=== 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: b1e573e5365af53cd5b64772cda048b47a415cd7 ################# 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, # HDF5 1.14.2, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, # PLUMED 2.9.0, 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: 04.10.2023 # # \ 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=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_HDF5 := 1.14.2 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.9.0 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(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)) 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__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 += $(LIBVDWXC_LIB)/libvdwxc.a endif ifneq ($(USE_HDF5),) USE_HDF5 := $(strip $(USE_HDF5)) HDF5_INC := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/include HDF5_LIB := $(INSTALL_PATH)/hdf5-$(USE_HDF5)/lib CFLAGS += -I$(HDF5_INC) DFLAGS += -D__HDF5 LIBS += $(HDF5_LIB)/libhdf5_fortran.a LIBS += $(HDF5_LIB)/libhdf5_hl.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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/01 job id: 49679452 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/02 job id: 49679453 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/03 job id: 49679454 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/04 job id: 49679455 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/05 job id: 49679456 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/06 job id: 49679457 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/07 job id: 49679458 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/08 job id: 49679459 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/09 job id: 49679460 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/10 job id: 49679461 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/11 job id: 49679462 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/12 job id: 49679463 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/13 job id: 49679464 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/14 job id: 49679465 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/15 job id: 49679466 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/16 job id: 49679467 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/17 job id: 49679468 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/18 job id: 49679469 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/19 job id: 49679471 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/20 job id: 49679472 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/21 job id: 49679473 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/22 job id: 49679474 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/23 job id: 49679475 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/24 job id: 49679476 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/25 job id: 49679477 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/26 job id: 49679478 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/27 job id: 49679479 --- 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/b1e573e5365af53cd5b64772cda048b47a415cd7_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 0.019 0.032 133.518 133.519 farming_run 1 2.0 132.891 132.896 133.485 133.489 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.488278E+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 0.024 114.220 114.220 qs_energies 1 2.0 0.000 0.000 113.973 113.975 mp2_main 1 3.0 0.000 0.000 111.937 111.940 mp2_gpw_main 1 4.0 0.027 0.034 110.964 110.966 mp2_ri_gpw_compute_in 1 5.0 0.172 0.174 92.035 92.319 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.004 54.580 54.864 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.162 40.967 45.995 get_2c_integrals 1 6.0 0.008 0.009 36.640 37.283 integrate_v_rspace 273 8.0 0.440 0.454 24.586 29.437 pw_transfer 6555 10.6 0.377 0.393 26.699 27.091 grid_integrate_task_list 273 9.0 20.501 25.813 20.501 25.813 fft_wrap_pw1pw2 5465 11.4 0.045 0.050 25.331 25.806 fft_wrap_pw1pw2_100 2178 12.4 1.094 1.280 22.885 23.372 rpa_ri_compute_en 1 5.0 0.033 0.038 18.810 18.925 compute_2c_integrals 1 7.0 0.004 0.006 18.912 18.913 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 18.515 18.592 mp2_eri_2c_integrate_gpw 1 9.0 2.395 2.434 18.512 18.591 cp_fm_cholesky_decompose 12 8.2 17.650 18.304 17.650 18.304 cholesky_decomp 1 7.0 0.000 0.000 16.572 17.231 fft3d_s 5443 13.4 16.067 16.346 16.090 16.369 ao_to_mo_and_store_B_mult_1 272 7.0 10.774 15.321 10.774 15.321 calculate_wavefunction 272 8.0 5.380 5.531 12.292 12.984 rpa_num_int 1 6.0 0.001 0.014 10.623 10.623 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.581 10.585 calc_mat_Q 8 8.0 0.000 0.000 9.455 9.533 contract_S_to_Q 8 9.0 0.000 0.000 8.876 8.955 calc_potential_gpw 544 9.5 0.005 0.005 8.237 8.658 parallel_gemm_fm 14 9.1 0.000 0.000 8.457 8.564 parallel_gemm_fm_cosma 14 10.1 8.457 8.564 8.457 8.564 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.002 8.045 8.249 potential_pw2rs 545 10.0 0.107 0.109 7.439 8.064 create_integ_mat 1 6.0 0.021 0.028 7.793 7.793 collocate_single_gaussian 272 10.0 0.039 0.042 7.271 7.490 array2fm 1 7.0 0.000 0.000 6.621 7.068 pw_scatter_s 2720 13.7 4.347 4.567 4.347 4.567 pw_gather_s 2722 13.2 3.424 3.826 3.424 3.826 array2fm_buffer_send 1 8.0 2.912 3.129 2.912 3.129 pw_poisson_solve 545 10.5 1.116 1.166 2.188 2.302 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=110.965376, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2804.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 0.025 0.036 421.039 421.040 farming_run 1 2.0 420.342 420.347 421.005 421.006 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.244692E+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 9307640. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.011 0.033 217.940 217.941 qs_energies 1 2.0 0.000 0.000 217.721 217.730 scf_env_do_scf 1 3.0 0.000 0.000 115.206 115.206 qs_ks_update_qs_env 5 5.0 0.000 0.000 114.306 114.313 rebuild_ks_matrix 4 6.0 0.000 0.000 114.305 114.312 qs_ks_build_kohn_sham_matrix 4 7.0 0.054 0.061 114.305 114.312 hfx_ks_matrix 4 8.0 0.001 0.001 113.957 113.962 integrate_four_center 4 9.0 0.154 0.466 113.957 113.961 integrate_four_center_main 4 10.0 0.118 0.526 101.853 105.451 integrate_four_center_bin 267 11.0 101.735 105.447 101.735 105.447 mp2_main 1 3.0 0.002 0.017 102.218 102.228 mp2_gpw_main 1 4.0 0.046 0.076 101.363 101.373 init_scf_loop 1 4.0 0.000 0.000 97.316 97.317 mp2_ri_gpw_compute_in 1 5.0 0.064 0.064 74.019 75.143 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.002 53.788 54.913 mp2_eri_3c_integrate_gpw 91 7.0 0.145 0.164 41.567 46.599 integrate_v_rspace 95 8.0 0.397 0.562 28.023 32.866 pw_transfer 2240 10.6 0.143 0.169 29.577 30.036 mp2_ri_gpw_compute_en 1 5.0 0.056 0.066 27.185 29.253 fft_wrap_pw1pw2 1868 11.4 0.017 0.021 28.575 29.090 grid_integrate_task_list 95 9.0 23.430 28.482 23.430 28.482 ao_to_mo_and_store_B_mult_1 91 7.0 10.548 26.892 10.548 26.892 fft_wrap_pw1pw2_100 730 12.4 1.301 1.500 26.322 26.883 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.824 1.902 25.402 25.414 get_2c_integrals 1 6.0 0.000 0.000 20.139 20.166 compute_2c_integrals 1 7.0 0.002 0.003 19.124 19.126 fft3d_s 1823 13.4 18.504 18.996 18.518 19.010 compute_2c_integrals_loop_lm 1 8.0 0.001 0.001 18.707 19.002 mp2_eri_2c_integrate_gpw 1 9.0 1.726 1.862 18.706 19.001 scf_env_do_scf_inner_loop 4 4.0 0.000 0.000 17.888 17.889 calculate_wavefunction 91 8.0 2.014 2.047 9.630 9.846 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.580 8.746 9.273 potential_pw2rs 186 10.0 0.033 0.035 8.435 8.981 mp2_ri_gpw_compute_en_comm 22 7.0 0.490 0.509 8.447 8.923 local_gemm 172 8.0 8.188 8.697 8.188 8.697 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.205 8.577 calc_potential_gpw 182 9.5 0.002 0.002 7.958 8.166 collocate_single_gaussian 91 10.0 0.017 0.025 7.789 8.029 mp_sendrecv_dm3 2068 8.0 6.503 6.968 6.503 6.968 integrate_four_center_load 4 10.0 0.000 0.000 6.751 6.755 hfx_load_balance 1 11.0 0.000 0.000 6.751 6.755 mp_sync 37 10.5 4.094 6.446 4.094 6.446 mp2_ri_gpw_compute_en_ener 172 7.0 6.341 6.400 6.341 6.400 pw_gather_s 912 13.2 4.492 4.918 4.492 4.918 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.360761, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1504.