=== 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: 5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2023.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/01 job id: 48399482 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/02 job id: 48399483 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/03 job id: 48399484 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/04 job id: 48399485 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/05 job id: 48399486 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/06 job id: 48399487 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/07 job id: 48399488 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/08 job id: 48399489 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/09 job id: 48399490 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/10 job id: 48399491 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/11 job id: 48399492 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/12 job id: 48399493 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/13 job id: 48399494 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/14 job id: 48399495 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/15 job id: 48399496 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/16 job id: 48399497 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/17 job id: 48399498 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/18 job id: 48399499 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/19 job id: 48399500 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/20 job id: 48399501 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/21 job id: 48399502 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/22 job id: 48399504 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/23 job id: 48399505 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/24 job id: 48399506 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/25 job id: 48399507 --- 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/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/26 job id: 48399508 --- Point --- name: 510 plot: h2o_32_nrep3_ls regex: CP2K label: (8n/1r/12t) --- Point --- name: 511 plot: h2o_32_nrep3_ls_mem regex: Estimated peak process memory label: (8n/1r/12t) ~~~~~~~ END TEST ~~~~~~~ ~~~~~~~~~ TEST ~~~~~~~~~ description: 512 H2O (4 NVE MD steps on 64 nodes) input file: benchmarks/QS/00512_H2O/H2O-512_md.inp required files: [] output file: result.log # nodes = 64 # ranks/node = 12 # threads/rank = 1 nrepeat = 1 time[min] = 10 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/27 job id: 48399509 --- Point --- name: 601 plot: h2o_512_md regex: CP2K label: (64n/12r/1t) --- Point --- name: 602 plot: h2o_512_md_mem regex: Estimated peak process memory label: (64n/12r/1t) ~~~~~~~ END TEST ~~~~~~~ === END TESTS (description) === ===== PLOTS (description) ===== ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_ri_rpa_mp2_mem", title="32 H2O molecules (RI-MP2, RI-RPA)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_64_md_mem", title="64 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_128_md_mem", title="128 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_256_md_mem", title="256 H2O molecules (10 MD steps)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_32_nrep3_ls_mem", title="864 H2O molecules (LS SCF)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Time [s]" ~~~~~~~~~ PLOT ~~~~~~~~~ Plot: name="h2o_512_md_mem", title="512 H2O (4 NVE MD steps on 64 nodes)", xlabel="Revision", ylabel="Est. peak process memory [MiB]" === END PLOTS (description) === ============ RESULTS ============ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 132.812 132.813 farming_run 1 2.0 132.130 132.133 132.782 132.786 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.490026E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 2592 MPI messages size (bytes): total size 1.140326E+09 min size 0.000000E+00 max size 1.663488E+06 average size 439.940750E+03 MPI breakdown and total messages size (bytes): size <= 128 132 0 128 < size <= 8192 348 2850816 8192 < size <= 32768 0 0 32768 < size <= 131072 1536 179306496 131072 < size <= 4194304 576 958169088 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 2308 54. MP_Alltoall 4670 822215. MP_ISend 2604 90577. MP_IRecv 2604 90574. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 230 1134128. MP_Allreduce 571 1938539. MP_Sync 25 MP_Alltoall 38 9316958. MP_SendRecv 120 384007. MP_ISendRecv 45 235435. MP_Wait 191 MP_comm_split 10 MP_ISend 127 3867574. MP_IRecv 127 3866554. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.006 0.027 114.316 114.317 qs_energies 1 2.0 0.000 0.000 114.001 114.009 mp2_main 1 3.0 0.000 0.000 111.963 111.971 mp2_gpw_main 1 4.0 0.028 0.034 110.989 110.997 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 92.274 92.602 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.751 55.078 mp2_eri_3c_integrate_gpw 272 7.0 0.152 0.167 41.143 46.161 get_2c_integrals 1 6.0 0.008 0.009 36.740 37.352 integrate_v_rspace 273 8.0 0.438 0.452 24.681 29.480 pw_transfer 6555 10.6 0.376 0.395 26.907 27.369 fft_wrap_pw1pw2 5465 11.4 0.045 0.049 25.539 26.039 grid_integrate_task_list 273 9.0 20.571 25.852 20.571 25.852 fft_wrap_pw1pw2_100 2178 12.4 1.144 1.339 23.102 23.586 compute_2c_integrals 1 7.0 0.002 0.004 19.030 19.032 rpa_ri_compute_en 1 5.0 0.048 0.068 18.599 18.807 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.620 18.748 mp2_eri_2c_integrate_gpw 1 9.0 2.363 2.410 18.617 18.744 cp_fm_cholesky_decompose 12 8.2 17.637 18.283 17.637 18.283 cholesky_decomp 1 7.0 0.000 0.000 16.560 17.203 fft3d_s 5443 13.4 16.175 16.462 16.197 16.484 ao_to_mo_and_store_B_mult_1 272 7.0 10.773 15.312 10.773 15.312 calculate_wavefunction 272 8.0 5.418 5.567 12.370 13.023 rpa_num_int 1 6.0 0.002 0.025 10.618 10.637 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.586 10.623 calc_mat_Q 8 8.0 0.000 0.000 9.464 9.565 contract_S_to_Q 8 9.0 0.000 0.000 8.889 8.987 calc_potential_gpw 544 9.5 0.005 0.006 8.241 8.717 parallel_gemm_fm 14 9.1 0.000 0.000 8.475 8.589 parallel_gemm_fm_cosma 14 10.1 8.475 8.589 8.475 8.589 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.096 8.300 potential_pw2rs 545 10.0 0.107 0.111 7.514 8.198 collocate_single_gaussian 272 10.0 0.039 0.041 7.354 7.535 create_integ_mat 1 6.0 0.021 0.036 7.514 7.523 array2fm 1 7.0 0.000 0.000 6.610 6.948 pw_scatter_s 2720 13.7 4.369 4.633 4.369 4.633 pw_gather_s 2722 13.2 3.457 3.763 3.457 3.763 array2fm_buffer_send 1 8.0 2.950 3.140 2.950 3.140 pw_poisson_solve 545 10.5 1.116 1.173 2.190 2.357 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.997496, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2807.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/02/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 22 205321. MP_Allreduce 424 9. MP_Sync 4 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.054 422.192 422.194 farming_run 1 2.0 421.264 421.268 422.138 422.141 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.237905E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 61440 MPI messages size (bytes): total size 6.073508E+09 min size 0.000000E+00 max size 642.960000E+03 average size 98.852664E+03 MPI breakdown and total messages size (bytes): size <= 128 32004 0 128 < size <= 8192 1820 14909440 8192 < size <= 32768 0 0 32768 < size <= 131072 18640 1081442304 131072 < size <= 4194304 8976 4977156096 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 1003 44. MP_Alltoall 1797 713538. MP_ISend 3686 54943. MP_IRecv 3622 54292. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 12 MP_Bcast 757 478553. MP_Allreduce 2021 21391. MP_Sync 37 MP_Alltoall 77 27167482. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.012 0.034 217.466 217.467 qs_energies 1 2.0 0.001 0.003 217.156 217.165 scf_env_do_scf 1 3.0 0.000 0.000 115.245 115.246 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.344 114.350 rebuild_ks_matrix 4 6.0 0.000 0.000 114.343 114.349 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 114.342 114.349 hfx_ks_matrix 4 8.0 0.001 0.001 113.991 113.996 integrate_four_center 4 9.0 0.153 0.472 113.991 113.995 integrate_four_center_main 4 10.0 0.103 0.555 102.316 105.892 integrate_four_center_bin 267 11.0 102.213 105.866 102.213 105.866 mp2_main 1 3.0 0.001 0.016 101.612 101.622 mp2_gpw_main 1 4.0 0.049 0.076 100.738 100.752 init_scf_loop 1 4.0 0.000 0.000 96.923 96.923 mp2_ri_gpw_compute_in 1 5.0 0.067 0.083 73.898 74.939 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.721 54.764 mp2_eri_3c_integrate_gpw 91 7.0 0.143 0.162 41.658 46.547 integrate_v_rspace 95 8.0 0.398 0.564 28.144 33.096 pw_transfer 2240 10.6 0.146 0.174 29.520 29.901 ao_to_mo_and_store_B_mult_1 91 7.0 10.390 29.084 10.390 29.084 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.527 28.900 mp2_ri_gpw_compute_en 1 5.0 0.062 0.097 26.675 28.873 grid_integrate_task_list 95 9.0 23.563 28.703 23.563 28.703 fft_wrap_pw1pw2_100 730 12.4 1.301 1.469 26.280 26.586 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.833 1.904 24.929 24.941 get_2c_integrals 1 6.0 0.003 0.021 20.081 20.110 compute_2c_integrals 1 7.0 0.003 0.004 19.058 19.061 compute_2c_integrals_loop_lm 1 8.0 0.001 0.002 18.668 18.919 mp2_eri_2c_integrate_gpw 1 9.0 1.725 1.867 18.667 18.919 fft3d_s 1823 13.4 18.490 18.779 18.504 18.793 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.321 18.321 calculate_wavefunction 91 8.0 2.005 2.035 9.612 9.860 mp2_ri_gpw_compute_en_expansio 172 7.0 0.556 0.584 8.825 9.478 potential_pw2rs 186 10.0 0.034 0.035 8.439 8.971 local_gemm 172 8.0 8.269 8.898 8.269 8.898 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.185 8.563 mp2_ri_gpw_compute_en_comm 22 7.0 0.501 0.526 7.893 8.211 calc_potential_gpw 182 9.5 0.002 0.002 7.913 8.143 collocate_single_gaussian 91 10.0 0.017 0.022 7.768 8.010 mp_sync 37 10.5 3.623 7.346 3.623 7.346 integrate_four_center_load 4 10.0 0.000 0.000 6.782 6.786 hfx_load_balance 1 11.0 0.000 0.000 6.782 6.786 mp2_ri_gpw_compute_en_ener 172 7.0 6.332 6.424 6.332 6.424 mp_sendrecv_dm3 2068 8.0 5.918 6.218 5.918 6.218 pw_gather_s 912 13.2 4.473 5.012 4.473 5.012 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=100.735233, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1497.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 452.411392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.066 0.106 55.094 55.102 qs_mol_dyn_low 1 2.0 0.003 0.003 54.625 54.641 qs_forces 11 3.9 0.003 0.012 54.565 54.567 qs_energies 11 4.9 0.002 0.003 53.045 53.076 scf_env_do_scf 11 5.9 0.000 0.001 46.846 46.847 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 44.794 44.794 qs_scf_new_mos 108 7.5 0.000 0.001 34.867 35.188 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.867 35.188 dbcsr_multiply_generic 2286 12.5 0.094 0.098 34.349 34.839 ot_scf_mini 108 9.5 0.002 0.002 33.143 33.365 multiply_cannon 2286 13.5 0.189 0.196 26.557 28.129 multiply_cannon_loop 2286 14.5 1.830 1.946 25.904 27.509 velocity_verlet 10 3.0 0.001 0.002 27.211 27.211 ot_mini 108 10.5 0.001 0.001 20.157 20.405 qs_ot_get_derivative 108 11.5 0.001 0.001 17.047 17.290 mp_waitall_1 245248 16.5 8.893 15.377 8.893 15.377 multiply_cannon_metrocomm3 54864 15.5 0.075 0.082 6.346 13.471 multiply_cannon_multrec 54864 15.5 3.620 5.822 7.538 10.922 qs_ot_get_p 119 10.4 0.001 0.001 8.256 8.523 rebuild_ks_matrix 119 8.3 0.000 0.000 7.806 7.954 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.012 7.805 7.954 mp_sum_l 7287 12.8 5.914 7.672 5.914 7.672 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.869 6.998 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.995 6.459 multiply_cannon_sync_h2d 54864 15.5 5.111 6.411 5.111 6.411 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.481 5.602 dbcsr_mm_accdrv_process 76910 16.1 1.814 2.899 3.832 5.520 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.224 5.264 init_scf_run 11 5.9 0.000 0.001 4.921 4.921 scf_env_initial_rho_setup 11 6.9 0.001 0.001 4.921 4.921 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.439 4.439 sum_up_and_integrate 119 10.3 0.012 0.014 4.419 4.428 integrate_v_rspace 119 11.3 0.002 0.003 4.407 4.417 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.251 4.251 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.088 4.249 calculate_rho_elec 119 8.7 0.011 0.016 4.087 4.249 cp_fm_redistribute_end 50 14.0 2.166 4.223 2.172 4.225 cp_fm_diag_elpa_base 50 14.0 2.047 4.116 2.051 4.127 multiply_cannon_metrocomm1 54864 15.5 0.059 0.065 1.921 3.326 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.937 3.142 apply_single 119 13.6 0.000 0.000 2.936 3.142 calculate_dm_sparse 119 9.5 0.000 0.001 2.940 3.114 acc_transpose_blocks 54864 15.5 0.232 0.266 2.223 2.797 ot_diis_step 108 11.5 0.006 0.006 2.785 2.786 calculate_first_density_matrix 1 7.0 0.000 0.000 2.663 2.668 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.538 2.582 jit_kernel_multiply 13 15.8 1.954 2.539 1.954 2.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.280 2.282 mp_sum_d 4135 12.0 1.507 2.212 1.507 2.212 wfi_extrapolate 11 7.9 0.001 0.001 2.188 2.188 density_rs2pw 119 9.7 0.004 0.004 2.074 2.187 grid_integrate_task_list 119 12.3 2.017 2.120 2.017 2.120 init_scf_loop 11 6.9 0.000 0.000 2.037 2.037 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.954 2.022 potential_pw2rs 119 12.3 0.004 0.004 1.795 1.808 pw_transfer 1439 11.6 0.052 0.056 1.618 1.691 make_m2s 4572 13.5 0.053 0.055 1.585 1.632 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.542 1.619 mp_alltoall_d11v 2130 13.8 1.322 1.562 1.322 1.562 make_images 4572 14.5 0.133 0.140 1.503 1.549 acc_transpose_blocks_sync 164592 16.5 1.196 1.471 1.196 1.471 grid_collocate_task_list 119 9.7 1.349 1.435 1.349 1.435 transfer_rs2pw 487 10.6 0.005 0.006 1.326 1.433 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.384 1.422 mp_waitany 12084 13.8 1.246 1.418 1.246 1.418 fft3d_ps 1201 14.6 0.372 0.476 1.256 1.328 transfer_pw2rs 487 13.2 0.006 0.006 1.310 1.318 fft_wrap_pw1pw2_140 487 13.2 0.142 0.155 1.203 1.280 dbcsr_dot_sd 1205 11.9 0.049 0.059 0.834 1.242 parallel_gemm_fm 81 9.0 0.000 0.000 1.166 1.170 parallel_gemm_fm_cosma 81 10.0 1.166 1.170 1.166 1.170 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=55.102000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.090909, yerr=1.311110 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 487.739392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2194560 MPI messages size (bytes): total size 310.646604E+09 min size 0.000000E+00 max size 1.145520E+06 average size 141.553031E+03 MPI breakdown and total messages size (bytes): size <= 128 724648 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 281952 4619501568 32768 < size <= 131072 494448 39143342080 131072 < size <= 4194304 440000 264807943488 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3672 62658. MP_Allreduce 10306 303. MP_Sync 54 MP_Alltoall 2060 197481. 