsr 209 11.7 0.001 0.002 25.848 38.554 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 23.557 36.163 qs_ks_update_qs_env 129 7.6 0.001 0.001 35.706 35.744 mp_alltoall_i22 718 14.1 21.326 33.916 21.326 33.916 cp_fm_cholesky_invert 11 10.9 33.067 33.075 33.067 33.075 mp_waitall_1 104500 16.8 28.219 32.098 28.219 32.098 qs_ot_get_p 129 10.4 0.001 0.001 25.067 25.103 qs_ot_get_derivative 118 11.6 0.002 0.002 24.217 24.239 qs_ot_p2m_diag 83 11.4 0.880 0.885 21.052 21.084 qs_rho_update_rho_low 129 7.7 0.001 0.001 20.853 20.880 calculate_rho_elec 129 8.7 0.487 0.488 20.853 20.880 make_m2s 5054 13.6 0.077 0.081 19.652 20.537 ot_diis_step 118 11.6 0.022 0.023 20.246 20.247 multiply_cannon_metrocomm3 10108 15.6 0.024 0.024 19.109 20.189 make_images 5054 14.6 3.804 3.919 19.169 20.057 sum_up_and_integrate 129 10.3 0.324 0.325 19.977 20.051 integrate_v_rspace 129 11.3 0.004 0.004 19.652 19.726 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.358 19.616 apply_single 129 13.6 0.001 0.001 19.357 19.615 cp_dbcsr_syevd 83 12.4 0.006 0.006 19.336 19.338 multiply_cannon_multrec 10108 15.6 10.629 12.490 18.802 18.944 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.185 16.186 cp_fm_diag_elpa_base 83 14.4 11.875 13.428 16.181 16.181 multiply_cannon_sync_h2d 10108 15.6 15.785 15.792 15.785 15.792 init_scf_run 11 5.9 0.000 0.001 12.995 12.996 scf_env_initial_rho_setup 11 6.9 0.000 0.001 12.995 12.995 pw_transfer 1559 11.6 0.094 0.095 12.323 12.333 hybrid_alltoall_any 5240 16.5 1.310 3.059 10.491 12.291 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 12.086 12.096 make_images_data 5054 15.6 0.063 0.068 10.330 12.074 fft_wrap_pw1pw2_140 527 13.2 0.545 0.547 10.748 10.762 fft3d_ps 1301 14.7 2.758 2.764 10.112 10.120 dbcsr_mm_accdrv_process 20926 16.1 3.823 5.667 7.933 9.862 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.343 9.364 wfi_extrapolate 11 7.9 0.001 0.001 9.107 9.107 mp_alltoall_d11v 2423 14.1 8.216 9.032 8.216 9.032 grid_integrate_task_list 129 12.3 8.651 8.841 8.651 8.841 cp_fm_cholesky_decompose 22 10.9 8.791 8.805 8.791 8.805 density_rs2pw 129 9.7 0.005 0.005 8.326 8.453 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.050 8.051 calculate_dm_sparse 129 9.5 0.001 0.001 7.394 7.536 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.413 6.475 grid_collocate_task_list 129 9.7 6.412 6.454 6.412 6.454 rs_scatter_matrices 140 9.7 3.698 4.719 6.093 6.324 copy_dbcsr_to_fm 186 11.8 0.004 0.004 6.014 6.066 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=298.127000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2606.545455, yerr=171.078591 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 1.262207E+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 0.034 0.070 85.363 85.364 qs_energies 1 2.0 0.000 0.001 84.630 84.644 ls_scf 1 3.0 0.000 0.000 83.603 83.616 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.514 72.709 multiply_cannon 111 7.7 0.017 0.020 55.820 57.032 multiply_cannon_loop 111 8.7 0.209 0.225 52.411 53.711 ls_scf_main 1 4.0 0.000 0.000 52.197 52.207 density_matrix_trs4 2 5.0 0.002 0.003 46.715 46.790 ls_scf_init_scf 1 4.0 0.018 0.024 28.381 28.387 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.242 27.301 mp_waitall_1 11031 10.9 22.408 25.313 22.408 25.313 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.998 25.017 multiply_cannon_multrec 2664 9.7 8.177 8.918 15.482 17.211 multiply_cannon_sync_h2d 2664 9.7 13.613 15.334 13.613 15.334 make_m2s 222 7.7 0.008 0.011 13.014 13.523 make_images 222 8.7 0.099 0.109 12.992 13.503 multiply_cannon_metrocomm1 2664 9.7 0.009 0.011 9.709 12.124 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.439 8.479 make_images_data 222 9.7 0.004 