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 452.685824E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 592243. MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.007 0.024 53.432 53.434 qs_mol_dyn_low 1 2.0 0.003 0.003 53.193 53.202 qs_forces 11 3.9 0.002 0.003 53.131 53.132 qs_energies 11 4.9 0.001 0.002 51.622 51.638 scf_env_do_scf 11 5.9 0.000 0.001 45.316 45.316 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 43.264 43.264 dbcsr_multiply_generic 2286 12.5 0.094 0.098 33.420 33.852 qs_scf_new_mos 108 7.5 0.000 0.001 33.349 33.647 qs_scf_loop_do_ot 108 8.5 0.000 0.001 33.348 33.646 ot_scf_mini 108 9.5 0.002 0.002 31.657 31.864 multiply_cannon 2286 13.5 0.188 0.196 26.345 27.850 multiply_cannon_loop 2286 14.5 1.811 1.916 25.692 27.207 velocity_verlet 10 3.0 0.001 0.001 26.480 26.480 ot_mini 108 10.5 0.001 0.001 19.059 19.312 qs_ot_get_derivative 108 11.5 0.001 0.001 16.035 16.246 mp_waitall_1 245248 16.5 8.322 14.305 8.322 14.305 multiply_cannon_metrocomm3 54864 15.5 0.073 0.081 6.010 12.516 multiply_cannon_multrec 54864 15.5 3.672 5.797 7.834 11.243 qs_ot_get_p 119 10.4 0.001 0.001 7.967 8.216 rebuild_ks_matrix 119 8.3 0.000 0.000 7.822 7.978 qs_ks_build_kohn_sham_matrix 119 9.3 0.010 0.011 7.821 7.978 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.873 7.009 mp_sum_l 7287 12.8 5.146 6.656 5.146 6.656 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.841 6.324 multiply_cannon_sync_h2d 54864 15.5 5.181 6.207 5.181 6.207 dbcsr_mm_accdrv_process 76910 16.1 1.814 2.850 4.076 5.803 qs_ot_p2m_diag 50 11.0 0.004 0.006 5.168 5.208 init_scf_run 11 5.9 0.000 0.001 5.046 5.046 scf_env_initial_rho_setup 11 6.9 0.005 0.019 5.046 5.046 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 4.932 5.045 sum_up_and_integrate 119 10.3 0.001 0.002 4.437 4.443 integrate_v_rspace 119 11.3 0.002 0.002 4.426 4.433 cp_dbcsr_syevd 50 12.0 0.003 0.003 4.346 4.346 qs_rho_update_rho_low 119 7.7 0.000 0.001 4.056 4.194 calculate_rho_elec 119 8.7 0.011 0.016 4.055 4.194 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.153 4.153 cp_fm_redistribute_end 50 14.0 2.119 4.128 2.126 4.131 cp_fm_diag_elpa_base 50 14.0 2.000 4.025 2.004 4.032 calculate_dm_sparse 119 9.5 0.000 0.000 3.041 3.175 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.929 3.131 apply_single 119 13.6 0.000 0.000 2.929 3.131 jit_kernel_multiply 13 15.8 2.200 2.870 2.200 2.870 acc_transpose_blocks 54864 15.5 0.227 0.259 2.272 2.841 calculate_first_density_matrix 1 7.0 0.000 0.000 2.789 2.800 multiply_cannon_metrocomm1 54864 15.5 0.056 0.063 1.666 2.749 ot_diis_step 108 11.5 0.006 0.006 2.739 2.739 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.409 2.491 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.466 2.468 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.130 2.187 density_rs2pw 119 9.7 0.004 0.004 2.099 2.185 wfi_extrapolate 11 7.9 0.001 0.001 2.181 2.181 grid_integrate_task_list 119 12.3 2.037 2.113 2.037 2.113 init_scf_loop 11 6.9 0.000 0.000 2.036 2.037 mp_sum_d 4135 12.0 1.351 2.032 1.351 2.032 potential_pw2rs 119 12.3 0.004 0.004 1.797 1.808 pw_transfer 1439 11.6 0.051 0.056 1.632 1.700 make_m2s 4572 13.5 0.054 0.056 1.614 1.654 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 1.557 1.627 make_images 4572 14.5 0.133 0.138 1.532 1.572 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.001 1.514 1.548 mp_alltoall_d11v 2130 13.8 1.278 1.481 1.278 1.481 acc_transpose_blocks_sync 164592 16.5 1.193 1.424 1.193 1.424 transfer_rs2pw 487 10.6 0.005 0.006 1.330 1.419 grid_collocate_task_list 119 9.7 1.350 1.419 1.350 1.419 mp_waitany 12084 13.8 1.223 1.397 1.223 1.397 fft3d_ps 1201 14.6 0.369 0.475 1.272 1.339 transfer_pw2rs 487 13.2 0.006 0.006 1.314 1.320 fft_wrap_pw1pw2_140 487 13.2 0.141 0.156 1.212 1.285 dbcsr_dot_sd 1205 11.9 0.048 0.059 0.772 1.198 acc_transpose_blocks_kernels 54864 16.5 0.248 0.383 0.808 1.123 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=53.434000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.272727, yerr=0.962091 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/04/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 57.173320E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3066240 0.0% 0.0% 100.0% average stack size 0.0 0.0 47.9 marketing flops 2.107592E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 488.542208E+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 1753186. 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 0.013 0.030 39.165 39.166 qs_mol_dyn_low 1 2.0 0.003 0.003 38.752 38.759 qs_forces 11 3.9 0.012 0.034 38.697 38.697 qs_energies 11 4.9 0.010 0.016 36.995 37.006 scf_env_do_scf 11 5.9 0.000 0.001 31.883 31.884 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 29.450 29.451 dbcsr_multiply_generic 2286 12.5 0.101 0.103 22.062 22.410 qs_scf_new_mos 108 7.5 0.001 0.001 20.779 21.005 qs_scf_loop_do_ot 108 8.5 0.001 0.001 20.778 21.004 ot_scf_mini 108 9.5 0.002 0.003 19.847 20.011 velocity_verlet 10 3.0 0.001 0.002 18.464 18.465 multiply_cannon 2286 13.5 0.210 0.217 16.898 18.404 multiply_cannon_loop 2286 14.5 1.186 1.255 15.729 17.206 ot_mini 108 10.5 0.001 0.001 12.214 12.445 mp_waitall_1 200699 16.5 5.620 10.734 5.620 10.734 qs_ot_get_derivative 108 11.5 0.001 0.001 9.848 10.016 multiply_cannon_metrocomm3 27432 15.5 0.071 0.074 4.087 9.433 multiply_cannon_multrec 27432 15.5 1.822 4.047 6.125 8.963 rebuild_ks_matrix 119 8.3 0.000 0.000 6.792 6.937 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.014 6.792 6.937 dbcsr_mm_accdrv_process 47894 16.0 3.511 5.658 4.223 6.287 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.971 6.103 qs_ot_get_p 119 10.4 0.001 0.001 4.696 4.923 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.745 4.592 mp_sum_l 7287 12.8 2.067 4.136 2.067 4.136 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.014 4.059 apply_single 119 13.6 0.000 0.000 3.014 4.059 init_scf_run 11 5.9 0.000 0.001 3.861 3.862 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.861 3.861 sum_up_and_integrate 119 10.3 0.001 0.002 3.713 3.718 integrate_v_rspace 119 11.3 0.002 0.008 3.699 3.705 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.661 3.689 calculate_rho_elec 119 8.7 0.021 0.024 3.661 3.689 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.049 3.069 make_m2s 4572 13.5 0.052 0.054 2.648 2.981 make_images 4572 14.5 0.205 0.245 2.558 2.895 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.646 2.646 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.337 2.424 init_scf_loop 11 6.9 0.000 0.000 2.412 2.412 calculate_first_density_matrix 1 7.0 0.000 0.001 2.318 2.320 ot_diis_step 108 11.5 0.011 0.011 2.317 2.318 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.307 2.308 cp_fm_redistribute_end 50 14.0 1.170 2.282 1.174 2.284 cp_fm_diag_elpa_base 50 14.0 1.078 2.190 1.106 2.222 multiply_cannon_sync_h2d 27432 15.5 1.695 2.201 1.695 2.201 calculate_dm_sparse 119 9.5 0.001 0.001 2.075 2.160 density_rs2pw 119 9.7 0.004 0.004 1.958 2.027 grid_integrate_task_list 119 12.3 1.842 1.952 1.842 1.952 acc_transpose_blocks 27432 15.5 0.113 0.118 1.603 1.945 pw_transfer 1439 11.6 0.064 0.068 1.863 1.907 jit_kernel_multiply 10 16.4 0.655 1.894 0.655 1.894 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.882 1.884 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.772 1.819 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.666 1.708 make_images_data 4572 15.5 0.047 0.054 1.241 1.572 prepare_preconditioner 11 7.9 0.000 0.000 1.525 1.549 make_preconditioner 11 8.9 