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 0.056 0.095 39.237 39.241 qs_mol_dyn_low 1 2.0 0.004 0.016 38.746 38.755 qs_forces 11 3.9 0.003 0.004 38.656 38.663 qs_energies 11 4.9 0.004 0.014 36.981 36.995 scf_env_do_scf 11 5.9 0.001 0.004 31.586 31.586 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 29.167 29.168 dbcsr_multiply_generic 2286 12.5 0.101 0.108 22.014 22.353 qs_scf_new_mos 108 7.5 0.001 0.001 20.559 20.784 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.559 20.784 ot_scf_mini 108 9.5 0.003 0.004 19.633 19.790 multiply_cannon 2286 13.5 0.210 0.218 16.895 18.385 velocity_verlet 10 3.0 0.002 0.008 18.289 18.292 multiply_cannon_loop 2286 14.5 1.192 1.250 15.718 17.189 ot_mini 108 10.5 0.001 0.002 12.129 12.356 mp_waitall_1 200699 16.5 5.565 10.663 5.565 10.663 qs_ot_get_derivative 108 11.5 0.001 0.002 9.759 9.918 multiply_cannon_metrocomm3 27432 15.5 0.072 0.075 4.039 9.405 multiply_cannon_multrec 27432 15.5 1.821 4.201 6.140 8.937 rebuild_ks_matrix 119 8.3 0.000 0.000 6.772 6.915 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.017 6.772 6.915 dbcsr_mm_accdrv_process 47894 16.0 3.542 5.605 4.237 6.352 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.964 6.096 qs_ot_get_p 119 10.4 0.001 0.001 4.598 4.811 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.714 4.558 mp_sum_l 7287 12.8 2.059 4.126 2.059 4.126 init_scf_run 11 5.9 0.000 0.001 4.117 4.118 scf_env_initial_rho_setup 11 6.9 0.001 0.002 4.117 4.117 apply_preconditioner_dbcsr 119 12.6 0.000 0.001 3.011 4.042 apply_single 119 13.6 0.000 0.000 3.011 4.041 sum_up_and_integrate 119 10.3 0.025 0.028 3.700 3.709 integrate_v_rspace 119 11.3 0.002 0.003 3.675 3.684 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.597 3.628 calculate_rho_elec 119 8.7 0.021 0.024 3.596 3.628 make_m2s 4572 13.5 0.052 0.054 2.634 2.988 qs_ot_p2m_diag 50 11.0 0.009 0.013 2.959 2.978 make_images 4572 14.5 0.205 0.243 2.547 2.904 calculate_first_density_matrix 1 7.0 0.001 0.003 2.609 2.611 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.562 2.562 init_scf_loop 11 6.9 0.002 0.011 2.399 2.400 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.312 2.396 ot_diis_step 108 11.5 0.011 0.014 2.322 2.323 multiply_cannon_sync_h2d 27432 15.5 1.702 2.299 1.702 2.299 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.249 2.250 cp_fm_redistribute_end 50 14.0 1.140 2.226 1.143 2.228 cp_fm_diag_elpa_base 50 14.0 1.053 2.132 1.081 2.176 calculate_dm_sparse 119 9.5 0.000 0.001 2.064 2.144 density_rs2pw 119 9.7 0.004 0.004 1.897 1.981 grid_integrate_task_list 119 12.3 1.842 1.957 1.842 1.957 acc_transpose_blocks 27432 15.5 0.113 0.119 1.612 1.944 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.940 1.942 jit_kernel_multiply 10 16.3 0.637 1.938 0.637 1.938 pw_transfer 1439 11.6 0.065 0.069 1.812 1.844 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.732 1.773 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.721 1.754 make_images_data 4572 15.5 0.048 0.054 1.260 1.609 prepare_preconditioner 11 7.9 0.000 0.000 1.501 1.528 make_preconditioner 11 8.9 0.000 0.001 1.501 1.528 hybrid_alltoall_any 4725 16.4 0.054 0.116 1.114 1.499 make_full_inverse_cholesky 11 9.9 0.000 0.001 1.403 1.459 wfi_extrapolate 11 7.9 0.001 0.001 1.449 1.449 potential_pw2rs 119 12.3 0.006 0.006 1.431 1.439 fft_wrap_pw1pw2_140 487 13.2 0.160 0.168 1.370 1.404 mp_alltoall_d11v 2130 13.8 1.195 1.392 1.195 1.392 fft3d_ps 1201 14.6 0.514 0.571 1.356 1.386 grid_collocate_task_list 119 9.7 1.285 1.358 1.285 1.358 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.247 1.255 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.195 1.242 transfer_rs2pw 487 10.6 0.005 0.006 1.120 1.225 mp_sum_d 4135 12.0 0.585 1.088 0.585 1.088 mp_allgather_i34 2286 14.5 0.596 1.027 0.596 1.027 qs_energies_init_hamiltonians 11 5.9 0.002 0.010 0.993 1.000 acc_transpose_blocks_sync 82296 16.5 0.814 0.941 0.814 0.941 transfer_pw2rs 487 13.2 0.005 0.005 0.903 0.911 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.871 0.884 acc_transpose_blocks_kernels 27432 16.5 0.190 0.280 0.659 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.241000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=464.909091, yerr=1.083307 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.589312E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 950976 MPI messages size (bytes): total size 203.844256E+09 min size 0.000000E+00 max size 1.638400E+06 average size 214.352688E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 253512 2076770304 8192 < size <= 32768 179424 2939682816 32768 < size <= 131072 181440 14863564800 131072 < size <= 4194304 330176 183964913216 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63490. MP_Allreduce 10155 305. MP_Sync 54 MP_Alltoall 1821 1607811. MP_SendRecv 11067 57667. MP_ISendRecv 11067 57667. MP_Wait 21987 MP_ISend 9880 92618. MP_IRecv 9880 92618. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.040 31.953 31.955 qs_mol_dyn_low 1 2.0 0.003 0.004 31.642 31.662 qs_forces 11 3.9 0.003 0.003 31.578 31.580 qs_energies 11 4.9 0.002 0.003 30.028 30.030 scf_env_do_scf 11 5.9 0.000 0.001 25.242 25.242 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 22.852 22.852 dbcsr_multiply_generic 2286 12.5 0.096 0.098 16.890 16.986 qs_scf_new_mos 108 7.5 0.001 0.001 15.290 15.308 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.289 15.308 velocity_verlet 10 3.0 0.002 0.002 14.969 14.971 ot_scf_mini 108 9.5 0.002 0.002 14.534 14.554 multiply_cannon 2286 13.5 0.200 0.204 13.718 14.464 multiply_cannon_loop 2286 14.5 0.860 0.886 12.970 13.710 ot_mini 108 10.5 0.001 0.001 9.016 9.041 qs_ot_get_derivative 108 11.5 0.001 0.001 7.542 7.562 multiply_cannon_multrec 18288 15.5 1.874 2.900 7.006 7.398 dbcsr_mm_accdrv_process 38222 16.0 4.925 5.867 5.038 5.933 rebuild_ks_matrix 119 8.3 0.000 0.000 5.914 5.929 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.914 5.929 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.207 5.222 mp_waitall_1 158411 16.6 2.612 3.712 2.612 3.712 init_scf_run 11 5.9 0.000 0.001 3.572 3.572 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.572 3.572 sum_up_and_integrate 119 10.3 0.030 0.031 3.513 3.516 integrate_v_rspace 119 11.3 0.003 0.003 3.483 3.489 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.863 3.427 qs_ot_get_p 119 10.4 0.001 0.001 3.294 3.316 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.198 3.204 calculate_rho_elec 119 8.7 0.031 0.031 3.197 3.203 multiply_cannon_metrocomm3 18288 15.5 0.048 0.050 1.467 2.452 init_scf_loop 11 6.9 0.000 0.000 2.375 2.376 calculate_first_density_matrix 1 7.0 0.000 0.000 2.373 2.374 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.964 2.340 apply_single 119 13.6 0.000 0.000 1.964 2.340 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.207 2.215 grid_integrate_task_list 119 12.3 1.809 1.934 1.809 1.934 calculate_dm_sparse 119 9.5 0.000 0.000 1.905 1.914 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.910 1.910 make_m2s 4572 13.5 0.044 0.045 1.779 1.906 density_rs2pw 119 9.7 0.004 0.004 1.769 1.877 make_images 4572 14.5 0.192 0.204 1.695 1.822 pw_transfer 1439 11.6 0.065 0.069 1.794 1.803 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.704 1.716 acc_transpose_blocks 18288 15.5 0.080 0.081 1.643 1.715 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.701 1.712 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.708 1.710 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.670 1.680 cp_fm_diag_elpa_base 50 14.0 1.647 1.661 1.668 1.678 prepare_preconditioner 11 7.9 0.000 0.000 1.620 1.621 make_preconditioner 11 8.9 0.000 0.000 1.620 1.621 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.477 1.561 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.538 1.543 ot_diis_step 108 11.5 0.011 0.011 1.461 1.461 mp_sum_l 7287 12.8 1.011 1.433 1.011 1.433 fft_wrap_pw1pw2_140 487 13.2 0.213 0.218 1.382 1.394 grid_collocate_task_list 119 9.7 1.234 1.304 1.234 1.304 potential_pw2rs 119 12.3 0.007 0.008 1.290 1.294 fft3d_ps 1201 14.6 0.532 0.550 1.268 1.279 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.247 multiply_cannon_sync_h2d 18288 15.5 1.028 1.189 1.028 1.189 wfi_extrapolate 11 7.9 0.001 0.001 1.146 1.146 transfer_rs2pw 487 10.6 0.005 0.005 0.949 1.051 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.943 0.961 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.957 0.959 make_images_data 4572 15.5 0.047 0.051 0.759 0.908 hybrid_alltoall_any 4725 16.4 0.058 0.116 0.657 0.842 jit_kernel_multiply 4 16.1 0.054 0.814 0.054 0.814 acc_transpose_blocks_kernels 18288 16.5 0.220 0.227 0.802 0.814 acc_transpose_blocks_sync 54864 16.5 0.742 0.810 0.742 0.810 transfer_pw2rs 487 13.2 0.004 0.004 0.780 0.782 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.776 0.778 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.674 0.757 mp_alltoall_d11v 2130 13.8 0.649 0.735 0.649 0.735 cp_fm_cholesky_invert 11 10.9 0.678 0.682 0.678 0.682 mp_alltoall_z22v 1201 16.6 0.590 0.655 0.590 0.655 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=31.955000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.818182, yerr=1.402477 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 558.333952E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1042416 MPI messages size (bytes): total size 150.443262E+09 min size 0.000000E+00 max size 1.188816E+06 average size 144.321719E+03 MPI breakdown and total messages size (bytes): size <= 128 228256 0 128 < size <= 8192 126888 1039466496 8192 < size <= 32768 191472 3137077248 32768 < size <= 131072 295800 25899827200 131072 < size <= 4194304 200000 120367247040 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63489. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 2412273. MP_SendRecv 8211 74133. MP_ISendRecv 8211 74133. MP_Wait 16271 MP_ISend 7280 135929. MP_IRecv 7280 135929. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.146 37.218 37.219 qs_mol_dyn_low 1 2.0 0.003 0.014 36.722 36.734 qs_forces 11 3.9 0.017 0.072 36.663 36.664 qs_energies 11 4.9 0.007 0.029 34.875 34.883 scf_env_do_scf 11 5.9 0.001 0.002 29.515 29.516 scf_env_do_scf_inner_loop 108 6.5 0.003 0.009 26.142 26.142 dbcsr_multiply_generic 2286 12.5 0.100 0.107 19.914 20.057 velocity_verlet 10 3.0 0.002 0.003 18.415 18.417 qs_scf_new_mos 108 7.5 0.001 0.001 17.942 17.994 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.941 17.993 ot_scf_mini 108 9.5 0.002 0.003 16.909 16.960 multiply_cannon 2286 13.5 0.221 0.244 16.108 16.501 multiply_cannon_loop 2286 14.5 1.549 1.649 15.130 15.493 ot_mini 108 10.5 0.001 0.001 10.372 10.430 multiply_cannon_multrec 27432 15.5 2.456 3.125 9.033 9.321 qs_ot_get_derivative 108 11.5 0.001 0.001 8.497 8.548 dbcsr_mm_accdrv_process 47916 15.9 5.960 7.633 6.474 7.808 rebuild_ks_matrix 119 8.3 0.000 0.000 6.312 6.367 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.312 6.366 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.598 5.646 init_scf_run 11 5.9 0.001 0.004 3.909 3.910 scf_env_initial_rho_setup 11 6.9 0.004 0.005 3.908 3.909 qs_ot_get_p 119 10.4 0.001 0.001 3.667 3.739 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.061 3.548 sum_up_and_integrate 119 10.3 0.035 0.037 3.494 3.503 integrate_v_rspace 119 11.3 0.003 0.005 3.459 3.469 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.331 3.374 calculate_rho_elec 119 8.7 0.040 0.046 3.331 3.373 init_scf_loop 11 6.9 0.001 0.003 3.353 3.353 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.201 2.650 apply_single 119 13.6 0.000 0.000 2.201 2.650 prepare_preconditioner 11 7.9 0.000 0.001 2.523 2.531 make_preconditioner 11 8.9 0.000 0.002 2.523 2.531 calculate_first_density_matrix 1 7.0 0.000 0.002 2.515 2.519 acc_transpose_blocks 27432 15.5 0.120 0.127 2.440 2.511 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.134 2.457 make_m2s 4572 13.5 0.054 0.056 2.292 2.404 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.373 2.400 make_images 4572 14.5 0.274 0.333 2.185 2.296 qs_ot_p2m_diag 50 11.0 0.016 0.024 2.237 2.249 mp_waitall_1 137007 16.6 1.666 2.228 1.666 2.228 calculate_dm_sparse 119 9.5 0.000 0.000 2.152 2.204 pw_transfer 1439 11.6 0.065 0.069 1.919 1.956 grid_integrate_task_list 119 12.3 1.830 1.908 1.830 1.908 density_rs2pw 119 9.7 0.004 0.004 1.793 1.899 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.827 1.867 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.866 1.867 ot_diis_step 108 11.5 0.012 0.012 1.833 1.833 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.828 1.829 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.654 1.666 mp_sum_l 7287 12.8 1.113 1.656 1.113 1.656 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.571 1.580 cp_fm_diag_elpa_base 50 14.0 1.536 1.553 1.568 1.577 fft_wrap_pw1pw2_140 487 13.2 0.246 0.259 1.508 1.552 acc_transpose_blocks_sync 82296 16.5 1.459 1.524 1.459 1.524 fft3d_ps 1201 14.6 0.560 0.613 1.345 1.369 grid_collocate_task_list 119 9.7 1.251 1.339 1.251 1.339 wfi_extrapolate 11 7.9 0.001 0.001 1.336 1.336 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.299 1.314 potential_pw2rs 119 12.3 0.009 0.010 1.278 1.282 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.250 1.258 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.693 1.167 qs_energies_init_hamiltonians 11 5.9 0.001 0.005 1.149 1.156 cp_fm_upper_to_full 72 14.2 0.806 1.147 0.806 1.147 transfer_rs2pw 487 10.6 0.005 0.005 0.959 1.078 jit_kernel_multiply 6 15.9 0.448 1.074 0.448 1.074 dbcsr_complete_redistribute 329 12.2 0.128 0.191 0.797 1.068 make_images_data 4572 15.5 0.047 0.052 0.848 0.966 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.744 0.903 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.010 0.802 0.891 mp_alltoall_d11v 2130 13.8 0.744 0.873 0.744 0.873 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.599 0.862 acc_transpose_blocks_kernels 27432 16.5 0.275 0.283 0.831 0.846 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.804 0.809 cp_fm_cholesky_invert 11 10.9 0.754 0.757 0.754 0.757 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.219000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.363636, yerr=2.603240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 610.713600E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 219456 MPI messages size (bytes): total size 97.042514E+09 min size 0.000000E+00 max size 3.276800E+06 average size 442.195750E+03 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 101892 3336634368 32768 < size <= 131072 0 0 131072 < size <= 4194304 116112 93705670464 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8156 20. MP_Alltoall 8655 64935. MP_ISend 36532 168375. MP_IRecv 36532 168349. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63488. MP_Allreduce 10154 346. MP_Sync 54 MP_Alltoall 1582 3682667. MP_SendRecv 5355 94533. MP_ISendRecv 5355 94533. MP_Wait 11335 MP_ISend 5200 225425. MP_IRecv 5200 225425. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.015 0.029 28.886 28.887 qs_mol_dyn_low 1 2.0 0.003 0.006 28.606 28.617 qs_forces 11 3.9 0.002 0.002 28.498 28.504 qs_energies 11 4.9 0.001 0.002 26.804 26.811 scf_env_do_scf 11 5.9 0.000 0.001 21.639 21.639 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 19.210 19.211 velocity_verlet 10 3.0 0.007 0.008 14.559 14.564 dbcsr_multiply_generic 2286 12.5 0.093 0.098 13.168 13.245 qs_scf_new_mos 108 7.5 0.001 0.001 11.692 11.715 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.691 11.714 multiply_cannon 2286 13.5 0.222 0.228 10.630 11.126 ot_scf_mini 108 9.5 0.002 0.002 10.984 11.009 multiply_cannon_loop 2286 14.5 0.647 0.678 9.717 9.887 ot_mini 108 10.5 0.001 0.001 6.344 6.375 multiply_cannon_multrec 9144 15.5 1.755 1.991 6.148 6.365 rebuild_ks_matrix 119 8.3 0.000 0.000 5.707 5.731 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.707 5.731 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.075 5.097 qs_ot_get_derivative 108 11.5 0.001 0.001 5.065 5.090 dbcsr_mm_accdrv_process 12550 15.8 3.585 4.297 4.283 4.351 init_scf_run 11 5.9 0.000 0.001 3.704 3.704 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.704 3.704 sum_up_and_integrate 119 10.3 0.037 0.041 3.409 3.412 integrate_v_rspace 119 11.3 0.003 0.006 3.372 3.376 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.270 3.277 calculate_rho_elec 119 8.7 0.060 0.061 3.270 3.277 qs_ot_get_p 119 10.4 0.001 0.001 2.777 2.807 calculate_first_density_matrix 1 7.0 0.000 0.000 2.562 2.563 init_scf_loop 11 6.9 0.000 0.000 2.409 2.410 pw_transfer 1439 11.6 0.065 0.068 1.935 1.945 grid_integrate_task_list 119 12.3 1.866 1.926 1.866 1.926 make_m2s 4572 13.5 0.033 0.035 1.768 1.918 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.842 1.855 calculate_dm_sparse 119 9.5 0.000 0.000 1.819 1.836 make_images 4572 14.5 0.268 0.300 1.680 1.829 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.814 1.817 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.727 1.746 density_rs2pw 119 9.7 0.004 0.004 1.661 1.732 mp_waitall_1 115863 16.7 1.243 1.701 1.243 1.701 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.698 1.699 prepare_preconditioner 11 7.9 0.000 0.000 1.682 1.687 make_preconditioner 11 8.9 0.000 0.000 1.682 1.687 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.576 1.603 acc_transpose_blocks 9144 15.5 0.041 0.043 1.526 1.580 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.572 1.572 fft_wrap_pw1pw2_140 487 13.2 0.323 0.331 1.523 1.537 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.483 1.490 jit_kernel_multiply 8 15.5 0.658 1.433 0.658 1.433 grid_collocate_task_list 119 9.7 1.297 1.354 1.297 1.354 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.341 1.353 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.309 1.317 cp_fm_diag_elpa_base 50 14.0 1.282 1.300 1.307 1.316 ot_diis_step 108 11.5 0.013 0.013 1.267 1.267 fft3d_ps 1201 14.6 0.558 0.570 1.256 1.266 qs_energies_init_hamiltonians 11 5.9 0.004 0.007 1.242 1.248 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.207 1.243 apply_single 119 13.6 0.000 0.000 1.207 1.243 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.208 1.213 potential_pw2rs 119 12.3 0.010 0.011 1.194 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.091 1.091 hybrid_alltoall_any 4725 16.4 0.065 0.175 0.738 0.943 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.873 0.921 make_images_data 4572 15.5 0.041 0.046 0.751 0.906 cp_fm_cholesky_invert 11 10.9 0.844 0.847 0.844 0.847 transfer_rs2pw 487 10.6 0.004 0.004 0.754 0.827 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.812 0.818 mp_alltoall_d11v 2130 13.8 0.750 0.803 0.750 0.803 acc_transpose_blocks_sync 27432 16.5 0.725 0.775 0.725 0.775 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.753 0.757 acc_transpose_blocks_kernels 9144 16.5 0.119 0.121 0.744 0.747 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.686 0.726 multiply_cannon_metrocomm3 9144 15.5 0.020 0.021 0.335 0.691 mp_allgather_i34 2286 14.5 0.225 0.639 0.225 0.639 transfer_pw2rs 487 13.2 0.003 0.004 0.637 0.638 jit_kernel_transpose 5 15.6 0.626 0.629 0.626 0.629 mp_alltoall_z22v 1201 16.6 0.571 0.599 0.571 0.599 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=28.887000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=580.181818, yerr=3.379569 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 780.189696E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 91440 MPI messages size (bytes): total size 85.748679E+09 min size 0.000000E+00 max size 6.553600E+06 average size 937.758938E+03 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 21148 692256768 32768 < size <= 131072 19224 1259864064 131072 < size <= 4194304 41040 21941452800 4194304 < size <= 16777216 9456 61855174464 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3622 63723. MP_Allreduce 10154 429. MP_Sync 54 MP_Alltoall 1582 7383731. MP_SendRecv 2499 189067. MP_ISendRecv 2499 189067. MP_Wait 6399 MP_ISend 3120 546875. MP_IRecv 3120 546875. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.043 42.616 42.618 qs_mol_dyn_low 1 2.0 0.003 0.004 42.239 42.247 qs_forces 11 3.9 0.004 0.014 42.157 42.159 qs_energies 11 4.9 0.002 0.005 40.156 40.162 scf_env_do_scf 11 5.9 0.001 0.001 34.200 34.201 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.479 26.481 velocity_verlet 10 3.0 0.002 0.002 23.779 23.785 dbcsr_multiply_generic 2286 12.5 0.101 0.105 18.985 19.242 qs_scf_new_mos 108 7.5 0.001 0.001 17.160 17.262 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.159 17.261 multiply_cannon 2286 13.5 0.297 0.301 15.149 16.103 ot_scf_mini 108 9.5 0.002 0.002 16.001 16.101 multiply_cannon_loop 2286 14.5 0.871 0.882 13.904 14.883 ot_mini 108 10.5 0.001 0.001 9.786 9.903 multiply_cannon_multrec 9144 15.5 3.396 4.743 8.906 9.038 qs_ot_get_derivative 108 11.5 0.001 0.001 7.755 7.856 init_scf_loop 11 6.9 0.000 0.000 7.695 7.698 rebuild_ks_matrix 119 8.3 0.000 0.001 7.058 7.201 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.057 7.201 prepare_preconditioner 11 7.9 0.000 0.000 6.735 6.749 make_preconditioner 11 8.9 0.000 0.000 6.735 6.749 dbcsr_mm_accdrv_process 12550 15.8 4.513 5.913 5.371 6.696 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.284 6.631 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.358 6.488 cp_fm_upper_to_full 72 14.2 3.159 4.558 3.159 4.558 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.975 3.981 calculate_rho_elec 119 8.7 0.118 0.121 3.975 3.981 init_scf_run 11 5.9 0.000 0.001 3.874 3.874 scf_env_initial_rho_setup 11 6.9 0.001 0.002 3.874 3.874 sum_up_and_integrate 119 10.3 0.065 0.066 3.780 3.786 integrate_v_rspace 119 11.3 0.004 0.004 3.715 3.722 qs_ot_get_p 119 10.4 0.001 0.001 3.313 3.450 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.803 3.199 mp_waitall_1 94719 16.7 2.158 3.102 2.158 3.102 dbcsr_complete_redistribute 329 12.2 0.287 0.293 1.978 2.813 pw_transfer 1439 11.6 0.069 0.070 2.544 2.548 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.678 2.499 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.177 2.484 apply_single 119 13.6 0.000 0.000 2.177 2.484 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.445 2.449 calculate_first_density_matrix 1 7.0 0.000 0.000 2.427 2.428 make_m2s 4572 13.5 0.038 0.038 2.270 2.413 calculate_dm_sparse 119 9.5 0.000 0.000 2.309 2.328 make_images 4572 14.5 0.356 0.390 2.149 2.292 mp_alltoall_i22 627 13.8 1.422 2.265 1.422 2.265 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.446 2.261 multiply_cannon_metrocomm3 9144 15.5 0.021 0.022 1.305 2.186 grid_integrate_task_list 119 12.3 2.081 2.123 2.081 2.123 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.050 2.100 fft_wrap_pw1pw2_140 487 13.2 0.576 0.582 2.076 2.080 density_rs2pw 119 9.7 0.004 0.004 2.003 2.032 ot_diis_step 108 11.5 0.014 0.014 2.005 2.005 qs_ot_p2m_diag 50 11.0 0.043 0.044 1.945 1.947 acc_transpose_blocks 9144 15.5 0.044 0.045 1.849 1.864 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.850 1.850 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 1.790 1.793 mp_sum_l 7287 12.8 1.021 1.716 1.021 1.716 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.592 1.636 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.588 1.588 grid_collocate_task_list 119 9.7 1.522 1.559 1.522 1.559 fft3d_ps 1201 14.6 0.596 0.607 1.541 1.544 wfi_extrapolate 11 7.9 0.001 0.001 1.383 1.383 cp_fm_cholesky_invert 11 10.9 1.375 1.378 1.375 1.378 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.307 1.329 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.317 1.326 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.313 1.313 cp_fm_diag_elpa_base 50 14.0 1.168 1.222 1.312 1.312 potential_pw2rs 119 12.3 0.014 0.014 1.233 1.235 hybrid_alltoall_any 4725 16.4 0.090 0.151 1.034 1.228 mp_alltoall_d11v 2130 13.8 1.160 1.190 1.160 1.190 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.116 1.133 acc_transpose_blocks_sync 27432 16.5 1.109 1.125 1.109 1.125 make_images_data 4572 15.5 0.046 0.049 0.934 1.104 jit_kernel_multiply 6 15.5 0.829 1.071 0.829 1.071 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 0.995 1.060 qs_create_task_list 11 7.9 0.001 0.003 0.935 0.947 generate_qs_task_list 11 8.9 0.368 0.387 0.933 0.945 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.909 0.923 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=42.618000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=730.181818, yerr=13.272105 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.280768E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 8483040 MPI messages size (bytes): total size 1.160510E+12 min size 0.000000E+00 max size 1.161504E+06 average size 136.803609E+03 MPI breakdown and total messages size (bytes): size <= 128 1836752 0 128 < size <= 8192 1040592 8524529664 8192 < size <= 32768 1486976 24362614784 32768 < size <= 131072 2491776 216971345920 131072 < size <= 4194304 1626944 910632720448 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66212. MP_Allreduce 9776 488. MP_Sync 52 MP_Alltoall 1938 1383689. MP_SendRecv 20900 9096. MP_ISendRecv 20900 9096. MP_Wait 37268 MP_ISend 14300 82312. MP_IRecv 14300 82312. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.017 0.034 81.781 81.782 qs_mol_dyn_low 1 2.0 0.003 0.003 81.510 81.519 qs_forces 11 3.9 0.003 0.003 81.444 81.445 qs_energies 11 4.9 0.001 0.001 78.584 78.613 scf_env_do_scf 11 5.9 0.001 0.001 69.572 69.575 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 64.074 64.074 dbcsr_multiply_generic 2055 12.4 0.106 0.109 51.048 51.347 qs_scf_new_mos 99 7.5 0.001 0.001 47.185 47.320 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.185 47.319 ot_scf_mini 99 9.5 0.002 0.002 44.836 44.936 multiply_cannon 2055 13.4 0.184 0.190 42.666 43.607 velocity_verlet 10 3.0 0.002 0.007 43.201 43.202 multiply_cannon_loop 2055 14.4 1.807 1.853 41.676 42.707 ot_mini 99 10.5 0.001 0.001 26.613 26.712 qs_ot_get_derivative 99 11.5 0.001 0.001 19.830 19.938 multiply_cannon_multrec 49320 15.4 11.411 12.115 17.402 18.179 rebuild_ks_matrix 110 8.3 0.000 0.001 14.417 14.543 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.013 14.417 14.543 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.641 12.754 mp_waitall_1 220248 16.4 10.785 11.905 10.785 11.905 multiply_cannon_sync_h2d 49320 15.4 9.617 10.231 9.617 10.231 qs_ot_get_p 110 10.4 0.001 0.001 9.788 9.886 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.779 8.327 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 6.577 8.122 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.208 7.714 apply_single 110 13.6 0.000 0.000 7.207 7.714 sum_up_and_integrate 110 10.3 0.036 0.043 6.945 6.956 init_scf_run 11 5.9 0.000 0.001 6.955 6.956 scf_env_initial_rho_setup 11 6.9 0.006 0.014 6.955 6.956 integrate_v_rspace 110 11.3 0.003 0.004 6.909 6.928 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.570 6.592 ot_diis_step 99 11.5 0.006 0.006 6.584 6.584 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.229 6.355 calculate_rho_elec 110 8.6 0.021 0.024 6.229 6.354 dbcsr_mm_accdrv_process 87628 16.1 3.033 3.137 5.860 6.169 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.713 5.713 init_scf_loop 11 6.9 0.000 0.000 5.472 5.472 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.317 5.384 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.152 5.181 cp_fm_diag_elpa_base 48 14.0 5.139 5.168 5.151 5.180 mp_sum_l 6594 12.7 4.003 4.754 4.003 4.754 make_m2s 4110 13.4 0.060 0.066 3.922 4.056 wfi_extrapolate 11 7.9 0.001 0.001 4.017 4.017 make_images 4110 14.4 0.177 0.190 3.827 3.963 calculate_dm_sparse 110 9.5 0.001 0.001 3.716 3.794 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.753 3.757 multiply_cannon_metrocomm1 49320 15.4 0.067 0.070 2.357 3.576 density_rs2pw 110 9.6 0.004 0.005 3.286 3.442 grid_integrate_task_list 110 12.3 3.243 3.424 3.243 3.424 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.291 3.345 prepare_preconditioner 11 7.9 0.000 0.000 3.301 3.316 make_preconditioner 11 8.9 0.000 0.000 3.301 3.316 qs_ot_get_orbitals 99 10.5 0.000 0.001 3.153 3.194 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.088 3.134 pw_transfer 1331 11.6 0.055 0.064 3.003 3.080 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 2.914 2.989 calculate_first_density_matrix 1 7.0 0.000 0.001 2.837 2.845 acc_transpose_blocks 49320 15.4 0.218 0.225 2.591 2.649 fft_wrap_pw1pw2_140 451 13.1 0.389 0.431 2.497 2.586 potential_pw2rs 110 12.3 0.006 0.006 2.551 2.572 jit_kernel_multiply 13 15.9 2.548 2.564 2.548 2.564 mp_alltoall_d11v 2046 13.8 2.020 2.550 2.020 2.550 grid_collocate_task_list 110 9.6 2.159 2.291 2.159 2.291 fft3d_ps 1111 14.6 0.796 0.888 2.210 2.276 mp_waitany 14300 13.8 1.803 2.063 1.803 2.063 transfer_rs2pw 451 10.6 0.005 0.006 1.870 2.045 make_images_data 4110 15.4 0.044 0.047 1.785 1.926 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.907 1.922 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.552 1.862 mp_sum_d 3889 11.9 1.355 1.794 1.355 1.794 cp_fm_cholesky_invert 11 10.9 1.787 1.790 1.787 1.790 transfer_pw2rs 451 13.1 0.006 0.007 1.736 1.744 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.655 1.680 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=81.782000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=476.272727, yerr=2.699862 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 594.038784E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1972800 MPI messages size (bytes): total size 1.077520E+12 min size 0.000000E+00 max size 4.537280E+06 average size 546.188250E+03 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 222984 1826684928 8192 < size <= 32768 520356 13399818240 32768 < size <= 131072 372336 35386294272 131072 < size <= 4194304 787758 788321309808 4194304 < size <= 16777216 54450 238588003280 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66430. MP_Allreduce 9775 566. MP_Sync 52 MP_Alltoall 1717 1347556. MP_SendRecv 10340 26400. MP_ISendRecv 10340 26400. MP_Wait 22352 MP_ISend 10164 155761. MP_IRecv 10164 155761. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.073 0.090 68.691 68.704 qs_mol_dyn_low 1 2.0 0.003 0.004 68.134 68.160 qs_forces 11 3.9 0.022 0.059 68.057 68.058 qs_energies 11 4.9 0.019 0.083 64.701 64.722 scf_env_do_scf 11 5.9 0.001 0.004 55.984 55.988 scf_env_do_scf_inner_loop 99 6.5 0.003 0.008 48.488 48.492 dbcsr_multiply_generic 2055 12.4 0.115 0.118 37.633 37.802 velocity_verlet 10 3.0 0.002 0.005 35.648 35.650 qs_scf_new_mos 99 7.5 0.001 0.001 32.817 32.963 qs_scf_loop_do_ot 99 8.5 0.001 0.001 32.817 32.962 multiply_cannon 2055 13.4 0.223 0.246 31.138 32.119 ot_scf_mini 99 9.5 0.003 0.004 31.157 31.287 multiply_cannon_loop 2055 14.4 1.166 1.198 30.036 30.885 ot_mini 99 10.5 0.001 0.001 18.236 18.385 multiply_cannon_multrec 24660 15.4 6.991 8.263 14.038 15.405 rebuild_ks_matrix 110 8.3 0.000 0.001 13.453 13.594 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 13.453 13.594 qs_ot_get_derivative 99 11.5 0.001 0.001 12.479 12.612 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.814 11.941 mp_waitall_1 176588 16.5 7.459 9.840 7.459 9.840 multiply_cannon_metrocomm3 24660 15.4 0.072 0.076 5.120 7.650 multiply_cannon_sync_h2d 24660 15.4 6.382 7.594 6.382 7.594 init_scf_loop 11 6.9 0.002 0.011 7.461 7.466 dbcsr_mm_accdrv_process 52282 16.1 5.594 6.305 6.876 7.182 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.447 7.051 apply_single 110 13.6 0.000 0.001 6.447 7.051 qs_ot_get_p 110 10.4 0.001 0.001 6.350 6.494 sum_up_and_integrate 110 10.3 0.053 0.060 6.341 6.356 integrate_v_rspace 110 11.3 0.002 0.003 6.288 6.304 init_scf_run 11 5.9 0.000 0.001 6.130 6.130 scf_env_initial_rho_setup 11 6.9 0.001 0.002 6.129 6.130 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.798 5.811 calculate_rho_elec 110 8.6 0.039 0.047 5.798 5.811 ot_diis_step 99 11.5 0.010 0.010 5.712 5.712 prepare_preconditioner 11 7.9 0.000 0.001 5.424 5.455 make_preconditioner 11 8.9 0.001 0.002 5.424 5.455 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.710 5.417 make_full_inverse_cholesky 11 9.9 0.000 0.002 5.006 5.159 qs_ot_p2m_diag 48 11.0 0.029 0.045 4.466 4.487 make_m2s 4110 13.4 0.056 0.061 4.158 4.477 make_images 4110 14.4 0.406 0.463 4.051 4.368 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.006 4.007 pw_transfer 1331 11.6 0.066 0.075 3.494 3.628 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.387 3.525 wfi_extrapolate 11 7.9 0.001 0.001 3.474 3.474 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.454 3.465 cp_fm_diag_elpa_base 48 14.0 3.409 3.429 3.451 3.462 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.250 3.334 grid_integrate_task_list 110 12.3 3.180 3.325 3.180 3.325 density_rs2pw 110 9.6 0.004 0.005 3.056 3.261 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.231 3.232 fft_wrap_pw1pw2_140 451 13.1 0.459 0.473 2.927 3.070 calculate_dm_sparse 110 9.5 0.001 0.001 2.962 2.992 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.886 2.937 make_images_data 4110 15.4 0.049 0.052 2.285 2.635 fft3d_ps 1111 14.6 1.110 1.335 2.485 2.631 hybrid_alltoall_any 4261 16.3 0.107 0.466 1.947 2.614 calculate_first_density_matrix 1 7.0 0.001 0.003 2.558 2.560 cp_fm_cholesky_invert 11 10.9 2.497 2.503 2.497 2.503 mp_sum_l 6594 12.7 1.736 2.457 1.736 2.457 grid_collocate_task_list 110 9.6 2.173 2.301 2.173 2.301 potential_pw2rs 110 12.3 0.008 0.009 2.202 2.215 jit_kernel_multiply 11 16.3 0.918 2.110 0.918 2.110 qs_energies_init_hamiltonians 11 5.9 0.001 0.008 1.998 2.018 acc_transpose_blocks 24660 15.4 0.115 0.119 1.980 2.007 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.951 1.978 multiply_cannon_metrocomm4 22605 15.4 0.078 0.086 0.773 1.841 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.799 1.813 mp_alltoall_d11v 2046 13.8 1.687 1.795 1.687 1.795 mp_irecv_dv 57340 16.2 0.644 1.617 0.644 1.617 transfer_rs2pw 451 10.6 0.006 0.008 1.466 1.615 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.535 1.546 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.389 1.493 mp_allgather_i34 2055 14.4 0.409 1.446 0.409 1.446 mp_waitany 10164 13.8 1.225 1.420 1.225 1.420 dbcsr_complete_redistribute 325 12.2 0.242 0.297 1.133 1.396 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=68.704000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=562.272727, yerr=7.136219 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 666.619904E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 854880 MPI messages size (bytes): total size 708.322787E+09 min size 0.000000E+00 max size 6.553600E+06 average size 828.564000E+03 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 222984 7302414336 32768 < size <= 131072 153888 10085203968 131072 < size <= 4194304 389376 200257044480 4194304 < size <= 16777216 82208 490679162176 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66421. MP_Allreduce 9774 562. MP_Sync 52 MP_Alltoall 1496 4511006. MP_SendRecv 6820 27424. MP_ISendRecv 6820 27424. MP_Wait 25498 MP_ISend 17072 115022. MP_IRecv 17072 115022. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.034 0.067 60.202 60.204 qs_mol_dyn_low 1 2.0 0.003 0.003 59.764 59.774 qs_forces 11 3.9 0.003 0.003 59.699 59.700 qs_energies 11 4.9 0.009 0.067 56.504 56.515 scf_env_do_scf 11 5.9 0.001 0.001 48.453 48.453 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 39.951 39.952 velocity_verlet 10 3.0 0.001 0.002 32.498 32.500 dbcsr_multiply_generic 2055 12.4 0.106 0.111 28.790 29.122 qs_scf_new_mos 99 7.5 0.001 0.001 25.328 25.440 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.328 25.440 ot_scf_mini 99 9.5 0.002 0.003 24.068 24.192 multiply_cannon 2055 13.4 0.213 0.225 22.430 23.637 multiply_cannon_loop 2055 14.4 0.814 0.849 21.368 22.399 ot_mini 99 10.5 0.001 0.001 13.734 13.858 rebuild_ks_matrix 110 8.3 0.000 0.001 12.150 12.319 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 12.149 12.318 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.699 10.848 multiply_cannon_multrec 16440 15.4 3.504 4.297 9.707 10.539 mp_waitall_1 139946 16.5 7.082 9.895 7.082 9.895 qs_ot_get_derivative 99 11.5 0.001 0.001 9.274 9.394 init_scf_loop 11 6.9 0.001 0.003 8.468 8.469 multiply_cannon_metrocomm3 16440 15.4 0.046 0.048 4.580 7.248 prepare_preconditioner 11 7.9 0.000 0.000 6.704 6.722 make_preconditioner 11 8.9 0.001 0.008 6.704 6.722 make_full_inverse_cholesky 11 9.9 0.000 0.001 5.996 6.348 sum_up_and_integrate 110 10.3 0.061 0.062 6.292 6.306 integrate_v_rspace 110 11.3 0.003 0.003 6.231 6.245 dbcsr_mm_accdrv_process 34862 16.1 5.338 5.709 6.047 6.154 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.717 5.728 calculate_rho_elec 110 8.6 0.058 0.059 5.717 5.727 qs_ot_get_p 110 10.4 0.001 0.001 5.497 5.646 init_scf_run 11 5.9 0.000 0.001 5.569 5.569 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.569 5.569 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.982 5.406 apply_single 110 13.6 0.000 0.000 4.982 5.406 make_m2s 4110 13.4 0.049 0.051 4.168 4.439 ot_diis_step 99 11.5 0.011 0.011 4.430 4.430 make_images 4110 14.4 0.397 0.518 4.055 4.324 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.852 3.857 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.126 3.795 multiply_cannon_sync_h2d 16440 15.4 3.194 3.762 3.194 3.762 pw_transfer 1331 11.6 0.066 0.073 3.567 3.579 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.480 3.480 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.460 3.474 grid_integrate_task_list 110 12.3 3.205 3.367 3.205 3.367 density_rs2pw 110 9.6 0.004 0.005 2.957 3.124 fft_wrap_pw1pw2_140 451 13.1 0.579 0.588 2.989 3.005 wfi_extrapolate 11 7.9 0.001 0.001 2.985 2.985 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.944 2.953 cp_fm_diag_elpa_base 48 14.0 2.879 2.907 2.943 2.952 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.845 2.847 make_images_data 4110 15.4 0.045 0.050 2.315 2.668 cp_fm_cholesky_invert 11 10.9 2.617 2.623 2.617 2.623 hybrid_alltoall_any 4261 16.3 0.109 0.383 2.062 2.605 calculate_dm_sparse 110 9.5 0.001 0.001 2.542 2.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.498 2.560 multiply_cannon_metrocomm4 14385 15.4 0.048 0.052 0.928 2.545 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.448 2.519 calculate_first_density_matrix 1 7.0 0.000 0.002 2.491 2.492 fft3d_ps 1111 14.6 1.100 1.110 2.403 2.422 mp_irecv_dv 48980 15.7 0.853 2.407 0.853 2.407 grid_collocate_task_list 110 9.6 2.225 2.391 2.225 2.391 mp_sum_l 6594 12.7 1.582 2.182 1.582 2.182 potential_pw2rs 110 12.3 0.011 0.011 2.117 2.123 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.992 1.993 mp_alltoall_d11v 2046 13.8 1.806 1.967 1.806 1.967 dbcsr_complete_redistribute 325 12.2 0.330 0.389 1.480 1.937 cp_fm_upper_to_full 70 14.2 1.362 1.742 1.362 1.742 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.646 1.665 acc_transpose_blocks 16440 15.4 0.077 0.079 1.569 1.610 cp_fm_cholesky_decompose 22 10.9 1.538 1.562 1.538 1.562 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.466 1.480 build_core_hamiltonian_matrix_ 11 4.9 0.002 0.009 1.381 1.465 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.022 1.454 transfer_rs2pw 451 10.6 0.005 0.006 1.279 1.451 mp_allgather_i34 2055 14.4 0.372 1.431 0.372 1.431 mp_waitany 17072 13.8 1.155 1.312 1.155 1.312 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.301 1.309 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.204000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.909091, yerr=8.532873 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 732.274688E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 937080 MPI messages size (bytes): total size 523.723932E+09 min size 0.000000E+00 max size 4.537280E+06 average size 558.889250E+03 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 264 2162688 8192 < size <= 32768 304932 8165326848 32768 < size <= 131072 110640 6338641920 131072 < size <= 4194304 489498 400769458320 4194304 < size <= 16777216 24750 108449092400 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66419. MP_Allreduce 9774 603. MP_Sync 52 MP_Alltoall 1496 5863162. MP_SendRecv 5060 43184. MP_ISendRecv 5060 43184. MP_Wait 20042 MP_ISend 13376 163145. MP_IRecv 13376 163145. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.047 66.232 66.234 qs_mol_dyn_low 1 2.0 0.003 0.003 65.942 65.953 qs_forces 11 3.9 0.004 0.005 65.880 65.881 qs_energies 11 4.9 0.001 0.001 62.486 62.490 scf_env_do_scf 11 5.9 0.000 0.001 54.021 54.024 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 42.523 42.524 velocity_verlet 10 3.0 0.002 0.002 37.161 37.163 dbcsr_multiply_generic 2055 12.4 0.113 0.120 31.257 31.487 qs_scf_new_mos 99 7.5 0.001 0.001 27.893 28.005 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.892 28.004 ot_scf_mini 99 9.5 0.002 0.003 26.207 26.309 multiply_cannon 2055 13.4 0.240 0.258 24.096 25.423 multiply_cannon_loop 2055 14.4 1.436 1.501 22.645 23.385 ot_mini 99 10.5 0.001 0.001 15.154 15.284 multiply_cannon_multrec 24660 15.4 4.158 6.904 13.313 14.604 rebuild_ks_matrix 110 8.3 0.000 0.000 12.021 12.132 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.013 12.021 12.131 init_scf_loop 11 6.9 0.000 0.000 11.457 11.458 qs_ot_get_derivative 99 11.5 0.001 0.001 10.996 11.109 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.623 10.722 dbcsr_mm_accdrv_process 52304 16.0 7.799 9.294 8.996 9.977 prepare_preconditioner 11 7.9 0.000 0.000 9.714 9.732 make_preconditioner 11 8.9 0.000 0.000 9.714 9.732 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.970 9.417 sum_up_and_integrate 110 10.3 0.068 0.071 6.258 6.274 integrate_v_rspace 110 11.3 0.003 0.003 6.190 6.205 mp_waitall_1 121746 16.5 4.389 6.191 4.389 6.191 qs_ot_get_p 110 10.4 0.001 0.001 5.870 6.017 make_m2s 4110 13.4 0.060 0.063 5.540 5.943 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.808 5.817 calculate_rho_elec 110 8.6 0.078 0.081 5.807 5.816 make_images 4110 14.4 0.577 0.697 5.399 5.798 init_scf_run 11 5.9 0.000 0.001 5.741 5.742 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.741 5.741 cp_fm_upper_to_full 70 14.2 3.319 4.810 3.319 4.810 ot_diis_step 99 11.5 0.011 0.012 4.120 4.120 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.043 4.106 apply_single 110 13.6 0.000 0.000 4.043 4.105 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.988 4.005 dbcsr_complete_redistribute 325 12.2 0.421 0.493 2.645 3.770 pw_transfer 1331 11.6 0.065 0.075 3.637 3.669 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.569 3.616 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.530 3.566 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.387 3.450 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.419 3.420 grid_integrate_task_list 110 12.3 3.270 3.415 3.270 3.415 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.127 3.232 hybrid_alltoall_any 4261 16.3 0.123 0.458 2.315 3.213 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.420 3.207 make_images_data 4110 15.4 0.049 0.053 2.785 3.184 calculate_dm_sparse 110 9.5 0.001 0.001 3.057 3.093 fft_wrap_pw1pw2_140 451 13.1 0.607 0.631 3.037 3.077 density_rs2pw 110 9.6 0.004 0.004 2.922 3.058 wfi_extrapolate 11 7.9 0.001 0.001 3.009 3.009 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.965 2.966 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.878 2.888 cp_fm_diag_elpa_base 48 14.0 2.730 2.785 2.876 2.886 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.735 2.825 mp_alltoall_i22 605 13.7 1.636 2.794 1.636 2.794 calculate_first_density_matrix 1 7.0 0.000 0.000 2.639 2.640 acc_transpose_blocks 24660 15.4 0.115 0.120 2.537 2.638 cp_fm_cholesky_invert 11 10.9 2.621 2.629 2.621 2.629 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.571 2.603 multiply_cannon_sync_h2d 24660 15.4 2.354 2.562 2.354 2.562 fft3d_ps 1111 14.6 1.087 1.118 2.415 2.429 grid_collocate_task_list 110 9.6 2.263 2.424 2.263 2.424 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.229 2.229 potential_pw2rs 110 12.3 0.013 0.013 2.027 2.035 mp_alltoall_d11v 2046 13.8 1.790 1.987 1.790 1.987 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.801 1.837 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.614 1.711 mp_allgather_i34 2055 14.4 0.530 1.704 0.530 1.704 cp_fm_cholesky_decompose 22 10.9 1.623 1.679 1.623 1.679 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.638 1.648 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.614 1.631 jit_kernel_multiply 8 15.8 0.853 1.599 0.853 1.599 multiply_cannon_metrocomm4 20550 15.4 0.062 0.066 0.848 1.549 acc_transpose_blocks_sync 73980 16.4 1.434 1.535 1.434 1.535 mp_irecv_dv 62702 16.1 0.743 1.467 0.743 1.467 mp_sum_l 6594 12.7 0.941 1.457 0.941 1.457 transfer_rs2pw 451 10.6 0.005 0.006 1.174 1.326 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=66.234000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.090909, yerr=7.971022 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/13/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 807.299199E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 1438408 0.0% 0.0% 100.0% average stack size 0.0 0.0 684.2 marketing flops 15.646297E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 846.766080E+06 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 197280 MPI messages size (bytes): total size 339.125567E+09 min size 0.000000E+00 max size 13.107200E+06 average size 1.719006E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 132 4325376 32768 < size <= 131072 88656 11620319232 131072 < size <= 4194304 89424 117209825280 4194304 < size <= 16777216 17616 210291069504 16777216 < size 0 0 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 7346 33. MP_Alltoall 8043 263767. MP_ISend 32836 654203. MP_IRecv 32836 654587. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3473 66417. MP_Allreduce 9774 644. MP_Sync 52 MP_Alltoall 1496 8504061. MP_SendRecv 3300 54848. MP_ISendRecv 3300 54848. MP_Wait 13926 MP_ISend 9240 278857. MP_IRecv 9240 278857. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.029 54.639 54.641 qs_mol_dyn_low 1 2.0 0.003 0.003 54.381 54.390 qs_forces 11 3.9 0.003 0.003 54.296 54.299 qs_energies 11 4.9 0.001 0.002 50.642 50.650 scf_env_do_scf 11 5.9 0.000 0.001 42.461 42.461 scf_env_do_scf_inner_loop 99 6.5 0.007 0.014 34.906 34.907 velocity_verlet 10 3.0 0.022 0.023 30.840 30.843 dbcsr_multiply_generic 2055 12.4 0.105 0.109 23.121 23.328 qs_scf_new_mos 99 7.5 0.001 0.001 20.539 20.613 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.539 20.612 ot_scf_mini 99 9.5 0.002 0.002 19.303 19.350 multiply_cannon 2055 13.4 0.240 0.250 17.715 18.784 multiply_cannon_loop 2055 14.4 0.603 0.625 16.504 16.879 rebuild_ks_matrix 110 8.3 0.000 0.000 11.498 11.541 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.498 11.540 ot_mini 99 10.5 0.001 0.001 10.779 10.819 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.227 10.265 multiply_cannon_multrec 8220 15.4 3.200 4.643 7.633 8.863 init_scf_loop 11 6.9 0.000 0.000 7.509 7.510 mp_waitall_1 103326 16.6 5.822 7.390 5.822 7.390 qs_ot_get_derivative 99 11.5 0.001 0.001 6.991 7.036 sum_up_and_integrate 110 10.3 0.080 0.082 6.161 6.174 integrate_v_rspace 110 11.3 0.004 0.004 6.081 6.093 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.883 5.900 calculate_rho_elec 110 8.6 0.114 0.114 5.883 5.899 prepare_preconditioner 11 7.9 0.000 0.000 5.880 5.885 make_preconditioner 11 8.9 0.000 0.000 5.880 5.885 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.475 5.550 dbcsr_mm_accdrv_process 17442 15.9 3.177 4.203 4.293 5.241 init_scf_run 11 5.9 0.000 0.001 5.048 5.048 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.048 5.048 qs_ot_get_p 110 10.4 0.001 0.001 4.615 4.672 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 3.094 4.569 make_m2s 4110 13.4 0.038 0.039 4.084 4.277 make_images 4110 14.4 0.647 0.706 3.954 4.144 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.797 3.874 apply_single 110 13.6 0.000 0.000 3.797 3.874 pw_transfer 1331 11.6 0.065 0.070 3.777 3.786 ot_diis_step 99 11.5 0.012 0.014 3.767 3.767 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.670 3.682 grid_integrate_task_list 110 12.3 3.387 3.512 3.387 3.512 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.278 3.281 fft_wrap_pw1pw2_140 451 13.1 0.779 0.791 3.192 3.205 density_rs2pw 110 9.6 0.004 0.005 2.855 3.025 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.943 2.944 cp_fm_cholesky_invert 11 10.9 2.823 2.826 2.823 2.826 qs_energies_init_hamiltonians 11 5.9 0.003 0.010 2.682 2.685 wfi_extrapolate 11 7.9 0.001 0.001 2.665 2.665 hybrid_alltoall_any 4261 16.3 0.200 0.850 2.189 2.557 grid_collocate_task_list 110 9.6 2.364 2.541 2.364 2.541 calculate_dm_sparse 110 9.5 0.001 0.001 2.470 2.508 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.485 2.486 multiply_cannon_sync_h2d 8220 15.4 2.352 2.461 2.352 2.461 make_images_data 4110 15.4 0.040 0.046 2.159 2.413 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.390 2.397 cp_fm_diag_elpa_base 48 14.0 2.335 2.356 2.388 2.395 fft3d_ps 1111 14.6 1.138 1.163 2.332 2.350 calculate_first_density_matrix 1 7.0 0.000 0.000 2.288 2.290 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.124 2.144 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.003 2.041 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.015 potential_pw2rs 110 12.3 0.016 0.016 1.913 1.916 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.877 1.901 mp_alltoall_d11v 2046 13.8 1.696 1.830 1.696 1.830 cp_fm_cholesky_decompose 22 10.9 1.624 1.641 1.624 1.641 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.504 1.618 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.601 1.605 dbcsr_complete_redistribute 325 12.2 0.567 0.588 1.398 1.482 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.436 1.446 mp_allgather_i34 2055 14.4 0.379 1.402 0.379 1.402 acc_transpose_blocks 8220 15.4 0.039 0.041 1.359 1.395 multiply_cannon_metrocomm4 6165 15.4 0.019 0.021 0.485 1.390 mp_irecv_dv 24056 15.7 0.458 1.346 0.458 1.346 qs_create_task_list 11 7.9 0.001 0.001 1.211 1.308 generate_qs_task_list 11 8.9 0.375 0.444 1.210 1.307 transfer_rs2pw 451 10.6 0.005 0.005 1.096 1.264 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.183 1.214 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 0.765 1.213 mp_waitany 9240 13.8 1.042 1.206 1.042 1.206 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.641000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=800.000000, yerr=12.105596 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 1.404645E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.066 86.588 86.605 qs_mol_dyn_low 1 2.0 0.004 0.010 86.228 86.243 qs_forces 11 3.9 0.004 0.006 86.142 86.153 qs_energies 11 4.9 0.001 0.001 82.008 82.018 scf_env_do_scf 11 5.9 0.001 0.001 71.830 71.830 velocity_verlet 10 3.0 0.002 0.002 55.158 55.164 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.936 43.937 dbcsr_multiply_generic 2055 12.4 0.120 0.126 29.714 29.851 init_scf_loop 11 6.9 0.000 0.000 27.824 27.826 qs_scf_new_mos 99 7.5 0.001 0.001 26.772 26.822 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.771 26.821 prepare_preconditioner 11 7.9 0.000 0.000 25.836 25.845 make_preconditioner 11 8.9 0.000 0.000 25.836 25.845 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.126 25.313 ot_scf_mini 99 9.5 0.002 0.002 24.965 25.007 multiply_cannon 2055 13.4 0.335 0.372 22.701 23.307 multiply_cannon_loop 2055 14.4 0.826 0.849 20.931 21.249 cp_fm_upper_to_full 70 14.2 12.518 17.951 12.518 17.951 ot_mini 99 10.5 0.001 0.001 14.172 14.218 rebuild_ks_matrix 110 8.3 0.001 0.001 13.568 13.630 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.568 13.630 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.311 12.368 dbcsr_complete_redistribute 325 12.2 1.022 1.053 7.274 10.494 multiply_cannon_multrec 8220 15.4 4.057 4.200 9.933 10.053 qs_ot_get_derivative 99 11.5 0.001 0.001 9.543 9.590 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.253 9.469 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.694 8.882 mp_alltoall_i22 605 13.7 5.330 8.512 5.330 8.512 mp_waitall_1 84994 16.7 7.443 8.323 7.443 8.323 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.986 7.018 calculate_rho_elec 110 8.6 0.225 0.226 6.986 7.018 sum_up_and_integrate 110 10.3 0.150 0.151 6.778 6.794 integrate_v_rspace 110 11.3 0.004 0.004 6.628 6.643 dbcsr_mm_accdrv_process 11614 15.7 3.951 4.155 5.727 5.944 init_scf_run 11 5.9 0.000 0.001 5.911 5.911 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.911 5.911 make_m2s 4110 13.4 0.043 0.044 5.218 5.689 qs_ot_get_p 110 10.4 0.001 0.001 5.493 5.551 make_images 4110 14.4 0.886 0.940 5.029 5.500 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.761 5.244 apply_single 110 13.6 0.000 0.000 4.760 5.243 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 4.937 5.235 cp_fm_cholesky_invert 11 10.9 5.135 5.139 5.135 5.139 pw_transfer 1331 11.6 0.075 0.075 4.850 4.859 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.733 4.742 ot_diis_step 99 11.5 0.015 0.017 4.607 4.607 fft_wrap_pw1pw2_140 451 13.1 1.277 1.281 4.144 4.154 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.943 3.953 grid_integrate_task_list 110 12.3 3.673 3.728 3.673 3.728 qs_energies_init_hamiltonians 11 5.9 0.002 0.005 3.663 3.674 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.074 3.551 density_rs2pw 110 9.6 0.004 0.004 3.450 3.464 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.426 3.427 hybrid_alltoall_any 4261 16.3 0.263 0.563 2.769 3.372 calculate_dm_sparse 110 9.5 0.001 0.001 3.291 3.323 wfi_extrapolate 11 7.9 0.001 0.001 3.290 3.290 make_images_data 4110 15.4 0.045 0.048 2.686 3.282 multiply_cannon_sync_h2d 8220 15.4 3.124 3.144 3.124 3.144 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.984 2.988 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.850 2.850 cp_fm_diag_elpa_base 48 14.0 2.317 2.511 2.848 2.849 fft3d_ps 1111 14.6 1.297 1.309 2.786 2.792 grid_collocate_task_list 110 9.6 2.663 2.683 2.663 2.683 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.650 2.678 calculate_first_density_matrix 1 7.0 0.000 0.000 2.508 2.509 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.478 2.499 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.174 2.224 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.105 2.223 potential_pw2rs 110 12.3 0.022 0.022 2.201 2.203 mp_alltoall_d11v 2046 13.8 1.996 2.045 1.996 2.045 cp_fm_cholesky_decompose 22 10.9 1.946 1.969 1.946 1.969 qs_create_task_list 11 7.9 0.000 0.000 1.880 1.928 generate_qs_task_list 11 8.9 0.731 0.783 1.879 1.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.855 1.860 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.783 1.802 acc_transpose_blocks 8220 15.4 0.041 0.041 1.752 1.799 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.744 1.787 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.605000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1270.636364, yerr=54.766069 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 1.094965E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755939872 0.0% 0.0% 100.0% number of processed stacks 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 631.742464E+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 3175955098256 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 57905. MP_Allreduce 11059 797. MP_Sync 87 MP_Alltoall 2226 2673823. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.066 205.395 205.397 qs_mol_dyn_low 1 2.0 0.005 0.018 204.775 204.813 qs_forces 11 3.9 0.005 0.007 204.633 204.635 qs_energies 11 4.9 0.005 0.028 199.027 199.047 scf_env_do_scf 11 5.9 0.001 0.002 182.558 182.563 scf_env_do_scf_inner_loop 117 6.6 0.005 0.016 162.185 162.188 dbcsr_multiply_generic 2507 12.6 0.179 0.183 125.127 125.889 qs_scf_new_mos 117 7.6 0.001 0.001 123.045 123.467 qs_scf_loop_do_ot 117 8.6 0.001 0.001 123.044 123.467 velocity_verlet 10 3.0 0.002 0.006 122.970 122.973 ot_scf_mini 117 9.6 0.003 0.003 116.390 116.739 multiply_cannon 2507 13.6 0.241 0.250 101.466 103.171 multiply_cannon_loop 2507 14.6 2.392 2.455 99.374 101.140 ot_mini 117 10.6 0.001 0.001 65.860 66.265 multiply_cannon_multrec 60168 15.6 31.994 33.903 41.635 43.371 qs_ot_get_derivative 117 11.6 0.001 0.001 40.968 41.332 rebuild_ks_matrix 128 8.3 0.001 0.001 33.482 33.830 qs_ks_build_kohn_sham_matrix 128 9.3 0.019 0.032 33.482 33.829 mp_waitall_1 267128 16.5 28.488 30.793 28.488 30.793 qs_ks_update_qs_env 128 7.6 0.001 0.001 30.059 30.379 multiply_cannon_sync_h2d 60168 15.6 26.502 29.344 26.502 29.344 qs_ot_get_p 128 10.4 0.001 0.001 28.384 28.826 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.370 25.082 apply_single 128 13.6 0.001 0.001 24.369 25.081 ot_diis_step 117 11.6 0.008 0.008 24.635 24.636 qs_ot_p2m_diag 83 11.4 0.079 0.094 21.852 21.940 init_scf_loop 11 6.9 0.000 0.001 20.300 20.301 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 19.572 19.923 cp_dbcsr_syevd 83 12.4 0.005 0.005 19.129 19.130 multiply_cannon_metrocomm3 60168 15.6 0.121 0.128 15.950 18.002 cp_fm_diag_elpa 83 13.4 0.000 0.000 16.243 16.277 cp_fm_diag_elpa_base 83 14.4 16.178 16.221 16.240 16.275 prepare_preconditioner 11 7.9 0.000 0.000 15.712 15.765 make_preconditioner 11 8.9 0.000 0.000 15.712 15.765 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.964 15.132 make_m2s 5014 13.6 0.105 0.114 13.769 14.243 make_images 5014 14.6 0.397 0.416 13.586 14.067 sum_up_and_integrate 128 10.3 0.089 0.107 13.842 13.855 integrate_v_rspace 128 11.3 0.004 0.005 13.752 13.768 qs_rho_update_rho_low 128 7.7 0.001 0.001 12.960 13.187 calculate_rho_elec 128 8.7 0.045 0.064 12.959 13.186 init_scf_run 11 5.9 0.000 0.001 12.311 12.311 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.310 12.311 mp_sum_l 7950 12.9 8.639 10.028 8.639 10.028 dbcsr_mm_accdrv_process 124484 16.2 4.602 4.823 9.205 9.772 wfi_extrapolate 11 7.9 0.001 0.001 9.028 9.028 cp_fm_cholesky_invert 11 10.9 8.966 8.975 8.966 8.975 calculate_dm_sparse 128 9.5 0.001 0.001 8.485 8.611 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.017 8.125 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.994 8.122 multiply_cannon_metrocomm1 60168 15.6 0.098 0.102 6.185 8.107 make_images_data 5014 15.6 0.068 0.073 6.748 7.705 pw_transfer 1547 11.6 0.074 0.090 7.486 7.657 grid_integrate_task_list 128 12.3 7.028 7.453 7.028 7.453 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 7.284 7.453 hybrid_alltoall_any 5200 16.5 0.295 2.261 5.952 7.420 density_rs2pw 128 9.7 0.006 0.007 6.593 7.331 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.005 6.668 6.679 fft_wrap_pw1pw2_140 523 13.2 1.137 1.180 6.447 6.616 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.806 5.915 fft3d_ps 1291 14.7 2.191 2.834 5.278 5.585 mp_alltoall_d11v 2415 14.1 4.285 5.289 4.285 5.289 grid_collocate_task_list 128 9.7 4.835 5.247 4.835 5.247 cp_fm_cholesky_decompose 22 10.9 4.586 4.600 4.586 4.600 potential_pw2rs 128 12.3 0.009 0.011 4.347 4.363 mp_sum_d 4465 12.1 3.448 4.243 3.448 4.243 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=205.397000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=596.909091, yerr=7.102555 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.166472E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5925696 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.6 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 846.929920E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2385600 MPI messages size (bytes): total size 4.069300E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.705776E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70188 2295595008 32768 < size <= 131072 716032 54973693952 131072 < size <= 4194304 1363760 1386318135296 4194304 < size <= 16777216 153648 1453843137296 16777216 < size 67056 1171888537600 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58284. MP_Allreduce 11071 960. MP_Sync 86 MP_Alltoall 1955 5001485. MP_SendRecv 11938 47072. MP_ISendRecv 11938 47072. MP_Wait 25718 MP_ISend 11660 212488. MP_IRecv 11660 212488. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.056 0.076 187.740 187.745 qs_mol_dyn_low 1 2.0 0.003 0.004 187.210 187.222 qs_forces 11 3.9 0.005 0.007 187.115 187.119 qs_energies 11 4.9 0.005 0.030 180.358 180.372 scf_env_do_scf 11 5.9 0.001 0.002 163.987 163.997 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 131.531 131.533 velocity_verlet 10 3.0 0.004 0.009 117.537 117.538 dbcsr_multiply_generic 2485 12.5 0.187 0.192 96.489 97.769 qs_scf_new_mos 116 7.6 0.001 0.001 93.408 93.928 qs_scf_loop_do_ot 116 8.6 0.001 0.001 93.407 93.927 ot_scf_mini 116 9.6 0.004 0.005 88.636 89.205 multiply_cannon 2485 13.5 0.499 0.553 76.619 81.058 multiply_cannon_loop 2485 14.5 1.562 1.631 73.319 76.067 ot_mini 116 10.6 0.001 0.001 49.663 50.173 mp_waitall_1 212858 16.6 23.555 37.377 23.555 37.377 multiply_cannon_multrec 29820 15.5 20.991 25.396 31.501 36.217 rebuild_ks_matrix 127 8.3 0.001 0.001 32.037 32.701 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 32.036 32.700 init_scf_loop 11 6.9 0.001 0.007 32.368 32.369 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.772 29.379 multiply_cannon_metrocomm3 29820 15.5 0.097 0.102 15.293 28.609 qs_ot_get_derivative 116 11.6 0.001 0.001 27.970 28.533 prepare_preconditioner 11 7.9 0.000 0.001 28.057 28.108 make_preconditioner 11 8.9 0.000 0.002 28.057 28.108 make_full_inverse_cholesky 11 9.9 0.000 0.001 26.789 27.327 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 21.869 22.998 apply_single 127 13.6 0.001 0.001 21.869 22.998 qs_ot_get_p 127 10.4 0.001 0.001 21.190 21.831 ot_diis_step 116 11.6 0.014 0.015 21.522 21.524 multiply_cannon_sync_h2d 29820 15.5 17.964 20.284 17.964 20.284 cp_fm_cholesky_invert 11 10.9 16.412 16.424 16.412 16.424 qs_ot_p2m_diag 82 11.4 0.186 0.215 16.327 16.361 make_m2s 4970 13.5 0.089 0.094 13.951 15.511 make_images 4970 14.5 1.152 1.347 13.744 15.304 cp_dbcsr_syevd 82 12.4 0.005 0.006 15.142 15.143 sum_up_and_integrate 127 10.3 0.116 0.133 13.910 13.948 integrate_v_rspace 127 11.3 0.004 0.004 13.794 13.838 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.204 13.230 calculate_rho_elec 127 8.7 0.086 0.103 13.204 13.230 cp_fm_diag_elpa 82 13.4 0.000 0.000 12.051 12.091 cp_fm_diag_elpa_base 82 14.4 11.802 11.885 12.045 12.081 init_scf_run 11 5.9 0.000 0.001 11.464 11.465 scf_env_initial_rho_setup 11 6.9 0.001 0.002 11.463 11.465 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.005 11.417 multiply_cannon_metrocomm4 27335 15.5 0.104 0.116 3.801 10.827 dbcsr_mm_accdrv_process 61726 16.2 5.365 6.107 9.960 10.519 mp_irecv_dv 68888 16.3 3.598 10.433 3.598 10.433 make_images_data 4970 15.5 0.067 0.076 8.250 10.215 hybrid_alltoall_any 5155 16.4 0.350 1.501 7.053 9.537 pw_transfer 1535 11.6 0.084 0.096 8.443 8.487 wfi_extrapolate 11 7.9 0.001 0.001 8.346 8.346 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.221 8.269 grid_integrate_task_list 127 12.3 7.135 7.488 7.135 7.488 fft_wrap_pw1pw2_140 519 13.2 1.216 1.243 7.307 7.368 density_rs2pw 127 9.7 0.006 0.007 6.958 7.347 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.282 7.029 cp_fm_cholesky_decompose 22 10.9 6.809 6.892 6.809 6.892 calculate_dm_sparse 127 9.5 0.001 0.001 6.432 6.567 mp_sum_l 7884 12.9 4.147 6.200 4.147 6.200 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.091 6.101 fft3d_ps 1281 14.7 2.824 2.971 5.786 5.819 qs_ot_get_orbitals 116 10.6 0.001 0.001 5.308 5.377 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.207 5.376 grid_collocate_task_list 127 9.7 4.968 5.375 4.968 5.375 mp_allgather_i34 2485 14.5 1.827 4.968 1.827 4.968 mp_alltoall_d11v 2401 14.1 3.993 4.545 3.993 4.545 potential_pw2rs 127 12.3 0.015 0.017 4.430 4.449 dbcsr_complete_redistribute 393 12.7 0.773 0.879 3.083 3.940 mp_sum_d 4449 12.1 2.600 3.914 2.600 3.914 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=187.745000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=803.000000, yerr=6.822423 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 944.267264E+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 4002 58208. MP_Allreduce 11082 999. MP_Sync 87 MP_Alltoall 1712 9388896. MP_SendRecv 7936 75008. MP_ISendRecv 7936 75008. MP_Wait 21820 MP_ISend 11748 275205. MP_IRecv 11748 275205. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.059 173.055 173.056 qs_mol_dyn_low 1 2.0 0.003 0.003 172.566 172.579 qs_forces 11 3.9 0.005 0.006 172.381 172.390 qs_energies 11 4.9 0.006 0.039 165.829 165.855 scf_env_do_scf 11 5.9 0.001 0.002 150.295 150.295 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 115.886 115.886 velocity_verlet 10 3.0 0.002 0.002 111.054 111.059 dbcsr_multiply_generic 2507 12.6 0.180 0.188 81.411 82.522 qs_scf_new_mos 117 7.6 0.001 0.001 79.170 79.529 qs_scf_loop_do_ot 117 8.6 0.001 0.001 79.169 79.528 ot_scf_mini 117 9.6 0.004 0.004 74.949 75.380 multiply_cannon 2507 13.6 0.506 0.527 61.519 65.716 multiply_cannon_loop 2507 14.6 1.127 1.195 58.947 61.390 ot_mini 117 10.6 0.001 0.001 42.277 42.696 mp_waitall_1 170520 16.6 24.711 34.660 24.711 34.660 init_scf_loop 11 6.9 0.001 0.010 34.312 34.313 rebuild_ks_matrix 128 8.3 0.001 0.001 30.007 30.505 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 30.006 30.505 prepare_preconditioner 11 7.9 0.000 0.000 30.336 30.391 make_preconditioner 11 8.9 0.000 0.002 30.336 30.391 make_full_inverse_cholesky 11 9.9 0.000 0.002 28.051 29.387 qs_ks_update_qs_env 128 7.6 0.001 0.001 27.023 27.478 multiply_cannon_multrec 20056 15.6 13.090 16.476 22.399 25.803 multiply_cannon_metrocomm3 20056 15.6 0.063 0.067 15.155 24.734 qs_ot_get_derivative 117 11.6 0.001 0.002 22.672 23.096 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 19.678 20.649 apply_single 128 13.6 0.001 0.001 19.678 20.648 ot_diis_step 117 11.6 0.018 0.019 19.495 19.496 qs_ot_get_p 128 10.4 0.001 0.001 18.067 18.530 make_m2s 5014 13.6 0.080 0.086 14.955 15.666 multiply_cannon_sync_h2d 20056 15.6 13.549 15.597 13.549 15.597 make_images 5014 14.6 1.143 1.243 14.722 15.437 cp_fm_cholesky_invert 11 10.9 14.150 14.158 14.150 14.158 qs_ot_p2m_diag 83 11.4 0.266 0.273 13.920 13.929 sum_up_and_integrate 128 10.3 0.133 0.145 13.902 13.928 integrate_v_rspace 128 11.3 0.004 0.004 13.768 13.797 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.378 13.420 calculate_rho_elec 128 8.7 0.130 0.145 13.377 13.420 cp_dbcsr_syevd 83 12.4 0.005 0.006 12.879 12.880 init_scf_run 11 5.9 0.000 0.001 10.484 10.484 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.484 10.484 make_images_data 5014 15.6 0.063 0.072 9.081 10.232 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.784 9.803 cp_fm_diag_elpa_base 83 14.4 9.384 9.529 9.781 9.800 hybrid_alltoall_any 5200 16.5 0.451 2.051 7.746 9.671 multiply_cannon_metrocomm4 17549 15.6 0.066 0.076 3.512 9.520 mp_irecv_dv 50230 16.2 3.383 9.264 3.383 9.264 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 8.922 9.230 dbcsr_mm_accdrv_process 41502 16.2 5.590 5.975 8.762 8.957 pw_transfer 1547 11.6 0.085 0.102 8.580 8.677 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.356 8.460 grid_integrate_task_list 128 12.3 7.303 7.675 7.303 7.675 fft_wrap_pw1pw2_140 523 13.2 1.312 1.357 7.462 7.571 wfi_extrapolate 11 7.9 0.001 0.001 7.356 7.356 cp_fm_cholesky_decompose 22 10.9 7.183 7.212 7.183 7.212 density_rs2pw 128 9.7 0.006 0.007 6.765 7.107 cp_fm_upper_to_full 105 14.8 5.626 7.072 5.626 7.072 dbcsr_complete_redistribute 395 12.7 1.179 1.207 4.492 6.189 calculate_dm_sparse 128 9.5 0.001 0.001 5.820 5.901 fft3d_ps 1291 14.7 2.762 2.975 5.723 5.797 grid_collocate_task_list 128 9.7 5.162 5.554 5.162 5.554 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.479 5.485 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.512 5.242 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.348 5.042 mp_alltoall_d11v 2415 14.1 4.189 4.707 4.189 4.707 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.570 4.703 mp_sum_l 7950 12.9 3.104 4.405 3.104 4.405 mp_allgather_i34 2507 14.6 1.064 4.380 1.064 4.380 potential_pw2rs 128 12.3 0.021 0.023 4.284 4.299 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.947 3.978 transfer_fm_to_dbcsr 11 9.9 0.020 0.024 2.266 3.949 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.764 3.774 mp_alltoall_i22 716 14.1 1.862 3.695 1.862 3.695 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.483 3.526 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=173.056000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=888.454545, yerr=10.360446 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.187332E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837692480 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4005 58161. MP_Allreduce 11090 1083. MP_Sync 86 MP_Alltoall 1700 12496371. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.075 0.154 185.393 185.401 qs_mol_dyn_low 1 2.0 0.003 0.004 184.705 184.735 qs_forces 11 3.9 0.004 0.006 184.257 184.265 qs_energies 11 4.9 0.006 0.019 177.160 177.169 scf_env_do_scf 11 5.9 0.001 0.002 160.460 160.470 velocity_verlet 10 3.0 0.002 0.003 122.256 122.274 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 114.504 114.515 dbcsr_multiply_generic 2485 12.5 0.190 0.205 79.397 79.926 qs_scf_new_mos 116 7.6 0.001 0.001 78.835 79.158 qs_scf_loop_do_ot 116 8.6 0.001 0.001 78.834 79.157 ot_scf_mini 116 9.6 0.004 0.005 74.283 74.603 multiply_cannon 2485 13.5 0.546 0.582 55.097 57.788 multiply_cannon_loop 2485 14.5 1.834 1.950 51.581 52.940 init_scf_loop 11 6.9 0.002 0.010 45.836 45.875 ot_mini 116 10.6 0.001 0.001 42.188 42.507 prepare_preconditioner 11 7.9 0.000 0.004 41.589 41.609 make_preconditioner 11 8.9 0.003 0.025 41.589 41.609 make_full_inverse_cholesky 11 9.9 0.011 0.021 35.338 40.269 multiply_cannon_multrec 29820 15.5 13.782 18.675 26.575 31.141 rebuild_ks_matrix 127 8.3 0.001 0.001 28.957 29.191 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.020 28.956 29.190 mp_waitall_1 146592 16.7 16.496 27.056 16.496 27.056 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.107 26.323 qs_ot_get_derivative 116 11.6 0.001 0.002 22.872 23.194 make_m2s 4970 13.5 0.096 0.100 20.019 20.803 make_images 4970 14.5 1.930 2.235 19.714 20.504 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 18.744 19.205 apply_single 127 13.6 0.001 0.001 18.744 19.205 ot_diis_step 116 11.6 0.017 0.018 19.172 19.173 qs_ot_get_p 127 10.4 0.002 0.008 18.594 18.954 cp_fm_upper_to_full 104 14.8 11.279 16.608 11.279 16.608 cp_fm_cholesky_invert 11 10.9 15.849 15.858 15.849 15.858 multiply_cannon_metrocomm3 29820 15.5 0.049 0.053 6.003 15.177 qs_ot_p2m_diag 82 11.4 0.339 0.385 14.394 14.444 sum_up_and_integrate 127 10.3 0.139 0.151 13.883 13.915 integrate_v_rspace 127 11.3 0.004 0.004 13.743 13.779 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.450 13.494 calculate_rho_elec 127 8.7 0.171 0.187 13.449 13.493 cp_dbcsr_syevd 82 12.4 0.005 0.005 13.017 13.018 dbcsr_mm_accdrv_process 61748 16.2 8.426 9.342 12.367 12.885 dbcsr_complete_redistribute 393 12.7 1.515 1.639 8.937 12.619 make_images_data 4970 15.5 0.066 0.073 10.684 12.359 hybrid_alltoall_any 5155 16.4 0.527 2.212 9.560 11.597 copy_fm_to_dbcsr 208 11.6 0.002 0.002 7.588 11.291 multiply_cannon_sync_h2d 29820 15.5 10.404 11.200 10.404 11.200 init_scf_run 11 5.9 0.000 0.001 10.738 10.740 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.738 10.739 cp_fm_diag_elpa 82 13.4 0.000 0.000 9.890 9.901 cp_fm_diag_elpa_base 82 14.4 8.949 9.271 9.884 9.894 transfer_fm_to_dbcsr 11 9.9 0.001 0.009 6.225 9.835 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.536 9.779 mp_alltoall_i22 712 14.1 5.454 9.064 5.454 9.064 pw_transfer 1535 11.6 0.085 0.099 8.809 8.892 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.585 8.671 grid_integrate_task_list 127 12.3 7.454 7.913 7.454 7.913 fft_wrap_pw1pw2_140 519 13.2 1.446 1.478 7.687 7.783 cp_fm_cholesky_decompose 22 10.9 7.457 7.546 7.457 7.546 wfi_extrapolate 11 7.9 0.001 0.001 7.455 7.455 density_rs2pw 127 9.7 0.006 0.006 6.593 7.080 multiply_cannon_metrocomm4 24850 15.5 0.084 0.094 2.697 6.588 mp_irecv_dv 75445 16.2 2.544 6.328 2.544 6.328 calculate_dm_sparse 127 9.5 0.001 0.001 6.223 6.306 fft3d_ps 1281 14.7 2.823 2.893 5.770 5.819 grid_collocate_task_list 127 9.7 5.245 5.636 5.245 5.636 mp_alltoall_d11v 2401 14.1 4.738 5.539 4.738 5.539 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.431 5.460 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.462 4.557 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 4.445 4.446 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.370 4.431 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.184 4.272 potential_pw2rs 127 12.3 0.023 0.024 4.190 4.204 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=185.401000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1097.545455, yerr=29.093466 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 5.820059E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.575645E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108580288 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4001 58216. MP_Allreduce 11080 1167. MP_Sync 86 MP_Alltoall 1700 18828158. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.016 0.031 167.643 167.644 qs_mol_dyn_low 1 2.0 0.003 0.003 167.209 167.221 qs_forces 11 3.9 0.004 0.004 167.117 167.119 qs_energies 11 4.9 0.001 0.002 159.709 159.718 scf_env_do_scf 11 5.9 0.001 0.001 142.532 142.543 velocity_verlet 10 3.0 0.002 0.002 109.882 109.885 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 107.006 107.007 dbcsr_multiply_generic 2485 12.5 0.182 0.193 71.445 71.966 qs_scf_new_mos 116 7.6 0.001 0.001 70.823 70.916 qs_scf_loop_do_ot 116 8.6 0.001 0.001 70.822 70.916 ot_scf_mini 116 9.6 0.003 0.004 66.449 66.540 multiply_cannon 2485 13.5 0.555 0.579 53.209 56.318 multiply_cannon_loop 2485 14.5 0.803 0.839 50.076 51.212 ot_mini 116 10.6 0.001 0.001 37.356 37.466 init_scf_loop 11 6.9 0.000 0.000 35.375 35.376 prepare_preconditioner 11 7.9 0.000 0.000 31.557 31.580 make_preconditioner 11 8.9 0.000 0.000 31.557 31.580 mp_waitall_1 124680 16.7 24.492 30.973 24.492 30.973 make_full_inverse_cholesky 11 9.9 0.013 0.025 29.413 29.666 rebuild_ks_matrix 127 8.3 0.001 0.001 28.685 28.820 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 28.684 28.820 qs_ks_update_qs_env 127 7.6 0.001 0.001 26.149 26.276 multiply_cannon_multrec 9940 15.5 10.246 16.130 17.762 22.286 qs_ot_get_derivative 116 11.6 0.001 0.002 20.252 20.346 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 12.830 20.270 cp_fm_cholesky_invert 11 10.9 18.036 18.042 18.036 18.042 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 17.163 17.380 apply_single 127 13.6 0.001 0.001 17.163 17.380 ot_diis_step 116 11.6 0.020 0.021 17.028 17.029 qs_ot_get_p 127 10.4 0.001 0.001 16.508 16.652 make_m2s 4970 13.5 0.066 0.070 14.411 15.347 make_images 4970 14.5 2.142 2.566 14.108 15.037 sum_up_and_integrate 127 10.3 0.180 0.190 14.031 14.084 integrate_v_rspace 127 11.3 0.004 0.004 13.851 13.912 qs_rho_update_rho_low 127 7.7 0.001 0.001 13.738 13.771 calculate_rho_elec 127 8.7 0.254 0.265 13.738 13.770 qs_ot_p2m_diag 82 11.4 0.490 0.495 12.803 12.818 cp_dbcsr_syevd 82 12.4 0.005 0.006 11.666 11.667 multiply_cannon_sync_h2d 9940 15.5 10.630 11.123 10.630 11.123 init_scf_run 11 5.9 0.000 0.001 10.359 10.359 scf_env_initial_rho_setup 11 6.9 0.001 0.002 10.358 10.359 make_images_data 4970 15.5 0.055 0.064 8.232 9.603 hybrid_alltoall_any 5155 16.4 0.838 3.760 8.101 9.218 pw_transfer 1535 11.6 0.085 0.093 9.114 9.132 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 8.894 8.921 cp_fm_diag_elpa 82 13.4 0.000 0.000 8.862 8.873 cp_fm_diag_elpa_base 82 14.4 8.629 8.705 8.858 8.870 grid_integrate_task_list 127 12.3 7.720 7.960 7.720 7.960 qs_ot_get_derivative_diag 76 12.4 0.002 0.003 7.888 7.954 fft_wrap_pw1pw2_140 519 13.2 1.801 1.844 7.904 7.938 cp_fm_cholesky_decompose 22 10.9 7.812 7.934 7.812 7.934 dbcsr_mm_accdrv_process 20590 16.1 3.144 4.128 7.165 7.883 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.221 7.370 wfi_extrapolate 11 7.9 0.001 0.001 7.273 7.273 density_rs2pw 127 9.7 0.006 0.006 6.553 7.041 calculate_dm_sparse 127 9.5 0.001 0.001 6.053 6.148 multiply_cannon_metrocomm4 7455 15.5 0.026 0.030 1.858 5.882 grid_collocate_task_list 127 9.7 5.522 5.864 5.522 5.864 mp_irecv_dv 28618 15.9 1.818 5.789 1.818 5.789 fft3d_ps 1281 14.7 2.748 2.828 5.633 5.666 dbcsr_complete_redistribute 393 12.7 2.103 2.159 5.254 5.612 qs_energies_init_hamiltonians 11 5.9 0.002 0.003 5.258 5.259 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.125 5.138 mp_alltoall_d11v 2401 14.1 4.611 5.000 4.611 5.000 mp_allgather_i34 2485 14.5 1.328 4.540 1.328 4.540 potential_pw2rs 127 12.3 0.027 0.027 4.150 4.163 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.041 4.096 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.581 3.936 copy_fm_to_dbcsr 208 11.6 0.001 0.002 3.396 3.708 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.586 3.675 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.613 3.665 copy_dbcsr_to_fm 185 11.7 0.004 0.004 3.450 3.523 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.432 3.443 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=167.644000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1477.636364, yerr=49.503485 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022121472 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444702699520 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796573E+12 0.0% 0.0% 100.0% flops max/rank 11.606413E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499744 0.0% 0.0% 100.0% number of processed stacks 1947808 0.0% 0.0% 100.0% average stack size 0.0 0.0 3442.6 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 3.088286E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 99400 MPI messages size (bytes): total size 1.127422E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.342275E+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 44768 34745614336 4194304 < size <= 16777216 43984 376564613120 16777216 < size 10032 716108580288 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 59121. MP_Allreduce 11085 1504. MP_Sync 86 MP_Alltoall 1700 36954383. MP_SendRecv 1778 218624. MP_ISendRecv 1778 218624. MP_Wait 9728 MP_ISend 6360 1080477. MP_IRecv 6360 1080477. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.053 0.097 289.302 289.312 qs_mol_dyn_low 1 2.0 0.003 0.004 288.716 288.749 qs_forces 11 3.9 0.004 0.005 288.598 288.601 qs_energies 11 4.9 0.002 0.003 279.775 279.780 scf_env_do_scf 11 5.9 0.001 0.001 257.140 257.148 velocity_verlet 10 3.0 0.013 0.018 209.183 209.194 scf_env_do_scf_inner_loop 116 6.6 0.011 0.015 132.933 132.935 init_scf_loop 11 6.9 0.000 0.000 123.931 123.933 prepare_preconditioner 11 7.9 0.000 0.000 119.076 119.105 make_preconditioner 11 8.9 0.000 0.000 119.076 119.105 make_full_inverse_cholesky 11 9.9 0.038 0.039 94.814 116.221 qs_scf_new_mos 116 7.6 0.001 0.001 89.822 89.971 qs_scf_loop_do_ot 116 8.6 0.001 0.001 89.821 89.970 ot_scf_mini 116 9.6 0.004 0.004 84.931 85.016 dbcsr_multiply_generic 2485 12.5 0.211 0.222 82.443 83.233 cp_fm_upper_to_full 104 14.8 52.496 75.779 52.496 75.779 multiply_cannon 2485 13.5 0.677 0.718 59.086 60.146 multiply_cannon_loop 2485 14.5 1.043 1.075 54.815 56.891 ot_mini 116 10.6 0.001 0.001 44.657 44.772 dbcsr_complete_redistribute 393 12.7 4.029 4.077 30.339 43.640 copy_fm_to_dbcsr 208 11.6 0.001 0.001 26.787 40.036 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.219 37.263 mp_alltoall_i22 712 14.1 22.023 35.433 22.023 35.433 rebuild_ks_matrix 127 8.3 0.001 0.001 33.928 34.078 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 33.928 34.077 mp_waitall_1 102768 16.8 28.390 33.652 28.390 33.652 cp_fm_cholesky_invert 11 10.9 33.225 33.231 33.225 33.231 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.452 31.589 qs_ot_get_p 127 10.4 0.001 0.001 25.171 25.289 qs_ot_get_derivative 116 11.6 0.002 0.002 24.520 24.607 multiply_cannon_metrocomm3 9940 15.5 0.025 0.027 19.076 21.141 qs_ot_p2m_diag 82 11.4 0.869 0.874 21.102 21.128 make_m2s 4970 13.5 0.075 0.078 18.618 20.337 ot_diis_step 116 11.6 0.022 0.022 20.090 20.090 make_images 4970 14.5 3.039 3.214 18.142 19.865 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 19.301 19.627 apply_single 127 13.6 0.001 0.001 19.300 19.627 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.337 19.339 multiply_cannon_multrec 9940 15.5 10.202 11.956 17.934 18.006 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.273 16.281 calculate_rho_elec 127 8.7 0.478 0.478 16.272 16.280 cp_fm_diag_elpa 82 13.4 0.000 0.000 16.216 16.217 cp_fm_diag_elpa_base 82 14.4 11.848 13.492 16.213 16.214 sum_up_and_integrate 127 10.3 0.318 0.321 16.060 16.148 integrate_v_rspace 127 11.3 0.004 0.005 15.741 15.830 multiply_cannon_sync_h2d 9940 15.5 14.189 14.221 14.189 14.221 hybrid_alltoall_any 5155 16.4 1.308 3.051 10.877 13.416 make_images_data 4970 15.5 0.064 0.068 10.598 13.004 init_scf_run 11 5.9 0.000 0.001 12.342 12.343 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.342 12.343 pw_transfer 1535 11.6 0.092 0.094 11.186 11.196 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 10.952 10.961 fft_wrap_pw1pw2_140 519 13.2 3.049 3.083 9.783 9.798 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.638 9.702 dbcsr_mm_accdrv_process 20590 16.0 4.227 6.107 7.484 9.361 wfi_extrapolate 11 7.9 0.001 0.001 9.109 9.110 cp_fm_cholesky_decompose 22 10.9 8.866 8.894 8.866 8.894 grid_integrate_task_list 127 12.3 8.492 8.678 8.492 8.678 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.073 8.073 density_rs2pw 127 9.7 0.006 0.006 7.385 7.433 mp_alltoall_d11v 2401 14.1 6.834 6.930 6.834 6.930 calculate_dm_sparse 127 9.5 0.001 0.001 6.776 6.848 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.546 6.592 grid_collocate_task_list 127 9.7 6.365 6.438 6.365 6.438 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.236 6.343 fft3d_ps 1281 14.7 2.830 2.845 6.208 6.242 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=289.312000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2836.909091, yerr=136.014341 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 1.259352E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 458208 MPI messages size (bytes): total size 3.456111E+12 min size 0.000000E+00 max size 18.735064E+06 average size 7.542668E+06 MPI breakdown and total messages size (bytes): size <= 128 112896 0 128 < size <= 8192 0 0 8192 < size <= 32768 224 5687808 32768 < size <= 131072 10528 813356544 131072 < size <= 4194304 36422 76284728544 4194304 < size <= 16777216 294266 3312457683808 16777216 < size 3872 66548597808 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 255646. MP_Allreduce 3139 6114. MP_Sync 4 MP_Alltoall 54 MP_SendRecv 285 19200. MP_ISendRecv 285 19200. MP_Wait 1017 MP_ISend 642 197829. MP_IRecv 642 197607. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.018 0.037 84.905 84.907 qs_energies 1 2.0 0.000 0.000 84.346 84.357 ls_scf 1 3.0 0.000 0.000 83.434 83.446 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.522 72.713 multiply_cannon 111 7.7 0.018 0.020 55.932 57.204 multiply_cannon_loop 111 8.7 0.224 0.238 52.476 54.006 ls_scf_main 1 4.0 0.000 0.000 52.166 52.166 density_matrix_trs4 2 5.0 0.002 0.003 46.631 46.706 ls_scf_init_scf 1 4.0 0.000 0.000 28.212 28.213 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.093 27.140 mp_waitall_1 11031 10.9 22.398 25.030 22.398 25.030 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 24.971 24.990 multiply_cannon_multrec 2664 9.7 8.130 8.891 15.504 17.252 multiply_cannon_sync_h2d 2664 9.7 13.515 15.793 13.515 15.793 make_m2s 222 7.7 0.009 0.010 12.946 13.516 make_images 222 8.7 0.098 0.109 12.924 13.495 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.683 11.987 make_images_data 222 9.7 0.004 0.005 7.521 7.981 dbcsr_mm_accdrv_process 4760 10.4 0.588 0.714 6.992 7.935 multiply_cannon_metrocomm3 2664 9.7 0.010 0.011 5.529 7.862 hybrid_alltoall_any 227 10.6 0.219 1.837 6.472 7.834 dbcsr_mm_accdrv_process_sort 4760 11.4 6.203 7.088 6.203 7.088 calculate_norms 4752 9.8 5.498 6.109 5.498 6.109 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.073 5.183 mp_sum_l 887 5.1 3.115 4.551 3.115 4.551 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.068 3.436 mp_irecv_dv 6231 10.9 2.051 3.405 2.051 3.405 make_images_sizes 222 9.7 0.000 0.000 0.674 3.327 mp_alltoall_i44 222 10.7 0.674 3.327 0.674 3.327 arnoldi_extremal 4 6.8 0.000 0.000 3.169 3.197 arnoldi_normal_ev 4 7.8 0.001 0.003 3.169 3.197 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.314 3.188 build_subspace 16 8.4 0.009 0.012 3.067 3.072 ls_scf_post 1 4.0 0.000 0.000 3.056 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.867 2.912 dbcsr_special_finalize 555 9.7 0.005 0.006 2.397 2.794 dbcsr_merge_single_wm 555 10.7 0.471 0.597 2.388 2.785 make_images_pack 222 9.7 2.207 2.614 2.209 2.616 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.304 2.549 dbcsr_sort_data 658 11.4 2.174 2.523 2.174 2.523 dbcsr_matrix_vector_mult_local 304 10.0 2.065 2.460 2.067 2.462 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.276 2.357 buffer_matrices_ensure_size 222 8.7 1.762 2.106 1.762 2.106 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.757 1.758 rebuild_ks_matrix 3 7.3 0.000 0.000 1.748 1.749 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.748 1.749 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=84.907000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1131.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 2.144571E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 106560 MPI messages size (bytes): total size 2.699093E+12 min size 0.000000E+00 max size 72.286792E+06 average size 25.329324E+06 MPI breakdown and total messages size (bytes): size <= 128 23040 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 3264 325830144 131072 < size <= 4194304 5280 3328561104 4194304 < size <= 16777216 12709 156766962056 16777216 < size 62267 2538670978840 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 10075. MP_Sync 4 MP_Alltoall 47 15335933. MP_SendRecv 141 57600. MP_ISendRecv 141 57600. MP_Wait 687 MP_ISend 462 414589. MP_IRecv 462 413870. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.043 0.073 89.501 89.503 qs_energies 1 2.0 0.000 0.001 88.816 88.820 ls_scf 1 3.0 0.000 0.001 87.400 87.403 dbcsr_multiply_generic 111 6.7 0.016 0.023 73.551 73.858 multiply_cannon 111 7.7 0.028 0.047 52.354 56.056 ls_scf_main 1 4.0 0.000 0.008 53.959 53.965 multiply_cannon_loop 111 8.7 0.135 0.147 49.639 52.785 density_matrix_trs4 2 5.0 0.003 0.011 48.289 48.492 ls_scf_init_scf 1 4.0 0.000 0.002 29.794 29.795 mp_waitall_1 9105 10.9 20.324 29.184 20.324 29.184 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.595 28.681 multiply_cannon_multrec 1332 9.7 13.377 17.544 22.655 27.948 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.219 26.233 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.161 19.896 make_m2s 222 7.7 0.007 0.022 14.949 15.511 make_images 222 8.7 1.368 1.699 14.917 15.482 dbcsr_mm_accdrv_process 4041 10.4 0.343 0.542 8.875 10.461 dbcsr_mm_accdrv_process_sort 4041 11.4 8.392 9.919 8.392 9.919 make_images_data 222 9.7 0.004 0.005 8.647 9.464 hybrid_alltoall_any 227 10.6 0.543 2.547 8.046 9.304 mp_sum_l 887 5.1 4.793 7.827 4.793 7.827 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.191 7.604 mp_irecv_dv 3311 11.0 3.171 7.554 3.171 7.554 calculate_norms 2376 9.8 6.068 6.840 6.068 6.840 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.626 6.266 multiply_cannon_sync_h2d 1332 9.7 4.906 6.060 4.906 6.060 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.919 5.160 arnoldi_extremal 4 6.8 0.000 0.000 4.650 4.669 arnoldi_normal_ev 4 7.8 0.001 0.005 4.650 4.669 build_subspace 16 8.4 0.014 0.021 4.382 4.385 ls_scf_post 1 4.0 0.000 0.001 3.648 3.652 ls_scf_store_result 1 5.0 0.000 0.000 3.391 3.458 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.116 3.346 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.214 2.739 3.216 mp_allgather_i34 111 8.7 0.706 2.865 0.706 2.865 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.616 2.734 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.165 2.536 dbcsr_data_new 4174 10.1 2.109 2.411 2.109 2.411 make_images_pack 222 9.7 1.829 2.138 1.831 2.140 dbcsr_sort_data 436 11.2 1.819 2.044 1.819 2.044 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.869 1.872 rebuild_ks_matrix 3 7.3 0.000 0.000 1.856 1.859 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.002 1.856 1.859 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=89.503000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1754.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 2.847912E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 46176 MPI messages size (bytes): total size 1.924064E+12 min size 0.000000E+00 max size 108.059888E+06 average size 41.668048E+06 MPI breakdown and total messages size (bytes): size <= 128 9984 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 3328 1170063360 4194304 < size <= 16777216 1870 19378539600 16777216 < size 30994 1903514987232 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265448. MP_Allreduce 3138 10896. MP_Sync 4 MP_Alltoall 47 23526250. MP_SendRecv 93 57600. MP_ISendRecv 93 57600. MP_Wait 639 MP_ISend 462 560046. MP_IRecv 462 560662. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.042 0.089 93.071 93.073 qs_energies 1 2.0 0.000 0.001 92.368 92.372 ls_scf 1 3.0 0.000 0.001 90.938 90.942 dbcsr_multiply_generic 111 6.7 0.017 0.023 75.626 75.937 ls_scf_main 1 4.0 0.009 0.043 56.785 56.789 multiply_cannon 111 7.7 0.039 0.095 52.183 56.141 multiply_cannon_loop 111 8.7 0.116 0.132 49.416 53.401 density_matrix_trs4 2 5.0 0.002 0.005 50.925 51.156 mp_waitall_1 7281 11.0 23.795 33.367 23.795 33.367 ls_scf_init_scf 1 4.0 0.000 0.002 30.471 30.474 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.183 29.265 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.820 26.834 multiply_cannon_multrec 888 9.7 12.630 15.224 21.198 24.341 multiply_cannon_metrocomm3 888 9.7 0.004 0.005 11.221 23.034 make_m2s 222 7.7 0.006 0.007 16.506 17.234 make_images 222 8.7 1.580 1.843 16.468 17.197 make_images_data 222 9.7 0.004 0.004 9.568 10.758 hybrid_alltoall_any 227 10.6 0.640 2.939 9.106 10.378 dbcsr_mm_accdrv_process 3754 10.4 0.309 0.488 8.084 9.332 dbcsr_mm_accdrv_process_sort 3754 11.4 7.650 8.844 7.650 8.844 mp_sum_l 887 5.1 5.257 8.696 5.257 8.696 multiply_cannon_sync_h2d 888 9.7 5.989 7.137 5.989 7.137 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.487 6.972 mp_irecv_dv 2335 11.1 2.471 6.926 2.471 6.926 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 4.027 6.887 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.651 6.884 arnoldi_extremal 4 6.8 0.000 0.000 5.054 5.076 arnoldi_normal_ev 4 7.8 0.001 0.005 5.054 5.076 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.715 4.951 calculate_norms 1584 9.8 4.397 4.856 4.397 4.856 build_subspace 16 8.4 0.014 0.020 4.729 4.735 mp_allgather_i34 111 8.7 0.877 3.807 0.877 3.807 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.418 3.754 ls_scf_post 1 4.0 0.000 0.000 3.681 3.686 dbcsr_matrix_vector_mult_local 304 10.0 3.018 3.594 3.020 3.597 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.507 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.822 2.908 make_images_sizes 222 9.7 0.000 0.000 1.120 2.505 mp_alltoall_i44 222 10.7 1.120 2.504 1.120 2.504 dbcsr_data_new 4116 9.9 2.101 2.452 2.101 2.452 dbcsr_sort_data 325 11.1 1.873 2.089 1.873 2.089 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.961 1.963 rebuild_ks_matrix 3 7.3 0.000 0.000 1.943 1.945 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.943 1.945 make_images_pack 222 9.7 1.625 1.888 1.628 1.891 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=93.073000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2209.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 3.409711E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 50616 MPI messages size (bytes): total size 1.536549E+12 min size 0.000000E+00 max size 72.286792E+06 average size 30.356986E+06 MPI breakdown and total messages size (bytes): size <= 128 10368 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 1056 104411904 131072 < size <= 4194304 3168 831638784 4194304 < size <= 16777216 3103 33613273640 16777216 < size 32921 1501999894888 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 266673. MP_Allreduce 3138 13030. MP_Sync 4 MP_Alltoall 47 30278988. MP_SendRecv 69 86400. MP_ISendRecv 69 86400. MP_Wait 531 MP_ISend 378 823502. MP_IRecv 378 823753. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.076 0.199 97.121 97.122 qs_energies 1 2.0 0.001 0.007 96.224 96.229 ls_scf 1 3.0 0.000 0.001 94.481 94.485 dbcsr_multiply_generic 111 6.7 0.020 0.030 78.140 78.390 ls_scf_main 1 4.0 0.002 0.022 58.476 58.476 multiply_cannon 111 7.7 0.041 0.076 51.627 56.350 density_matrix_trs4 2 5.0 0.005 0.020 52.368 52.471 multiply_cannon_loop 111 8.7 0.154 0.168 46.551 49.319 ls_scf_init_scf 1 4.0 0.000 0.002 32.721 32.723 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.404 31.485 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 28.890 28.905 mp_waitall_1 6369 11.0 22.489 28.707 22.489 28.707 multiply_cannon_multrec 1332 9.7 14.176 17.401 22.077 24.592 make_m2s 222 7.7 0.007 0.008 21.050 22.504 make_images 222 8.7 3.146 3.621 20.999 22.457 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.021 16.935 make_images_data 222 9.7 0.004 0.005 11.714 13.365 hybrid_alltoall_any 227 10.6 0.801 3.779 11.081 12.927 dbcsr_mm_accdrv_process 3641 10.4 0.302 0.478 7.540 9.073 dbcsr_mm_accdrv_process_sort 3641 11.4 7.099 8.578 7.099 8.578 mp_sum_l 887 5.1 3.878 7.250 3.878 7.250 multiply_cannon_sync_h2d 1332 9.7 5.475 6.067 5.475 6.067 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.095 6.012 mp_irecv_dv 3229 10.9 2.069 5.939 2.069 5.939 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.949 5.726 arnoldi_extremal 4 6.8 0.000 0.000 5.200 5.215 arnoldi_normal_ev 4 7.8 0.002 0.011 5.200 5.214 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.588 4.954 build_subspace 16 8.4 0.014 0.021 4.838 4.849 mp_allgather_i34 111 8.7 2.191 4.732 2.191 4.732 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.555 4.722 calculate_norms 2376 9.8 4.186 4.521 4.186 4.521 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 3.579 3.874 dbcsr_matrix_vector_mult_local 304 10.0 3.188 3.687 3.190 3.689 dbcsr_sort_data 658 11.4 3.116 3.391 3.116 3.391 ls_scf_post 1 4.0 0.000 0.001 3.284 3.290 dbcsr_special_finalize 555 9.7 0.006 0.007 2.862 3.259 dbcsr_merge_single_wm 555 10.7 0.542 0.662 2.854 3.251 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.000 3.068 ls_scf_store_result 1 5.0 0.000 0.000 2.989 3.046 dbcsr_data_release 10477 10.7 1.585 2.415 1.585 2.415 dbcsr_finalize 304 7.8 0.049 0.061 1.810 1.984 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.122000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2698.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 4.735627E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 10656 MPI messages size (bytes): total size 1.149035E+12 min size 0.000000E+00 max size 203.538048E+06 average size 107.829832E+06 MPI breakdown and total messages size (bytes): size <= 128 2304 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 768 702038016 4194304 < size <= 16777216 0 0 16777216 < size 7584 1148332810224 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 2 12. MP_Allreduce 705 128. MP_Alltoall 310 12920694. MP_ISend 1776 40180424. MP_IRecv 1776 40465030. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 265536. MP_Allreduce 3129 15263. MP_Sync 4 MP_Alltoall 47 46208988. MP_SendRecv 45 115200. MP_ISendRecv 45 115200. MP_Wait 528 MP_ISend 420 924980. MP_IRecv 420 924528. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.051 0.054 92.312 92.314 qs_energies 1 2.0 0.000 0.000 91.356 91.359 ls_scf 1 3.0 0.000 0.000 89.423 89.427 dbcsr_multiply_generic 111 6.7 0.018 0.019 71.139 71.354 ls_scf_main 1 4.0 0.000 0.000 56.802 56.803 multiply_cannon 111 7.7 0.085 0.152 52.777 55.748 multiply_cannon_loop 111 8.7 0.090 0.117 50.157 51.939 density_matrix_trs4 2 5.0 0.002 0.003 49.911 49.978 ls_scf_init_scf 1 4.0 0.000 0.000 29.197 29.199 mp_waitall_1 5436 11.0 24.805 28.794 24.805 28.794 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.944 27.980 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.875 25.887 multiply_cannon_multrec 444 9.7 13.654 16.083 20.731 22.205 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.794 15.410 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.388 15.174 make_m2s 222 7.7 0.005 0.005 13.639 14.601 make_images 222 8.7 2.033 2.472 13.571 14.531 hybrid_alltoall_any 227 10.6 0.802 3.822 8.219 9.903 make_images_data 222 9.7 0.003 0.004 8.439 9.736 multiply_cannon_sync_h2d 444 9.7 6.734 7.898 6.734 7.898 dbcsr_mm_accdrv_process 3003 10.4 0.328 0.399 6.763 7.888 dbcsr_mm_accdrv_process_sort 3003 11.4 6.400 7.491 6.400 7.491 arnoldi_extremal 4 6.8 0.000 0.000 5.781 5.795 arnoldi_normal_ev 4 7.8 0.002 0.005 5.781 5.795 build_subspace 16 8.4 0.015 0.020 5.391 5.403 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.398 4.595 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.158 4.361 mp_sum_l 887 5.1 2.724 4.252 2.724 4.252 dbcsr_matrix_vector_mult_local 304 10.0 3.697 4.164 3.699 4.166 mp_allgather_i34 111 8.7 1.197 3.807 1.197 3.807 calculate_norms 792 9.8 3.612 3.768 3.612 3.768 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.542 3.764 mp_irecv_dv 1241 11.2 1.528 3.718 1.528 3.718 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.475 3.529 ls_scf_post 1 4.0 0.000 0.000 3.424 3.427 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.942 3.371 make_images_sizes 222 9.7 0.000 0.000 0.821 3.288 mp_alltoall_i44 222 10.7 0.820 3.287 0.820 3.287 ls_scf_store_result 1 5.0 0.000 0.000 3.203 3.238 dbcsr_finalize 304 7.8 0.062 0.077 2.192 2.275 dbcsr_data_new 4608 9.7 1.792 2.237 1.792 2.237 dbcsr_merge_all 275 8.9 0.479 0.531 2.050 2.119 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.022 2.022 rebuild_ks_matrix 3 7.3 0.000 0.000 1.989 1.990 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.989 1.990 qs_energies_init_hamiltonians 1 3.0 0.012 0.029 1.917 1.917 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.314000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3788.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_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 8.871125E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.088 0.117 108.415 108.415 qs_energies 1 2.0 0.000 0.000 106.832 106.839 ls_scf 1 3.0 0.000 0.000 103.902 103.908 dbcsr_multiply_generic 111 6.7 0.024 0.027 77.691 77.825 ls_scf_main 1 4.0 0.000 0.000 63.309 63.309 density_matrix_trs4 2 5.0 0.002 0.003 54.378 54.432 multiply_cannon 111 7.7 0.111 0.179 48.547 50.732 multiply_cannon_loop 111 8.7 0.098 0.100 45.565 48.839 ls_scf_init_scf 1 4.0 0.000 0.000 36.809 36.810 ls_scf_init_matrix_S 1 5.0 0.000 0.000 35.303 35.329 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 32.615 32.626 mp_waitall_1 4527 11.1 21.646 25.107 21.646 25.107 make_m2s 222 7.7 0.005 0.005 22.607 23.695 make_images 222 8.7 3.573 3.890 22.500 23.585 multiply_cannon_multrec 444 9.7 17.768 18.547 22.400 22.958 hybrid_alltoall_any 227 10.6 1.655 3.619 12.837 15.753 make_images_data 222 9.7 0.003 0.004 13.035 15.135 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 10.016 10.504 multiply_cannon_sync_h2d 444 9.7 8.791 8.844 8.791 8.844 arnoldi_extremal 4 6.8 0.000 0.000 7.365 7.391 arnoldi_normal_ev 4 7.8 0.003 0.009 7.365 7.390 build_subspace 16 8.4 0.026 0.036 6.813 6.822 dbcsr_matrix_vector_mult 304 9.0 0.017 0.034 5.450 5.606 dbcsr_matrix_vector_mult_local 304 10.0 5.042 5.347 5.045 5.350 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.067 5.155 mp_sum_l 887 5.1 4.136 5.102 4.136 5.102 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.851 5.100 dbcsr_mm_accdrv_process 1814 10.4 0.269 0.362 4.453 4.585 dbcsr_mm_accdrv_process_sort 1814 11.4 4.114 4.248 4.114 4.248 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.522 4.170 compute_matrix_preconditioner 1 6.0 0.002 0.002 3.982 3.988 ls_scf_post 1 4.0 0.000 0.000 3.784 3.790 ls_scf_store_result 1 5.0 0.000 0.000 3.506 3.543 make_images_sizes 222 9.7 0.000 0.000 1.448 3.497 mp_alltoall_i44 222 10.7 1.448 3.496 1.448 3.496 mp_allgather_i34 111 8.7 1.092 3.487 1.092 3.487 acc_transpose_blocks 444 9.7 0.003 0.003 0.635 3.475 acc_transpose_blocks_kernels 444 10.7 0.006 0.006 0.552 3.392 jit_kernel_transpose 1 13.0 0.547 3.387 0.547 3.387 calculate_norms 792 9.8 3.229 3.275 3.229 3.275 dbcsr_finalize 304 7.8 0.082 0.089 3.072 3.169 dbcsr_merge_all 275 8.9 0.886 0.914 2.858 2.949 qs_energies_init_hamiltonians 1 3.0 0.001 0.002 2.900 2.900 dbcsr_complete_redistribute 5 7.6 1.433 1.477 2.739 2.855 dbcsr_sort_data 325 11.1 2.441 2.508 2.441 2.508 matrix_ls_to_qs 2 6.0 0.000 0.000 2.378 2.482 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.405 2.407 dbcsr_data_new 6591 9.6 1.882 2.389 1.882 2.389 rebuild_ks_matrix 3 7.3 0.000 0.000 2.339 2.341 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.339 2.341 dbcsr_new_transposed 4 7.5 0.242 0.253 2.295 2.323 dbcsr_frobenius_norm 74 6.6 2.058 2.137 2.200 2.228 dbcsr_add_d 103 6.2 0.000 0.000 2.126 2.201 dbcsr_add_anytype 103 7.2 0.859 0.891 2.126 2.200 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=108.415000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7053.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b_performance_tests/27/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 7009386627072 0.0% 0.0% 100.0% flops 9 x 9 x 32 7335108845568 0.0% 0.0% 100.0% flops 9 x 22 x 32 9866241589248 0.0% 0.0% 100.0% flops 22 x 9 x 32 9884108906496 0.0% 0.0% 100.0% flops 22 x 22 x 32 13354440523776 0.0% 0.0% 100.0% flops 32 x 32 x 9 20607185977344 0.0% 0.0% 100.0% flops 32 x 32 x 22 25186560638976 0.0% 0.0% 100.0% flops 9 x 32 x 32 28458319085568 0.0% 0.0% 100.0% flops 22 x 32 x 32 34782389993472 0.0% 0.0% 100.0% flops 9 x 32 x 9 42881542373376 0.0% 0.0% 100.0% flops 22 x 32 x 9 55680402235392 0.0% 0.0% 100.0% flops 9 x 32 x 22 55680402235392 0.0% 0.0% 100.0% flops 22 x 32 x 22 72328573419520 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 383.054662E+12 0.0% 0.0% 100.0% flops max/rank 733.641090E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 26899403712 0.0% 0.0% 100.0% number of processed stacks 118860288 0.0% 0.0% 100.0% average stack size 0.0 0.0 226.3 marketing flops 780.439111E+12 ------------------------------------------------------------------------------- # multiplications 1445 max memory usage/rank 588.447744E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 102097920 MPI messages size (bytes): total size 37.227590E+12 min size 0.000000E+00 max size 4.551360E+06 average size 364.626312E+03 MPI breakdown and total messages size (bytes): size <= 128 731472 0 128 < size <= 8192 11922720 97670922240 8192 < size <= 32768 24718992 614677610496 32768 < size <= 131072 20000256 1970081366016 131072 < size <= 4194304 42515668 24886801223040 4194304 < size <= 16777216 2208812 9656099886720 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4640 78072. MP_Allreduce 13232 2081. MP_Sync 1064 MP_Alltoall 2588 4804797. MP_SendRecv 168740 11136. MP_ISendRecv 92040 11136. MP_Wait 102830 MP_comm_split 40 MP_ISend 26090 85106. MP_IRecv 37890 59644. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.045 0.096 301.288 301.291 qs_mol_dyn_low 1 2.0 0.003 0.016 299.954 300.002 qs_forces 5 3.8 0.005 0.017 299.836 299.841 qs_energies 5 4.8 0.004 0.060 296.284 296.364 scf_env_do_scf 5 5.8 0.000 0.001 280.003 280.006 scf_env_do_scf_inner_loop 105 6.6 0.005 0.022 239.233 239.234 qs_scf_new_mos 105 7.6 0.001 0.001 185.628 186.762 qs_scf_loop_do_ot 105 8.6 0.001 0.001 185.628 186.761 ot_scf_mini 105 9.6 0.003 0.005 173.781 174.620 dbcsr_multiply_generic 1445 12.2 0.126 0.135 152.557 158.011 multiply_cannon 1445 13.2 0.280 0.292 121.900 138.336 multiply_cannon_loop 1445 14.2 2.855 3.020 118.286 135.780 velocity_verlet 4 3.0 0.003 0.012 135.226 135.227 mp_waitall_1 488190 16.1 47.224 76.845 47.224 76.845 ot_mini 105 10.6 0.001 0.016 70.668 70.940 qs_ot_get_p 112 10.4 0.001 0.001 69.506 70.175 multiply_cannon_metrocomm3 69360 15.2 0.203 0.213 31.053 64.669 qs_ot_p2m_diag 40 11.0 0.020 0.032 53.262 53.398 cp_dbcsr_syevd 40 12.0 0.002 0.002 49.228 49.230 qs_ot_get_derivative 55 11.6 0.001 0.001 48.053 48.450 multiply_cannon_multrec 69360 15.2 28.667 33.696 38.578 43.624 cp_fm_syevd 40 13.0 0.000 0.000 43.311 43.476 init_scf_loop 7 6.6 0.000 0.003 40.706 40.708 cp_fm_redistribute_end 40 14.0 19.099 38.067 19.109 38.077 cp_fm_syevd_base 40 14.0 18.953 37.946 18.953 37.946 rebuild_ks_matrix 110 8.4 0.000 0.000 35.279 36.266 qs_ks_build_kohn_sham_matrix 110 9.4 0.014 0.040 35.278 36.266 prepare_preconditioner 7 7.6 0.000 0.000 34.604 34.648 make_preconditioner 7 8.6 0.000 0.001 34.604 34.648 qs_ks_update_qs_env 112 7.6 0.001 0.001 32.664 33.531 multiply_cannon_metrocomm1 69360 15.2 0.101 0.109 7.729 33.449 multiply_cannon_sync_h2d 69360 15.2 27.714 32.604 27.714 32.604 qs_rho_update_rho_low 110 7.6 0.001 0.002 25.134 25.778 calculate_rho_elec 110 8.6 0.030 0.032 25.133 25.778 make_full_inverse_cholesky 7 9.6 0.000 0.001 25.354 25.474 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.607 23.896 apply_single 62 13.6 0.000 0.000 23.606 23.896 mp_sum_l 4764 12.2 18.668 22.408 18.668 22.408 ot_new_cg_direction 55 11.6 0.001 0.003 21.550 21.553 qs_ot_get_orbitals 105 10.6 0.001 0.001 17.805 19.010 density_rs2pw 110 9.6 0.005 0.007 17.693 18.164 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 17.878 18.134 cp_fm_cholesky_invert 7 10.6 17.591 17.623 17.591 17.623 pw_transfer 1645 12.4 0.080 0.106 15.034 15.489 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 14.894 15.352 calculate_dm_sparse 110 9.5 0.000 0.001 13.539 13.991 transfer_rs2pw 445 10.6 0.007 0.008 13.066 13.579 init_scf_run 5 5.8 0.000 0.001 13.294 13.295 scf_env_initial_rho_setup 5 6.8 0.001 0.005 13.293 13.294 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 12.893 13.091 fft_wrap_pw1pw2_240 915 15.0 1.137 1.227 12.586 13.044 make_m2s 2890 13.2 0.078 0.084 11.539 12.716 make_images 2890 14.2 0.241 0.259 11.434 12.607 sum_up_and_integrate 60 10.3 0.028 0.031 12.460 12.471 integrate_v_rspace 60 11.3 0.002 0.002 12.432 12.444 qs_vxc_create 110 10.4 0.006 0.048 12.174 12.239 fft3d_pb 915 16.0 2.361 2.662 10.799 11.370 check_diag 80 13.5 8.594 8.886 10.487 10.647 dbcsr_mm_accdrv_process 154766 15.8 6.151 6.403 9.785 10.607 mp_alltoall_z22v 2340 17.7 7.938 9.004 7.938 9.004 make_full_single_inverse 7 9.6 0.001 0.001 8.833 8.877 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 8.170 8.446 xc_vxc_pw_create 60 11.3 0.038 0.050 8.238 8.302 mp_alltoall_d11v 1300 13.8 6.763 7.988 6.763 7.988 potential_pw2rs 60 12.3 0.003 0.003 7.827 7.876 xc_pw_derive 510 13.4 0.005 0.006 7.747 7.848 acc_transpose_blocks 69360 15.2 0.361 0.377 7.191 7.812 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.007 7.683 7.701 calculate_first_density_matrix 1 7.0 0.000 0.012 7.186 7.208 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 6.785 7.144 cp_fm_cholesky_decompose 14 10.2 7.072 7.080 7.072 7.080 make_images_sizes 2890 15.2 0.004 0.004 5.648 6.920 mp_alltoall_i44 2890 16.2 5.644 6.917 5.644 6.917 transfer_rs2pw_30 110 11.6 1.299 1.352 6.354 6.877 transfer_pw2rs 245 13.2 0.003 0.004 6.611 6.661 mp_sendrecv_dv 168740 12.6 6.225 6.435 6.225 6.435 mp_waitany 7680 13.5 5.682 6.373 5.682 6.373 make_images_data 2890 15.2 0.070 0.078 4.593 6.027 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=301.291000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=561.000000, yerr=2.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: 5b040937cb7487e394f71cbb1d2cd1b7a7dc8f5b Summary: empty Status: OK