0.005 7.591 8.276 dbcsr_mm_accdrv_process 4760 10.4 0.508 0.638 6.923 7.923 hybrid_alltoall_any 227 10.6 0.217 1.833 6.558 7.766 dbcsr_mm_accdrv_process_sort 4760 11.4 6.216 7.091 6.216 7.091 calculate_norms 4752 9.8 5.517 6.138 5.517 6.138 mp_sum_l 807 5.4 3.148 5.210 3.148 5.210 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.991 5.160 multiply_cannon_metrocomm4 2442 9.7 0.012 0.016 2.049 4.684 mp_irecv_dv 6231 10.9 2.030 4.651 2.030 4.651 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.363 4.101 make_images_sizes 222 9.7 0.000 0.000 0.780 3.384 mp_alltoall_i44 222 10.7 0.780 3.384 0.780 3.384 arnoldi_extremal 4 6.8 0.000 0.000 3.216 3.244 arnoldi_normal_ev 4 7.8 0.003 0.017 3.216 3.244 build_subspace 16 8.4 0.009 0.012 3.117 3.119 ls_scf_post 1 4.0 0.000 0.000 3.026 3.040 ls_scf_store_result 1 5.0 0.000 0.000 2.852 2.888 dbcsr_special_finalize 555 9.7 0.005 0.006 2.290 2.719 dbcsr_merge_single_wm 555 10.7 0.452 0.600 2.282 2.711 make_images_pack 222 9.7 2.207 2.617 2.209 2.618 dbcsr_matrix_vector_mult 304 9.0 0.005 0.013 2.303 2.545 dbcsr_matrix_vector_mult_local 304 10.0 2.064 2.456 2.066 2.458 dbcsr_sort_data 658 11.4 2.088 2.443 2.088 2.443 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.278 2.347 buffer_matrices_ensure_size 222 8.7 1.757 2.102 1.757 2.102 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.790 1.802 rebuild_ks_matrix 3 7.3 0.000 0.000 1.780 1.793 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.012 1.780 1.792 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.364000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1130.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 2.100183E+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 0.048 0.087 91.356 91.358 qs_energies 1 2.0 0.001 0.012 90.742 90.753 ls_scf 1 3.0 0.002 0.019 89.406 89.433 dbcsr_multiply_generic 111 6.7 0.015 0.016 75.039 75.433 multiply_cannon 111 7.7 0.028 0.038 53.203 56.752 ls_scf_main 1 4.0 0.000 0.001 54.916 54.928 multiply_cannon_loop 111 8.7 0.116 0.122 49.907 52.562 density_matrix_trs4 2 5.0 0.003 0.025 49.169 49.338 ls_scf_init_scf 1 4.0 0.028 0.056 31.008 31.032 mp_waitall_1 9105 10.9 21.144 30.339 21.144 30.339 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.561 29.729 multiply_cannon_multrec 1332 9.7 13.100 16.817 22.298 27.145 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.030 27.040 multiply_cannon_metrocomm3 1332 9.7 0.006 0.007 11.836 21.152 make_m2s 222 7.7 0.006 0.007 15.297 15.963 make_images 222 8.7 1.576 1.954 15.267 15.933 dbcsr_mm_accdrv_process 4041 10.4 0.277 0.453 8.798 10.407 dbcsr_mm_accdrv_process_sort 4041 11.4 8.394 9.955 8.394 9.955 make_images_data 222 9.7 0.004 0.004 8.799 9.744 hybrid_alltoall_any 227 10.6 0.522 2.474 8.309 9.321 mp_sum_l 807 5.4 5.239 7.979 5.239 7.979 multiply_cannon_metrocomm4 1221 9.7 0.006 0.008 3.228 7.935 mp_irecv_dv 3311 11.0 3.209 7.880 3.209 7.880 calculate_norms 2376 9.8 5.989 6.792 5.989 6.792 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.018 6.319 multiply_cannon_sync_h2d 1332 9.7 4.840 6.172 4.840 6.172 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.986 5.184 arnoldi_extremal 4 6.8 0.000 0.000 4.612 4.627 arnoldi_normal_ev 4 7.8 0.001 0.005 4.612 4.627 build_subspace 16 8.4 0.014 0.021 4.353 4.357 ls_scf_post 1 4.0 0.000 0.001 3.479 3.483 dbcsr_matrix_vector_mult 304 9.0 0.009 0.021 3.119 3.343 ls_scf_store_result 1 5.0 0.000 0.000 3.195 3.305 dbcsr_matrix_vector_mult_local 304 10.0 2.739 3.212 2.741 3.214 ls_scf_dm_to_ks 2 5.0 0.001 0.016 2.623 2.716 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 1.143 2.690 mp_allgather_i34 111 8.7 0.994 2.492 0.994 2.492 make_images_pack 222 9.7 2.032 2.433 2.034 2.435 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.059 2.078 dbcsr_sort_data 436 11.2 1.841 2.070 1.841 2.070 rebuild_ks_matrix 3 7.3 0.000 0.000 2.047 2.065 qs_ks_build_kohn_sham_matrix 3 8.3 0.007 0.037 2.047 2.065 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=91.358000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1719.