0.000 0.000 1.525 1.549 wfi_extrapolate 11 7.9 0.001 0.001 1.484 1.484 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.426 1.481 potential_pw2rs 119 12.3 0.006 0.007 1.463 1.470 hybrid_alltoall_any 4725 16.4 0.053 0.115 1.107 1.462 fft3d_ps 1201 14.6 0.518 0.575 1.409 1.448 fft_wrap_pw1pw2_140 487 13.2 0.159 0.168 1.395 1.438 mp_alltoall_d11v 2130 13.8 1.223 1.345 1.223 1.345 grid_collocate_task_list 119 9.7 1.289 1.344 1.289 1.344 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.252 1.262 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.206 1.255 transfer_rs2pw 487 10.6 0.005 0.005 1.124 1.219 mp_sum_d 4135 12.0 0.553 1.049 0.553 1.049 mp_allgather_i34 2286 14.5 0.593 0.985 0.593 0.985 qs_energies_init_hamiltonians 11 5.9 0.000 0.002 0.971 0.979 acc_transpose_blocks_sync 82296 16.5 0.812 0.943 0.812 0.943 transfer_pw2rs 487 13.2 0.004 0.005 0.920 0.927 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.879 0.893 acc_transpose_blocks_kernels 27432 16.5 0.187 0.275 0.654 0.866 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=39.166000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.181818, yerr=1.526623 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 519.516160E+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 0.009 0.027 32.074 32.075 qs_mol_dyn_low 1 2.0 0.003 0.003 31.835 31.843 qs_forces 11 3.9 0.010 0.018 31.775 31.775 qs_energies 11 4.9 0.001 0.002 30.224 30.233 scf_env_do_scf 11 5.9 0.000 0.001 25.503 25.503 scf_env_do_scf_inner_loop 108 6.5 0.002 0.006 23.081 23.082 dbcsr_multiply_generic 2286 12.5 0.096 0.098 16.852 16.937 qs_scf_new_mos 108 7.5 0.001 0.001 15.435 15.447 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.434 15.447 velocity_verlet 10 3.0 0.001 0.002 15.086 15.088 ot_scf_mini 108 9.5 0.002 0.002 14.679 14.692 multiply_cannon 2286 13.5 0.195 0.200 13.622 14.298 multiply_cannon_loop 2286 14.5 0.852 0.908 12.843 13.551 ot_mini 108 10.5 0.001 0.001 9.037 9.052 qs_ot_get_derivative 108 11.5 0.001 0.001 7.538 7.551 multiply_cannon_multrec 18288 15.5 1.878 2.831 6.922 7.187 rebuild_ks_matrix 119 8.3 0.000 0.000 5.960 5.977 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.959 5.976 dbcsr_mm_accdrv_process 38222 16.0 4.869 5.712 4.951 5.776 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.246 5.261 mp_waitall_1 158411 16.6 2.641 3.742 2.641 3.742 sum_up_and_integrate 119 10.3 0.001 0.001 3.542 3.547 integrate_v_rspace 119 11.3 0.002 0.003 3.529 3.536 init_scf_run 11 5.9 0.000 0.001 3.451 3.452 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.451 3.451 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.877 3.439 qs_ot_get_p 119 10.4 0.001 0.001 3.390 3.412 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.242 3.247 calculate_rho_elec 119 8.7 0.031 0.032 3.242 3.247 multiply_cannon_metrocomm3 18288 15.5 0.047 0.049 1.466 2.435 init_scf_loop 11 6.9 0.000 0.000 2.405 2.405 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.983 2.352 apply_single 119 13.6 0.000 0.000 1.982 2.352 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.275 2.282 calculate_first_density_matrix 1 7.0 0.000 0.000 2.239 2.240 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.978 1.979 make_m2s 4572 13.5 0.044 0.046 1.821 1.975 density_rs2pw 119 9.7 0.004 0.004 1.807 1.953 grid_integrate_task_list 119 12.3 1.806 1.900 1.806 1.900 make_images 4572 14.5 0.192 0.205 1.736 1.889 calculate_dm_sparse 119 9.5 0.000 0.001 1.840 1.849 pw_transfer 1439 11.6 0.064 0.067 1.810 1.820 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.723 1.735 cp_fm_diag_elpa_base 50 14.0 1.700 1.717 1.721 1.734 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.718 1.729 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.704 1.712 acc_transpose_blocks 18288 15.5 0.079 0.081 1.612 1.696 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.664 1.666 prepare_preconditioner 11 7.9 0.000 0.000 1.653 1.656 make_preconditioner 11 8.9 0.000 0.000 1.653 1.656 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.511 1.594 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.505 1.511 ot_diis_step 108 11.5 0.011 0.011 1.484 1.485 mp_sum_l 7287 12.8 1.016 1.405 1.016 1.405 fft_wrap_pw1pw2_140 487 13.2 0.211 0.217 1.385 1.395 potential_pw2rs 119 12.3 0.007 0.009 1.336 1.340 grid_collocate_task_list 119 9.7 1.234 1.308 1.234 1.308 fft3d_ps 1201 14.6 0.527 0.543 1.287 1.297 multiply_cannon_sync_h2d 18288 15.5 1.055 1.219 1.055 1.219 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.180 1.184 wfi_extrapolate 11 7.9 0.001 0.001 1.159 1.159 transfer_rs2pw 487 10.6 0.005 0.005 0.985 1.156 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.050 1.058 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.949 0.969 make_images_data 4572 15.5 0.047 0.052 0.779 0.936 hybrid_alltoall_any 4725 16.4 0.058 0.117 0.684 0.882 transfer_pw2rs 487 13.2 0.004 0.004 0.817 0.819 acc_transpose_blocks_sync 54864 16.5 0.731 0.812 0.731 0.812 acc_transpose_blocks_kernels 18288 16.5 0.217 0.223 0.783 0.796 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.783 0.785 mp_alltoall_d11v 2130 13.8 0.678 0.784 0.678 0.784 jit_kernel_multiply 4 15.0 0.024 0.771 0.024 0.771 mp_waitany 9880 13.7 0.533 0.733 0.533 0.733 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.675 0.726 transfer_rs2pw_140 130 11.5 0.117 0.125 0.527 0.700 cp_fm_cholesky_invert 11 10.9 0.693 0.697 0.693 0.697 mp_alltoall_z22v 1201 16.6 0.610 0.679 0.610 0.679 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.622 0.678 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=32.075000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=495.727273, yerr=0.445362 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 559.493120E+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 0.020 0.033 37.543 37.545 qs_mol_dyn_low 1 2.0 0.003 0.004 37.372 37.380 qs_forces 11 3.9 0.002 0.003 37.309 37.310 qs_energies 11 4.9 0.001 0.001 35.582 35.587 scf_env_do_scf 11 5.9 0.000 0.001 30.284 30.286 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.631 26.632 dbcsr_multiply_generic 2286 12.5 0.100 0.103 20.232 20.340 velocity_verlet 10 3.0 0.001 0.002 18.919 18.921 qs_scf_new_mos 108 7.5 0.001 0.001 18.427 18.481 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.427 18.481 ot_scf_mini 108 9.5 0.002 0.003 17.386 17.451 multiply_cannon 2286 13.5 0.219 0.226 16.307 16.659 multiply_cannon_loop 2286 14.5 1.517 1.571 15.305 15.641 ot_mini 108 10.5 0.001 0.001 10.594 10.672 multiply_cannon_multrec 27432 15.5 2.457 3.118 9.084 9.379 qs_ot_get_derivative 108 11.5 0.001 0.001 8.700 8.765 dbcsr_mm_accdrv_process 47916 15.9 6.174 7.810 6.526 7.898 rebuild_ks_matrix 119 8.3 0.000 0.000 6.538 6.589 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.538 6.588 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.804 5.849 qs_ot_get_p 119 10.4 0.001 0.002 3.915 3.997 init_scf_run 11 5.9 0.000 0.001 3.936 3.937 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.936 3.936 sum_up_and_integrate 119 10.3 0.001 0.001 3.595 3.645 integrate_v_rspace 119 11.3 0.002 0.003 3.583 3.633 init_scf_loop 11 6.9 0.000 0.000 3.625 3.625 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.205 3.608 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.434 3.467 calculate_rho_elec 119 8.7 0.040 0.046 3.433 3.467 prepare_preconditioner 11 7.9 0.000 0.000 2.650 2.660 make_preconditioner 11 8.9 0.000 0.000 2.650 2.660 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.262 2.584 mp_waitall_1 137007 16.6 1.973 2.570 1.973 2.570 make_m2s 4572 13.5 0.054 0.056 2.398 2.514 calculate_first_density_matrix 1 7.0 0.000 0.000 2.476 2.477 acc_transpose_blocks 27432 15.5 0.118 0.123 2.365 2.444 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.403 2.413 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.372 2.406 make_images 4572 14.5 0.274 0.333 2.290 2.404 calculate_dm_sparse 119 9.5 0.000 0.000 2.258 2.312 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.122 