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 2.711978E+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 0.032 0.051 94.827 94.829 qs_energies 1 2.0 0.000 0.001 94.207 94.214 ls_scf 1 3.0 0.000 0.002 92.802 92.812 dbcsr_multiply_generic 111 6.7 0.015 0.016 77.460 77.745 ls_scf_main 1 4.0 0.000 0.001 58.248 58.254 multiply_cannon 111 7.7 0.049 0.123 53.263 57.048 multiply_cannon_loop 111 8.7 0.100 0.113 49.739 54.119 density_matrix_trs4 2 5.0 0.002 0.005 52.289 52.462 mp_waitall_1 7281 11.0 24.388 34.731 24.388 34.731 ls_scf_init_scf 1 4.0 0.010 0.036 30.941 30.952 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.732 29.823 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.300 27.319 multiply_cannon_multrec 888 9.7 12.586 15.419 21.089 24.573 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.362 23.950 make_m2s 222 7.7 0.006 0.007 17.180 18.394 make_images 222 8.7 1.966 2.285 17.142 18.354 hybrid_alltoall_any 227 10.6 0.616 2.809 9.471 10.888 make_images_data 222 9.7 0.003 0.004 9.810 10.832 mp_sum_l 807 5.4 5.459 9.451 5.459 9.451 dbcsr_mm_accdrv_process 3754 10.4 0.250 0.409 8.030 9.219 dbcsr_mm_accdrv_process_sort 3754 11.4 7.662 8.810 7.662 8.810 multiply_cannon_sync_h2d 888 9.7 6.081 7.619 6.081 7.619 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.240 7.530 multiply_cannon_metrocomm1 888 9.7 0.002 0.003 3.836 7.228 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.490 7.050 mp_irecv_dv 2335 11.1 2.476 7.005 2.476 7.005 arnoldi_extremal 4 6.8 0.000 0.000 5.051 5.084 arnoldi_normal_ev 4 7.8 0.001 0.006 5.051 5.084 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.882 5.082 build_subspace 16 8.4 0.014 0.020 4.738 4.743 calculate_norms 1584 9.8 4.261 4.725 4.261 4.725 mp_allgather_i34 111 8.7 1.366 3.906 1.366 3.906 dbcsr_matrix_vector_mult 304 9.0 0.009 0.020 3.427 3.757 ls_scf_post 1 4.0 0.000 0.001 3.612 3.617 dbcsr_matrix_vector_mult_local 304 10.0 3.030 3.605 3.032 3.607 ls_scf_store_result 1 5.0 0.000 0.000 3.316 3.415 ls_scf_dm_to_ks 2 5.0 0.000 0.001 2.862 2.970 make_images_sizes 222 9.7 0.000 0.000 1.052 2.310 mp_alltoall_i44 222 10.7 1.052 2.309 1.052 2.309 dbcsr_sort_data 325 11.1 1.877 2.205 1.877 2.205 make_images_pack 222 9.7 1.815 2.127 1.818 2.130 dbcsr_finalize 304 7.8 0.026 0.032 1.607 1.952 dbcsr_data_release 9322 10.9 1.313 1.927 1.313 1.927 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.921 1.923 rebuild_ks_matrix 3 7.3 0.000 0.000 1.903 1.906 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.903 1.906 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=94.829000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2213.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 3.291939E+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 0.041 0.069 97.229 97.232 qs_energies 1 2.0 0.000 0.002 96.353 96.360 ls_scf 1 3.0 0.000 0.000 94.692 94.697 dbcsr_multiply_generic 111 6.7 0.017 0.022 78.414 78.696 ls_scf_main 1 4.0 0.000 0.001 58.868 58.870 multiply_cannon 111 7.7 0.064 0.118 51.542 55.400 density_matrix_trs4 2 5.0 0.002 0.005 52.734 52.850 multiply_cannon_loop 111 8.7 0.114 0.127 46.559 49.690 ls_scf_init_scf 1 4.0 0.001 0.002 32.569 32.571 ls_scf_init_matrix_S 1 5.0 0.000 0.001 31.324 31.391 mp_waitall_1 6369 11.0 22.892 29.766 22.892 29.766 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.865 28.875 multiply_cannon_multrec 1332 9.7 14.170 17.395 22.010 24.628 make_m2s 222 7.7 0.007 0.008 21.237 22.574 make_images 222 8.7 3.145 3.608 21.186 22.525 multiply_cannon_metrocomm3 1332 9.7 0.003 0.003 9.403 16.703 make_images_data 222 9.7 0.004 0.004 11.896 13.449 hybrid_alltoall_any 227 10.6 0.797 3.832 11.214 12.887 dbcsr_mm_accdrv_process 3641 10.4 0.202 0.410 7.480 9.002 dbcsr_mm_accdrv_process_sort 3641 11.4 7.109 8.596 7.109 8.596 mp_sum_l 807 5.4 4.042 7.111 4.042 7.111 multiply_cannon_sync_h2d 1332 9.7 5.551 6.237 5.551 6.237 multiply_cannon_metrocomm4 1110 9.7 0.004 0.006 2.063 5.968 mp_irecv_dv 3229 10.9 2.040 5.890 2.040 5.890 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.099 5.556 arnoldi_extremal 4 6.8 0.000 0.000 5.231 5.248 arnoldi_normal_ev 4 7.8 0.001 0.005 5.231 5.248 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.426 4.985 build_subspace 16 8.4 0.014 0.021 4.898 4.906 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.521 4.706 calculate_norms 2376 9.8 4.171 4.557 4.171 4.557 mp_allgather_i34 111 8.7 2.112 4.392 2.112 4.392 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.596 3.920 dbcsr_matrix_vector_mult_local 304 10.0 3.214 3.746 3.216 3.748 dbcsr_sort_data 658 11.4 3.070 3.556 3.070 3.556 ls_scf_post 1 4.0 0.000 0.000 3.254 3.261 dbcsr_special_finalize 555 9.7 0.006 0.007 2.806 3.236 dbcsr_merge_single_wm 555 10.7 0.534 0.670 2.798 3.228 ls_scf_dm_to_ks 2 5.0 0.000 0.001 3.061 3.127 ls_scf_store_result 1 5.0 0.000 0.000 2.998 3.057 dbcsr_data_release 10477 10.7 1.594 2.456 1.594 2.456 dbcsr_finalize 304 7.8 0.049 0.061 1.806 2.069 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.989 1.991 rebuild_ks_matrix 3 7.3 0.000 0.000 1.965 1.967 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.003 1.965 1.967 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.232000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2695.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 4.625080E+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 99.573 99.574 qs_energies 1 2.0 0.000 0.000 98.770 98.776 ls_scf 1 3.0 0.000 0.000 96.810 96.816 dbcsr_multiply_generic 111 6.7 0.017 0.018 78.024 78.240 ls_scf_main 1 4.0 0.000 0.000 62.318 62.318 multiply_cannon 111 7.7 0.074 0.129 55.295 60.836 density_matrix_trs4 2 5.0 0.002 0.003 55.027 55.115 multiply_cannon_loop 111 8.7 0.069 0.076 50.901 52.472 mp_waitall_1 5436 11.0 26.692 31.904 26.692 31.904 ls_scf_init_scf 1 4.0 0.000 0.001 30.872 30.878 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.509 29.544 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 27.318 27.325 multiply_cannon_multrec 444 9.7 13.977 16.451 21.002 22.717 make_m2s 222 7.7 0.004 0.005 17.845 20.211 make_images 222 8.7 3.721 4.430 17.783 20.151 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 11.190 16.553 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 6.268 14.277 make_images_data 222 9.7 0.003 0.004 10.036 12.423 hybrid_alltoall_any 227 10.6 0.789 3.762 9.899 12.319 dbcsr_mm_accdrv_process 3003 10.4 0.176 0.349 6.731 7.838 multiply_cannon_sync_h2d 444 9.7 6.569 7.583 6.569 7.583 dbcsr_mm_accdrv_process_sort 3003 11.4 6.417 7.497 6.417 7.497 mp_allgather_i34 111 8.7 2.631 7.001 2.631 7.001 arnoldi_extremal 4 6.8 0.000 0.000 5.833 5.843 arnoldi_normal_ev 4 7.8 0.016 0.037 5.833 5.842 build_subspace 16 8.4 0.015 0.020 5.418 5.427 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.533 4.704 mp_sum_l 807 5.4 2.923 4.593 2.923 4.593 dbcsr_matrix_vector_mult 304 9.0 0.011 0.020 4.182 4.384 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.676 4.366 mp_irecv_dv 1241 11.2 1.658 4.342 1.658 4.342 dbcsr_matrix_vector_mult_local 304 10.0 3.728 4.199 3.730 4.201 