2.257 apply_single 119 13.6 0.000 0.000 2.122 2.256 pw_transfer 1439 11.6 0.064 0.069 2.011 2.092 density_rs2pw 119 9.7 0.004 0.004 1.892 2.033 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.919 2.003 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.993 1.993 grid_integrate_task_list 119 12.3 1.828 1.930 1.828 1.930 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.891 1.892 ot_diis_step 108 11.5 0.012 0.012 1.848 1.849 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.708 1.718 cp_fm_diag_elpa_base 50 14.0 1.673 1.691 1.706 1.715 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.680 1.693 fft_wrap_pw1pw2_140 487 13.2 0.245 0.258 1.548 1.633 fft3d_ps 1201 14.6 0.554 0.609 1.439 1.515 acc_transpose_blocks_sync 82296 16.5 1.396 1.485 1.396 1.485 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.813 1.473 potential_pw2rs 119 12.3 0.009 0.009 1.389 1.443 wfi_extrapolate 11 7.9 0.001 0.001 1.406 1.406 mp_sum_l 7287 12.8 1.095 1.365 1.095 1.365 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.353 1.362 grid_collocate_task_list 119 9.7 1.250 1.349 1.250 1.349 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.316 1.334 cp_fm_upper_to_full 72 14.2 0.802 1.140 0.802 1.140 dbcsr_complete_redistribute 329 12.2 0.126 0.157 0.841 1.108 transfer_rs2pw 487 10.6 0.004 0.005 0.961 1.103 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.084 1.085 jit_kernel_multiply 6 16.5 0.288 1.067 0.288 1.067 make_images_data 4572 15.5 0.047 0.051 0.878 0.992 hybrid_alltoall_any 4725 16.4 0.065 0.156 0.776 0.934 mp_alltoall_d11v 2130 13.8 0.812 0.919 0.812 0.919 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.801 0.876 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.602 0.863 acc_transpose_blocks_kernels 27432 16.5 0.270 0.276 0.823 0.842 mp_alltoall_z22v 1201 16.6 0.767 0.833 0.767 0.833 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.825 0.829 cp_fm_cholesky_invert 11 10.9 0.825 0.829 0.825 0.829 transfer_pw2rs 487 13.2 0.004 0.004 0.795 0.799 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.545000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=530.000000, yerr=4.472136 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 620.265472E+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 0.009 0.028 29.233 29.234 qs_mol_dyn_low 1 2.0 0.003 0.003 29.039 29.046 qs_forces 11 3.9 0.002 0.003 28.981 28.982 qs_energies 11 4.9 0.001 0.001 27.281 27.283 scf_env_do_scf 11 5.9 0.000 0.001 22.001 22.001 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.539 19.539 velocity_verlet 10 3.0 0.002 0.002 14.538 14.540 dbcsr_multiply_generic 2286 12.5 0.094 0.099 13.619 13.687 qs_scf_new_mos 108 7.5 0.001 0.001 12.021 12.043 qs_scf_loop_do_ot 108 8.5 0.001 0.001 12.021 12.042 multiply_cannon 2286 13.5 0.223 0.229 11.043 11.484 ot_scf_mini 108 9.5 0.002 0.002 11.317 11.342 multiply_cannon_loop 2286 14.5 0.641 0.662 10.163 10.346 multiply_cannon_multrec 9144 15.5 1.720 1.910 6.412 6.818 ot_mini 108 10.5 0.001 0.001 6.688 6.720 rebuild_ks_matrix 119 8.3 0.000 0.000 5.730 5.749 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.014 5.729 5.748 qs_ot_get_derivative 108 11.5 0.001 0.001 5.405 5.431 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.102 5.119 dbcsr_mm_accdrv_process 12550 15.8 4.032 4.801 4.583 4.844 init_scf_run 11 5.9 0.000 0.001 3.854 3.854 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.853 3.853 sum_up_and_integrate 119 10.3 0.001 0.001 3.381 3.386 integrate_v_rspace 119 11.3 0.003 0.003 3.371 3.376 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.282 3.290 calculate_rho_elec 119 8.7 0.060 0.061 3.282 3.290 qs_ot_get_p 119 10.4 0.002 0.002 2.770 2.799 calculate_first_density_matrix 1 7.0 0.000 0.000 2.729 2.730 init_scf_loop 11 6.9 0.000 0.000 2.443 2.444 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.059 2.129 mp_waitall_1 115863 16.7 1.488 2.074 1.488 2.074 calculate_dm_sparse 119 9.5 0.000 0.000 1.966 1.981 pw_transfer 1439 11.6 0.065 0.067 1.944 1.953 grid_integrate_task_list 119 12.3 1.867 1.944 1.867 1.944 make_m2s 4572 13.5 0.034 0.035 1.743 1.866 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.851 1.862 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.807 1.809 density_rs2pw 119 9.7 0.003 0.004 1.674 1.796 make_images 4572 14.5 0.269 0.301 1.654 1.776 prepare_preconditioner 11 7.9 0.000 0.000 1.683 1.687 make_preconditioner 11 8.9 0.000 0.000 1.683 1.687 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.616 1.618 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.575 1.599 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.263 1.587 apply_single 119 13.6 0.000 0.000 1.263 1.587 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.568 1.568 fft_wrap_pw1pw2_140 487 13.2 0.323 0.332 1.533 1.544 acc_transpose_blocks 9144 15.5 0.041 0.042 1.470 1.503 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.457 1.464 grid_collocate_task_list 119 9.7 1.295 1.412 1.295 1.412 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.359 1.364 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.316 1.329 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.293 1.301 cp_fm_diag_elpa_base 50 14.0 1.266 1.283 1.291 1.300 jit_kernel_multiply 6 15.3 0.512 1.274 0.512 1.274 fft3d_ps 1201 14.6 0.559 0.570 1.265 1.273 ot_diis_step 108 11.5 0.013 0.013 1.271 1.271 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.214 1.214 potential_pw2rs 119 12.3 0.010 0.011 1.196 1.198 wfi_extrapolate 11 7.9 0.001 0.001 1.073 1.073 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.400 1.050 hybrid_alltoall_any 4725 16.4 0.064 0.175 0.754 0.935 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.880 0.934 make_images_data 4572 15.5 0.042 0.045 0.740 0.886 transfer_rs2pw 487 10.6 0.004 0.005 0.762 0.872 mp_alltoall_d11v 2130 13.8 0.770 0.853 0.770 0.853 cp_fm_cholesky_invert 11 10.9 0.848 0.850 0.848 0.850 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.809 0.816 acc_transpose_blocks_sync 27432 16.5 0.720 0.752 0.720 0.752 multiply_cannon_metrocomm1 9144 15.5 0.023 0.024 0.390 0.752 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.748 0.750 qs_env_update_s_mstruct 11 6.9 0.000 0.001 0.663 0.710 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.694 0.698 mp_sum_l 7287 12.8 0.383 0.668 0.383 0.668 parallel_gemm_fm 81 9.0 0.000 0.000 0.645 0.646 parallel_gemm_fm_cosma 81 10.0 0.645 0.646 0.645 0.646 transfer_pw2rs 487 13.2 0.003 0.004 0.633 0.635 mp_alltoall_z22v 1201 16.6 0.580 0.612 0.580 0.612 make_basis_sm 11 9.8 0.001 0.001 0.610 0.611 mp_allgather_i34 2286 14.5 0.200 0.601 0.200 0.601 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.234000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=588.909091, yerr=5.759649 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 772.210688E+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 0.011 0.030 43.017 43.019 qs_mol_dyn_low 1 2.0 0.003 0.004 42.805 42.812 qs_forces 11 3.9 0.002 0.003 42.604 42.605 qs_energies 11 4.9 0.001 0.001 40.608 40.612 scf_env_do_scf 11 5.9 0.001 0.001 34.569 34.569 scf_env_do_scf_inner_loop 108 6.5 0.008 0.016 26.793 26.794 velocity_verlet 10 3.0 0.002 0.002 23.824 23.847 dbcsr_multiply_generic 2286 12.5 0.102 0.105 19.238 19.486 qs_scf_new_mos 108 7.5 0.001 0.001 17.489 17.592 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.489 17.591 ot_scf_mini 108 9.5 0.002 0.002 16.326 16.429 multiply_cannon 2286 13.5 0.297 0.308 15.184 16.313 multiply_cannon_loop 2286 14.5 0.856 0.871 13.941 15.102 ot_mini 108 10.5 0.001 0.001 10.002 10.120 multiply_cannon_multrec 9144 15.5 3.436 4.820 8.852 8.936 qs_ot_get_derivative 108 11.5 0.001 0.001 7.969 8.075 init_scf_loop 11 6.9 0.000 0.000 7.748 7.750 rebuild_ks_matrix 119 8.3 0.000 0.000 7.116 7.269 qs_ks_build_kohn_sham_matrix 119 9.3 0.036 0.049 7.116 7.268 prepare_preconditioner 11 7.9 0.000 0.000 6.724 6.752 make_preconditioner 11 8.9 0.000 0.000 6.724 6.752 dbcsr_mm_accdrv_process 12550 15.8 4.462 6.610 5.281 6.642 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.301 6.630 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.417 6.556 cp_fm_upper_to_full 72 14.2 3.140 4.506 3.140 4.506 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.983 3.993 calculate_rho_elec 119 8.7 0.118 0.121 3.982 3.993 init_scf_run 11 5.9 0.000 0.001 3.970 3.970 scf_env_initial_rho_setup 11 6.9 0.001 0.001 3.969 3.969 sum_up_and_integrate 119 10.3 0.001 0.001 3.746 3.752 integrate_v_rspace 119 11.3 0.003 0.003 3.735 3.741 qs_ot_get_p 119 10.4 0.002 0.002 3.418 3.554 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.937 3.523 mp_waitall_1 94719 16.7 2.287 3.521 2.287 3.521 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.360 2.809 apply_single 119 13.6 0.000 0.000 2.359 2.809 dbcsr_complete_redistribute 329 12.2 0.285 0.291 1.960 2.763 pw_transfer 1439 11.6 0.068 0.069 2.581 2.586 calculate_first_density_matrix 1 7.0 0.000 0.000 2.533 2.536 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.358 2.523 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.483 2.489 make_m2s 4572 13.5 0.038 0.038 2.321 2.463 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.660 2.463 calculate_dm_sparse 119 9.5 0.000 0.001 2.349 2.365 make_images 4572 14.5 0.354 0.387 2.201 2.343 mp_alltoall_i22 627 13.8 1.392 2.235 1.392 2.235 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.418 2.214 acc_transpose_blocks 9144 15.5 0.043 0.044 1.887 2.113 fft_wrap_pw1pw2_140 487 13.2 0.575 0.580 2.099 2.104 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.051 2.102 grid_integrate_task_list 119 12.3 2.070 2.096 2.070 2.096 mp_sum_l 7287 12.8 1.182 2.088 1.182 2.088 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.032 2.034 density_rs2pw 119 9.7 0.003 0.004 2.024 2.034 ot_diis_step 108 11.5 0.014 0.014 2.008 2.008 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.837 1.838 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 1.781 1.782 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.675 1.676 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.595 1.638 fft3d_ps 1201 14.6 0.594 0.607 1.580 1.583 grid_collocate_task_list 119 9.7 1.515 1.527 1.515 1.527 cp_fm_cholesky_invert 11 10.9 1.411 1.414 1.411 1.414 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.385 1.385 cp_fm_diag_elpa_base 50 14.0 1.240 1.293 1.383 1.383 wfi_extrapolate 11 7.9 0.001 0.001 1.369 1.369 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.354 1.361 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.304 1.321 potential_pw2rs 119 12.3 0.014 0.014 1.272 1.273 hybrid_alltoall_any 4725 16.4 0.090 0.150 1.039 1.255 mp_alltoall_d11v 2130 13.8 1.138 1.165 1.138 1.165 make_images_data 4572 15.5 0.046 0.049 0.971 1.154 qs_env_update_s_mstruct 11 6.9 0.001 0.001 1.107 1.126 acc_transpose_blocks_sync 27432 16.5 1.091 1.122 1.091 1.122 jit_kernel_multiply 6 15.7 0.791 1.055 0.791 1.055 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.001 1.030 acc_transpose_blocks_kernels 9144 16.5 0.121 0.122 0.738 0.966 qs_create_task_list 11 7.9 0.000 0.000 0.931 0.942 generate_qs_task_list 11 8.9 0.369 0.388 0.930 0.941 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.910 0.924 mp_alltoall_z22v 1201 16.6 0.850 0.867 0.850 0.867 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.019000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=720.363636, yerr=16.366666 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 499.728384E+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 0.020 0.070 80.680 80.688 qs_mol_dyn_low 1 2.0 0.003 0.003 80.302 80.321 qs_forces 11 3.9 0.003 0.003 80.234 80.234 qs_energies 11 4.9 0.001 0.002 77.367 77.384 scf_env_do_scf 11 5.9 0.000 0.001 67.777 67.780 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 62.267 62.267 dbcsr_multiply_generic 2055 12.4 0.105 0.107 49.563 49.849 qs_scf_new_mos 99 7.5 0.000 0.001 45.189 45.306 qs_scf_loop_do_ot 99 8.5 0.001 0.001 45.188 45.306 ot_scf_mini 99 9.5 0.002 0.002 42.887 42.979 multiply_cannon 2055 13.4 0.185 0.189 41.948 42.951 multiply_cannon_loop 2055 14.4 1.756 1.788 40.979 41.955 velocity_verlet 10 3.0 0.001 0.002 41.516 41.517 ot_mini 99 10.5 0.001 0.001 25.177 25.288 qs_ot_get_derivative 99 11.5 0.001 0.001 18.418 18.530 multiply_cannon_multrec 49320 15.4 11.735 12.463 17.766 18.447 rebuild_ks_matrix 110 8.3 0.000 0.001 14.575 14.750 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.012 14.575 14.750 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.784 12.939 mp_waitall_1 220248 16.4 9.537 10.732 9.537 10.732 multiply_cannon_sync_h2d 49320 15.4 9.957 10.588 9.957 10.588 qs_ot_get_p 110 10.4 0.003 0.006 9.483 9.600 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.412 7.904 apply_single 110 13.6 0.000 0.000 7.412 7.904 init_scf_run 11 5.9 0.000 0.001 7.508 7.508 scf_env_initial_rho_setup 11 6.9 0.001 0.001 7.508 7.508 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 6.625 7.325 multiply_cannon_metrocomm3 49320 15.4 0.085 0.088 5.745 7.259 sum_up_and_integrate 110 10.3 0.002 0.003 7.036 7.052 integrate_v_rspace 110 11.3 0.003 0.003 7.011 7.033 ot_diis_step 99 11.5 0.006 0.006 6.604 6.604 qs_ot_p2m_diag 48 11.0 0.012 0.019 6.390 6.406 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.309 6.402 calculate_rho_elec 110 8.6 0.021 0.025 6.308 6.401 dbcsr_mm_accdrv_process 87628 16.1 2.887 3.007 5.899 6.325 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.612 5.612 init_scf_loop 11 6.9 0.000 0.000 5.480 5.480 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.198 5.243 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.041 5.064 cp_fm_diag_elpa_base 48 14.0 5.028 5.050 5.039 5.063 mp_sum_l 6594 12.7 3.231 4.120 3.231 4.120 wfi_extrapolate 11 7.9 0.001 0.001 4.097 4.097 make_m2s 4110 13.4 0.061 0.066 3.983 4.090 make_images 4110 14.4 0.177 0.189 3.886 3.998 calculate_dm_sparse 110 9.5 0.001 0.001 3.847 3.965 density_rs2pw 110 9.6 0.004 0.005 3.403 3.617 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.582 3.587 grid_integrate_task_list 110 12.3 3.251 3.423 3.251 3.423 prepare_preconditioner 11 7.9 0.000 0.000 3.302 3.320 make_preconditioner 11 8.9 0.000 0.000 3.302 3.320 calculate_first_density_matrix 1 7.0 0.000 0.000 3.314 3.315 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.244 3.305 pw_transfer 1331 11.6 0.055 0.069 3.147 3.230 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.103 3.159 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.059 3.143 multiply_cannon_metrocomm1 49320 15.4 0.066 0.069 1.933 3.062 jit_kernel_multiply 13 15.9 2.740 2.973 2.740 2.973 qs_ot_get_orbitals 99 10.5 0.000 0.001 2.905 2.944 fft_wrap_pw1pw2_140 451 13.1 0.389 0.429 2.590 2.676 potential_pw2rs 110 12.3 0.005 0.006 2.620 2.640 mp_alltoall_d11v 2046 13.8 2.029 2.566 2.029 2.566 acc_transpose_blocks 49320 15.4 0.205 0.213 2.493 2.566 fft3d_ps 1111 14.6 0.795 0.882 2.355 2.419 grid_collocate_task_list 110 9.6 2.159 2.297 2.159 2.297 transfer_rs2pw 451 10.6 0.005 0.006 1.916 2.119 mp_waitany 14300 13.8 1.815 2.011 1.815 2.011 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.921 1.945 make_images_data 4110 15.4 0.043 0.046 1.779 1.925 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.829 1.863 hybrid_alltoall_any 4261 16.3 0.084 0.481 1.550 1.840 parallel_gemm_fm 81 9.0 0.000 0.000 1.833 1.839 parallel_gemm_fm_cosma 81 10.0 1.833 1.838 1.833 1.838 cp_fm_cholesky_invert 11 10.9 1.812 1.816 1.812 1.816 transfer_pw2rs 451 13.1 0.006 0.007 1.762 1.776 mp_sum_d 3889 11.9 1.274 1.775 1.274 1.775 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=80.688000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.090909, yerr=1.880742 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/10/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 390.715586E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 5019072 0.0% 0.0% 100.0% average stack size 0.0 0.0 196.1 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 592.756736E+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 2716832. 