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.730 3.805 calculate_norms 792 9.8 3.561 3.689 3.561 3.689 ls_scf_post 1 4.0 0.000 0.000 3.620 3.626 make_images_sizes 222 9.7 0.000 0.000 1.163 3.518 mp_alltoall_i44 222 10.7 1.163 3.518 1.163 3.518 ls_scf_store_result 1 5.0 0.000 0.000 3.405 3.441 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.000 3.395 dbcsr_finalize 304 7.8 0.062 0.078 2.195 2.278 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.260 2.261 rebuild_ks_matrix 3 7.3 0.000 0.000 2.228 2.229 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.228 2.229 dbcsr_merge_all 275 8.9 0.474 0.531 2.047 2.114 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=99.574000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3634.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/15b8176f3f1ac853f85861d42f1522ebe17008b3_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 8.715342E+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 0.081 0.120 108.687 108.694 qs_energies 1 2.0 0.000 0.000 107.246 107.258 ls_scf 1 3.0 0.000 0.000 104.310 104.322 dbcsr_multiply_generic 111 6.7 0.023 0.028 77.651 77.793 ls_scf_main 1 4.0 0.000 0.000 66.477 66.492 density_matrix_trs4 2 5.0 0.002 0.003 57.176 57.226 multiply_cannon 111 7.7 0.149 0.226 50.099 52.106 multiply_cannon_loop 111 8.7 0.067 0.069 46.618 47.529 ls_scf_init_scf 1 4.0 0.001 0.001 34.121 34.168 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.436 32.459 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 29.621 29.634 mp_waitall_1 4527 11.1 22.347 26.152 22.347 26.152 make_m2s 222 7.7 0.005 0.005 23.872 24.882 make_images 222 8.7 4.582 4.954 23.766 24.774 multiply_cannon_multrec 444 9.7 17.927 18.772 22.519 23.149 hybrid_alltoall_any 227 10.6 1.660 3.622 12.913 15.592 make_images_data 222 9.7 0.003 0.003 13.121 15.572 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.623 11.111 multiply_cannon_sync_h2d 444 9.7 8.847 8.890 8.847 8.890 arnoldi_extremal 4 6.8 0.000 0.000 7.430 7.440 arnoldi_normal_ev 4 7.8 0.003 0.009 7.430 7.440 build_subspace 16 8.4 0.026 0.036 6.886 6.896 dbcsr_matrix_vector_mult 304 9.0 0.016 0.033 5.505 5.690 dbcsr_matrix_vector_mult_local 304 10.0 5.091 5.430 5.094 5.432 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.232 5.326 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.987 5.245 dbcsr_mm_accdrv_process 1814 10.4 0.255 0.322 4.429 4.561 dbcsr_mm_accdrv_process_sort 1814 11.4 4.128 4.256 4.128 4.256 ls_scf_post 1 4.0 0.000 0.000 3.712 3.723 mp_allgather_i34 111 8.7 1.134 3.581 1.134 3.581 make_images_sizes 222 9.7 0.000 0.000 1.427 3.498 mp_alltoall_i44 222 10.7 1.427 3.497 1.427 3.497 ls_scf_store_result 1 5.0 0.000 0.000 3.413 3.426 calculate_norms 792 9.8 3.245 3.277 3.245 3.277 dbcsr_finalize 304 7.8 0.082 0.089 3.088 3.166 dbcsr_merge_all 275 8.9 0.893 0.915 2.874 2.945 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.905 2.906 dbcsr_complete_redistribute 5 7.6 1.434 1.471 2.767 2.884 dbcsr_data_release 12724 10.6 2.334 2.866 2.334 2.866 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.594 2.643 rebuild_ks_matrix 3 7.3 0.000 0.000 2.528 2.577 qs_ks_build_kohn_sham_matrix 3 8.3 0.015 0.062 2.528 2.577 matrix_ls_to_qs 2 6.0 0.000 0.000 2.407 2.532 dbcsr_sort_data 325 11.1 2.441 2.501 2.441 2.501 dbcsr_new_transposed 4 7.5 0.243 0.249 2.248 2.259 dbcsr_frobenius_norm 74 6.6 2.058 2.130 2.192 2.231 dbcsr_add_d 103 6.2 0.000 0.000 2.133 2.204 dbcsr_add_anytype 103 7.2 0.859 0.891 2.133 2.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.694000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6884.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 15b8176f3f1ac853f85861d42f1522ebe17008b3 Summary: empty Status: OK