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 0.011 0.025 69.556 69.557 qs_mol_dyn_low 1 2.0 0.003 0.004 69.281 69.327 qs_forces 11 3.9 0.003 0.003 69.215 69.216 qs_energies 11 4.9 0.002 0.003 65.877 65.880 scf_env_do_scf 11 5.9 0.000 0.001 57.330 57.333 scf_env_do_scf_inner_loop 99 6.5 0.002 0.007 49.616 49.617 dbcsr_multiply_generic 2055 12.4 0.116 0.120 38.130 38.288 velocity_verlet 10 3.0 0.001 0.002 36.438 36.439 qs_scf_new_mos 99 7.5 0.001 0.001 33.598 33.738 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.597 33.737 multiply_cannon 2055 13.4 0.224 0.243 31.386 32.616 ot_scf_mini 99 9.5 0.003 0.003 31.924 32.036 multiply_cannon_loop 2055 14.4 1.162 1.191 30.085 30.991 ot_mini 99 10.5 0.001 0.001 18.538 18.666 multiply_cannon_multrec 24660 15.4 7.041 8.563 14.073 15.746 rebuild_ks_matrix 110 8.3 0.000 0.000 13.628 13.736 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.627 13.736 qs_ot_get_derivative 99 11.5 0.001 0.001 12.743 12.863 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.991 12.087 mp_waitall_1 176588 16.5 7.529 10.219 7.529 10.219 multiply_cannon_metrocomm3 24660 15.4 0.072 0.074 5.087 7.781 init_scf_loop 11 6.9 0.000 0.000 7.675 7.676 multiply_cannon_sync_h2d 24660 15.4 6.457 7.463 6.457 7.463 dbcsr_mm_accdrv_process 52282 16.1 5.505 6.433 6.862 7.211 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.504 7.165 apply_single 110 13.6 0.000 0.001 6.504 7.165 qs_ot_get_p 110 10.4 0.001 0.001 6.692 6.847 sum_up_and_integrate 110 10.3 0.001 0.003 6.436 6.450 integrate_v_rspace 110 11.3 0.002 0.003 6.409 6.423 init_scf_run 11 5.9 0.000 0.001 6.160 6.161 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.160 6.160 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.988 5.997 calculate_rho_elec 110 8.6 0.039 0.047 5.987 5.997 ot_diis_step 99 11.5 0.010 0.010 5.745 5.745 prepare_preconditioner 11 7.9 0.000 0.000 5.647 5.665 make_preconditioner 11 8.9 0.000 0.000 5.647 5.665 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.783 5.494 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.223 5.380 make_m2s 4110 13.4 0.057 0.060 4.335 4.802 qs_ot_p2m_diag 48 11.0 0.029 0.044 4.756 4.776 make_images 4110 14.4 0.408 0.463 4.225 4.689 cp_dbcsr_syevd 48 12.0 0.003 0.003 4.282 4.282 pw_transfer 1331 11.6 0.066 0.073 3.628 3.773 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.688 3.701 cp_fm_diag_elpa_base 48 14.0 3.638 3.660 3.685 3.697 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 3.522 3.668 wfi_extrapolate 11 7.9 0.001 0.001 3.549 3.549 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.327 3.398 density_rs2pw 110 9.6 0.004 0.005 3.223 3.391 grid_integrate_task_list 110 12.3 3.160 3.312 3.160 3.312 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 3.271 3.273 fft_wrap_pw1pw2_140 451 13.1 0.459 0.474 2.976 3.123 calculate_dm_sparse 110 9.5 0.001 0.001 2.973 3.005 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.923 2.969 hybrid_alltoall_any 4261 16.3 0.105 0.450 2.032 2.844 make_images_data 4110 15.4 0.048 0.052 2.325 2.834 fft3d_ps 1111 14.6 1.110 1.330 2.619 2.775 cp_fm_cholesky_invert 11 10.9 2.656 2.663 2.656 2.663 calculate_first_density_matrix 1 7.0 0.000 0.000 2.512 2.514 mp_sum_l 6594 12.7 1.778 2.389 1.778 2.389 grid_collocate_task_list 110 9.6 2.171 2.334 2.171 2.334 potential_pw2rs 110 12.3 0.008 0.009 2.289 2.304 mp_alltoall_d11v 2046 13.8 1.833 2.147 1.833 2.147 acc_transpose_blocks 24660 15.4 0.115 0.119 1.971 2.003 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.973 1.992 jit_kernel_multiply 10 16.3 0.994 1.927 0.994 1.927 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.859 1.860 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.793 1.808 transfer_rs2pw 451 10.6 0.006 0.008 1.470 1.703 mp_allgather_i34 2055 14.4 0.612 1.645 0.612 1.645 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.540 1.550 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.394 1.496 multiply_cannon_metrocomm4 22605 15.4 0.078 0.083 0.755 1.440 dbcsr_complete_redistribute 325 12.2 0.236 0.291 1.163 1.431 mp_waitany 10164 13.8 1.160 1.428 1.160 1.428 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=69.557000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=560.454545, yerr=7.536139 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 665.321472E+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 0.049 0.066 60.623 60.631 qs_mol_dyn_low 1 2.0 0.003 0.003 60.263 60.272 qs_forces 11 3.9 0.003 0.003 60.197 60.198 qs_energies 11 4.9 0.002 0.003 57.000 57.003 scf_env_do_scf 11 5.9 0.000 0.001 49.088 49.088 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 40.347 40.347 velocity_verlet 10 3.0 0.002 0.002 32.928 32.930 dbcsr_multiply_generic 2055 12.4 0.113 0.124 28.864 29.080 qs_scf_new_mos 99 7.5 0.001 0.001 25.566 25.669 qs_scf_loop_do_ot 99 8.5 0.001 0.001 25.565 25.668 ot_scf_mini 99 9.5 0.002 0.002 24.316 24.442 multiply_cannon 2055 13.4 0.213 0.222 22.477 23.684 multiply_cannon_loop 2055 14.4 0.812 0.837 21.270 22.408 ot_mini 99 10.5 0.001 0.001 13.833 13.965 rebuild_ks_matrix 110 8.3 0.000 0.000 12.249 12.371 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 12.248 12.371 multiply_cannon_multrec 16440 15.4 3.710 5.001 9.915 11.140 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.773 10.889 mp_waitall_1 139946 16.5 6.863 10.084 6.863 10.084 qs_ot_get_derivative 99 11.5 0.001 0.001 9.425 9.552 init_scf_loop 11 6.9 0.000 0.000 8.703 8.704 multiply_cannon_metrocomm3 16440 15.4 0.046 0.047 4.221 7.164 prepare_preconditioner 11 7.9 0.000 0.000 6.932 6.949 make_preconditioner 11 8.9 0.000 0.000 6.932 6.949 make_full_inverse_cholesky 11 9.9 0.000 0.000 6.243 6.601 sum_up_and_integrate 110 10.3 0.001 0.002 6.365 6.381 integrate_v_rspace 110 11.3 0.003 0.003 6.339 6.355 dbcsr_mm_accdrv_process 34862 16.1 5.334 5.665 6.050 6.209 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.818 5.828 calculate_rho_elec 110 8.6 0.059 0.059 5.817 5.828 qs_ot_get_p 110 10.4 0.001 0.002 5.617 5.758 init_scf_run 11 5.9 0.000 0.001 5.513 5.513 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.512 5.512 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.912 5.379 apply_single 110 13.6 0.000 0.000 4.912 5.379 make_m2s 4110 13.4 0.050 0.051 4.278 4.680 make_images 4110 14.4 0.400 0.522 4.163 4.564 ot_diis_step 99 11.5 0.011 0.011 4.379 4.379 multiply_cannon_sync_h2d 16440 15.4 3.247 3.987 3.247 3.987 qs_ot_p2m_diag 48 11.0 0.042 0.044 3.944 3.947 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.221 3.841 pw_transfer 1331 11.6 0.065 0.073 3.635 3.649 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.564 3.564 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.528 3.546 grid_integrate_task_list 110 12.3 3.212 3.335 3.212 3.335 density_rs2pw 110 9.6 0.004 0.005 3.043 3.190 fft_wrap_pw1pw2_140 451 13.1 0.578 0.588 3.020 3.037 cp_fm_diag_elpa 48 13.0 0.000 0.000 3.009 3.020 cp_fm_diag_elpa_base 48 14.0 2.942 2.974 3.007 3.018 wfi_extrapolate 11 7.9 0.001 0.001 3.000 3.000 make_images_data 4110 15.4 0.045 0.049 2.361 2.896 hybrid_alltoall_any 4261 16.3 0.108 0.378 2.128 2.876 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.805 2.807 cp_fm_cholesky_invert 11 10.9 2.689 2.696 2.689 2.696 calculate_dm_sparse 110 9.5 0.001 0.001 2.527 2.557 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.467 2.536 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.449 2.492 fft3d_ps 1111 14.6 1.089 1.100 2.472 2.490 multiply_cannon_metrocomm4 14385 15.4 0.048 0.053 0.886 2.458 calculate_first_density_matrix 1 7.0 0.000 0.000 2.414 2.416 grid_collocate_task_list 110 9.6 2.220 2.392 2.220 2.392 mp_irecv_dv 48980 15.7 0.812 2.323 0.812 2.323 potential_pw2rs 110 12.3 0.011 0.012 2.190 2.203 mp_sum_l 6594 12.7 1.487 2.093 1.487 2.093 dbcsr_complete_redistribute 325 12.2 0.345 0.394 1.543 1.999 mp_alltoall_d11v 2046 13.8 1.870 1.989 1.870 1.989 qs_energies_init_hamiltonians 11 5.9 0.000 0.001 1.952 1.953 cp_fm_upper_to_full 70 14.2 1.411 1.791 1.411 1.791 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.669 1.684 cp_fm_cholesky_decompose 22 10.9 1.620 1.645 1.620 1.645 acc_transpose_blocks 16440 15.4 0.076 0.078 1.573 1.638 mp_allgather_i34 2055 14.4 0.522 1.580 0.522 1.580 transfer_rs2pw 451 10.6 0.005 0.006 1.315 1.503 copy_fm_to_dbcsr 174 11.2 0.001 0.001 1.036 1.487 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.468 1.482 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.378 1.471 mp_waitany 17072 13.8 1.128 1.351 1.128 1.351 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.281 1.286 transfer_pw2rs 451 13.1 0.005 0.005 1.214 1.225 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=60.631000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=630.818182, yerr=8.277222 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 733.401088E+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 0.027 0.066 64.891 64.892 qs_mol_dyn_low 1 2.0 0.003 0.004 64.639 64.648 qs_forces 11 3.9 0.003 0.003 64.577 64.577 qs_energies 11 4.9 0.001 0.002 61.192 61.196 scf_env_do_scf 11 5.9 0.000 0.001 52.961 52.963 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 41.492 41.493 velocity_verlet 10 3.0 0.001 0.002 36.866 36.868 dbcsr_multiply_generic 2055 12.4 0.114 0.121 30.471 30.686 qs_scf_new_mos 99 7.5 0.001 0.001 27.044 27.157 qs_scf_loop_do_ot 99 8.5 0.001 0.001 27.043 27.157 ot_scf_mini 99 9.5 0.002 0.003 25.368 25.487 multiply_cannon 2055 13.4 0.240 0.259 23.595 24.849 multiply_cannon_loop 2055 14.4 1.409 1.482 22.053 22.695 ot_mini 99 10.5 0.001 0.001 14.649 14.798 multiply_cannon_multrec 24660 15.4 4.097 6.681 12.983 14.123 rebuild_ks_matrix 110 8.3 0.000 0.000 11.898 12.036 qs_ks_build_kohn_sham_matrix 110 9.3 0.016 0.027 11.897 12.036 init_scf_loop 11 6.9 0.000 0.000 11.429 11.430 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.507 10.629 qs_ot_get_derivative 99 11.5 0.001 0.001 10.490 10.612 prepare_preconditioner 11 7.9 0.000 0.000 9.701 9.720 make_preconditioner 11 8.9 0.000 0.000 9.701 9.720 dbcsr_mm_accdrv_process 52304 16.0 7.684 8.913 8.730 9.684 make_full_inverse_cholesky 11 9.9 0.000 0.000 7.944 9.383 sum_up_and_integrate 110 10.3 0.001 0.002 6.145 6.157 integrate_v_rspace 110 11.3 0.003 0.003 6.119 6.132 mp_waitall_1 121746 16.5 4.017 6.098 4.017 6.098 qs_ot_get_p 110 10.4 0.001 0.001 5.602 5.783 make_m2s 4110 13.4 0.060 0.061 5.253 5.757 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.727 5.740 calculate_rho_elec 110 8.6 0.078 0.081 5.727 5.740 make_images 4110 14.4 0.578 0.703 5.112 5.611 init_scf_run 11 5.9 0.000 0.001 5.488 5.488 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.488 5.488 cp_fm_upper_to_full 70 14.2 3.330 4.826 3.330 4.826 ot_diis_step 99 11.5 0.011 0.011 4.120 4.120 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.029 4.119 apply_single 110 13.6 0.000 0.000 4.029 4.119 qs_ot_p2m_diag 48 11.0 0.055 0.064 3.786 3.803 dbcsr_complete_redistribute 325 12.2 0.437 0.505 2.646 3.764 pw_transfer 1331 11.6 0.065 0.073 3.587 3.618 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.467 3.528 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.481 3.516 grid_integrate_task_list 110 12.3 3.272 3.429 3.272 3.429 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.319 3.378 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.137 3.243 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.228 3.229 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.351 3.191 hybrid_alltoall_any 4261 16.3 0.122 0.460 2.227 3.090 fft_wrap_pw1pw2_140 451 13.1 0.609 0.629 3.008 3.046 make_images_data 4110 15.4 0.048 0.053 2.573 3.039 calculate_dm_sparse 110 9.5 0.001 0.001 2.969 2.999 density_rs2pw 110 9.6 0.004 0.004 2.853 2.988 wfi_extrapolate 11 7.9 0.001 0.001 2.968 2.968 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.749 2.841 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.821 2.823 mp_alltoall_i22 605 13.7 1.647 2.794 1.647 2.794 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.701 2.711 cp_fm_diag_elpa_base 48 14.0 2.557 2.615 2.698 2.709 cp_fm_cholesky_invert 11 10.9 2.563 2.571 2.563 2.571 multiply_cannon_sync_h2d 24660 15.4 2.360 2.528 2.360 2.528 acc_transpose_blocks 24660 15.4 0.113 0.115 2.402 2.518 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.437 2.480 grid_collocate_task_list 110 9.6 2.262 2.466 2.262 2.466 calculate_first_density_matrix 1 7.0 0.000 0.000 2.426 2.429 fft3d_ps 1111 14.6 1.085 1.112 2.364 2.381 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.255 2.255 mp_alltoall_d11v 2046 13.8 1.739 1.991 1.739 1.991 potential_pw2rs 110 12.3 0.012 0.013 1.968 1.972 qs_ot_get_orbitals 99 10.5 0.001 0.001 1.779 1.820 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.617 1.713 cp_fm_cholesky_decompose 22 10.9 1.624 1.676 1.624 1.676 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.630 1.645 mp_allgather_i34 2055 14.4 0.622 1.617 0.622 1.617 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.534 1.551 mp_sum_l 6594 12.7 0.932 1.532 0.932 1.532 multiply_cannon_metrocomm4 20550 15.4 0.061 0.065 0.870 1.521 acc_transpose_blocks_sync 73980 16.4 1.397 1.504 1.397 1.504 mp_irecv_dv 62702 16.1 0.765 1.434 0.765 1.434 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=64.892000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=695.272727, yerr=8.550289 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 864.079872E+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 0.031 0.073 54.286 54.303 qs_mol_dyn_low 1 2.0 0.003 0.003 54.002 54.026 qs_forces 11 3.9 0.003 0.003 53.899 53.899 qs_energies 11 4.9 0.001 0.001 50.278 50.282 scf_env_do_scf 11 5.9 0.000 0.001 42.063 42.063 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 34.507 34.507 velocity_verlet 10 3.0 0.001 0.002 30.709 30.715 dbcsr_multiply_generic 2055 12.4 0.105 0.109 22.950 23.078 qs_scf_new_mos 99 7.5 0.001 0.001 20.199 20.249 qs_scf_loop_do_ot 99 8.5 0.001 0.001 20.199 20.249 ot_scf_mini 99 9.5 0.002 0.002 18.952 18.980 multiply_cannon 2055 13.4 0.239 0.252 17.516 18.892 multiply_cannon_loop 2055 14.4 0.602 0.625 16.260 16.562 rebuild_ks_matrix 110 8.3 0.000 0.000 11.462 11.483 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.014 11.462 11.483 ot_mini 99 10.5 0.001 0.001 10.507 10.530 qs_ks_update_qs_env 110 7.6 0.001 0.001 10.196 10.215 multiply_cannon_multrec 8220 15.4 3.190 4.316 7.590 8.454 init_scf_loop 11 6.9 0.000 0.000 7.511 7.512 mp_waitall_1 103326 16.6 5.671 7.160 5.671 7.160 qs_ot_get_derivative 99 11.5 0.001 0.001 6.839 6.866 sum_up_and_integrate 110 10.3 0.001 0.002 6.104 6.115 integrate_v_rspace 110 11.3 0.003 0.003 6.077 6.088 prepare_preconditioner 11 7.9 0.000 0.000 5.889 5.893 make_preconditioner 11 8.9 0.000 0.000 5.889 5.893 qs_rho_update_rho_low 110 7.6 0.001 0.001 5.854 5.871 calculate_rho_elec 110 8.6 0.114 0.115 5.854 5.870 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.477 5.555 dbcsr_mm_accdrv_process 17442 15.9 3.197 4.124 4.259 5.181 init_scf_run 11 5.9 0.000 0.001 5.101 5.101 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.101 5.101 qs_ot_get_p 110 10.4 0.002 0.002 4.521 4.535 make_m2s 4110 13.4 0.038 0.040 4.146 4.435 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 2.944 4.313 make_images 4110 14.4 0.650 0.713 4.016 4.306 pw_transfer 1331 11.6 0.065 0.071 3.754 3.769 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.708 3.759 apply_single 110 13.6 0.000 0.000 3.708 3.758 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.646 3.665 ot_diis_step 99 11.5 0.012 0.012 3.643 3.643 grid_integrate_task_list 110 12.3 3.375 3.502 3.375 3.502 qs_ot_p2m_diag 48 11.0 0.081 0.084 3.196 3.200 fft_wrap_pw1pw2_140 451 13.1 0.777 0.787 3.171 3.192 density_rs2pw 110 9.6 0.004 0.004 2.849 2.977 cp_dbcsr_syevd 48 12.0 0.003 0.003 2.863 2.863 cp_fm_cholesky_invert 11 10.9 2.783 2.786 2.783 2.786 hybrid_alltoall_any 4261 16.3 0.199 0.853 2.246 2.776 wfi_extrapolate 11 7.9 0.001 0.001 2.667 2.667 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 2.664 2.664 make_images_data 4110 15.4 0.040 0.046 2.225 2.613 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.512 2.513 calculate_dm_sparse 110 9.5 0.001 0.001 2.476 2.512 multiply_cannon_sync_h2d 8220 15.4 2.391 2.482 2.391 2.482 grid_collocate_task_list 110 9.6 2.368 2.446 2.368 2.446 calculate_first_density_matrix 1 7.0 0.000 0.000 2.338 2.341 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.329 2.336 cp_fm_diag_elpa_base 48 14.0 2.274 2.296 2.328 2.335 fft3d_ps 1111 14.6 1.131 1.161 2.312 2.322 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.138 2.158 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 1.970 2.001 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.796 1.999 potential_pw2rs 110 12.3 0.015 0.015 1.913 1.916 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 1.874 1.888 mp_alltoall_d11v 2046 13.8 1.657 1.864 1.657 1.864 cp_fm_cholesky_decompose 22 10.9 1.651 1.666 1.651 1.666 qs_env_update_s_mstruct 11 6.9 0.001 0.002 1.488 1.601 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.592 1.594 mp_allgather_i34 2055 14.4 0.429 1.511 0.429 1.511 dbcsr_complete_redistribute 325 12.2 0.571 0.599 1.421 1.500 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.432 1.444 acc_transpose_blocks 8220 15.4 0.039 0.040 1.357 1.382 qs_create_task_list 11 7.9 0.000 0.001 1.207 1.307 generate_qs_task_list 11 8.9 0.374 0.442 1.207 1.306 transfer_rs2pw 451 10.6 0.005 0.005 1.089 1.228 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.189 1.213 mp_waitany 9240 13.8 1.045 1.192 1.045 1.192 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=54.303000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=807.909091, yerr=12.161448 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 1.368257E+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 0.033 0.067 86.617 86.626 qs_mol_dyn_low 1 2.0 0.003 0.003 86.188 86.228 qs_forces 11 3.9 0.003 0.004 85.382 85.383 qs_energies 11 4.9 0.001 0.002 81.262 81.264 scf_env_do_scf 11 5.9 0.001 0.001 71.246 71.246 velocity_verlet 10 3.0 0.002 0.002 55.496 55.594 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 43.460 43.462 dbcsr_multiply_generic 2055 12.4 0.119 0.125 29.308 29.409 init_scf_loop 11 6.9 0.000 0.000 27.715 27.717 qs_scf_new_mos 99 7.5 0.001 0.001 26.455 26.492 qs_scf_loop_do_ot 99 8.5 0.001 0.001 26.454 26.492 prepare_preconditioner 11 7.9 0.000 0.000 25.740 25.748 make_preconditioner 11 8.9 0.000 0.000 25.740 25.748 make_full_inverse_cholesky 11 9.9 0.000 0.000 20.045 25.223 ot_scf_mini 99 9.5 0.002 0.002 24.657 24.683 multiply_cannon 2055 13.4 0.331 0.351 22.346 23.117 multiply_cannon_loop 2055 14.4 0.826 0.850 20.568 20.826 cp_fm_upper_to_full 70 14.2 12.546 17.969 12.546 17.969 ot_mini 99 10.5 0.001 0.001 13.972 13.998 rebuild_ks_matrix 110 8.3 0.000 0.001 13.408 13.442 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 13.408 13.442 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.162 12.192 dbcsr_complete_redistribute 325 12.2 1.011 1.028 7.224 10.324 multiply_cannon_multrec 8220 15.4 4.074 4.226 9.771 9.935 qs_ot_get_derivative 99 11.5 0.001 0.001 9.422 9.447 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.234 9.341 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 5.680 8.755 mp_alltoall_i22 605 13.7 5.302 8.433 5.302 8.433 mp_waitall_1 84994 16.7 7.303 8.134 7.303 8.134 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.965 6.998 calculate_rho_elec 110 8.6 0.224 0.224 6.964 6.997 sum_up_and_integrate 110 10.3 0.002 0.002 6.618 6.630 integrate_v_rspace 110 11.3 0.004 0.004 6.590 6.603 dbcsr_mm_accdrv_process 11614 15.7 3.949 4.131 5.549 5.773 init_scf_run 11 5.9 0.000 0.001 5.754 5.754 scf_env_initial_rho_setup 11 6.9 0.001 0.001 5.753 5.753 make_m2s 4110 13.4 0.044 0.044 5.246 5.745 make_images 4110 14.4 0.883 0.935 5.058 5.553 qs_ot_get_p 110 10.4 0.002 0.002 5.412 5.445 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.718 5.187 apply_single 110 13.6 0.000 0.000 4.718 5.187 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 4.749 5.080 cp_fm_cholesky_invert 11 10.9 5.070 5.074 5.070 5.074 pw_transfer 1331 11.6 0.074 0.074 4.813 4.820 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 4.697 4.704 ot_diis_step 99 11.5 0.015 0.016 4.532 4.532 fft_wrap_pw1pw2_140 451 13.1 1.278 1.284 4.117 4.125 qs_ot_p2m_diag 48 11.0 0.151 0.156 3.887 3.897 grid_integrate_task_list 110 12.3 3.669 3.730 3.669 3.730 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 3.660 3.661 hybrid_alltoall_any 4261 16.3 0.262 0.563 2.767 3.486 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.987 3.461 density_rs2pw 110 9.6 0.004 0.004 3.441 3.457 cp_dbcsr_syevd 48 12.0 0.003 0.003 3.371 3.372 make_images_data 4110 15.4 0.045 0.048 2.738 3.338 wfi_extrapolate 11 7.9 0.001 0.001 3.271 3.271 calculate_dm_sparse 110 9.5 0.001 0.001 3.179 3.203 multiply_cannon_sync_h2d 8220 15.4 3.121 3.150 3.121 3.150 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.877 2.879 cp_fm_diag_elpa 48 13.0 0.000 0.000 2.798 2.799 cp_fm_diag_elpa_base 48 14.0 2.267 2.462 2.796 2.797 fft3d_ps 1111 14.6 1.297 1.303 2.751 2.755 grid_collocate_task_list 110 9.6 2.662 2.674 2.662 2.674 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.623 2.637 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.388 2.403 calculate_first_density_matrix 1 7.0 0.000 0.000 2.368 2.370 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.113 2.254 qs_env_update_s_mstruct 11 6.9 0.001 0.001 2.172 2.222 potential_pw2rs 110 12.3 0.021 0.021 2.180 2.182 mp_alltoall_d11v 2046 13.8 1.976 2.044 1.976 2.044 cp_fm_cholesky_decompose 22 10.9 1.916 1.929 1.916 1.929 qs_create_task_list 11 7.9 0.000 0.000 1.878 1.922 generate_qs_task_list 11 8.9 0.735 0.786 1.878 1.922 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.844 1.849 acc_transpose_blocks 8220 15.4 0.040 0.040 1.743 1.783 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.712 1.747 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=86.626000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1248.727273, yerr=50.623550 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/15/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430460020736 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613089636352 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228663E+12 0.0% 0.0% 100.0% flops max/rank 1.103326E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806316384 0.0% 0.0% 100.0% number of processed stacks 12044928 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.1 marketing flops 145.647559E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 626.958336E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10431456 MPI messages size (bytes): total size 4.526012E+12 min size 0.000000E+00 max size 4.537280E+06 average size 433.881156E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3605352 96403587072 32768 < size <= 131072 1305088 74666999808 131072 < size <= 4194304 5190114 3200148350408 4194304 < size <= 16777216 263934 1154745839544 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4024 57450. MP_Allreduce 11139 794. MP_Sync 87 MP_Alltoall 2242 1541531. MP_SendRecv 24510 18752. MP_ISendRecv 24510 18752. MP_Wait 42802 MP_ISend 16140 108019. MP_IRecv 16140 108019. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.010 0.026 199.182 199.183 qs_mol_dyn_low 1 2.0 0.003 0.003 198.743 198.756 qs_forces 11 3.9 0.005 0.005 198.658 198.659 qs_energies 11 4.9 0.012 0.021 193.007 193.018 scf_env_do_scf 11 5.9 0.001 0.001 176.283 176.287 scf_env_do_scf_inner_loop 118 6.6 0.003 0.008 156.115 156.117 dbcsr_multiply_generic 2527 12.6 0.180 0.186 120.642 121.545 velocity_verlet 10 3.0 0.005 0.013 121.298 121.299 qs_scf_new_mos 118 7.6 0.001 0.001 116.882 117.305 qs_scf_loop_do_ot 118 8.6 0.001 0.001 116.881 117.304 ot_scf_mini 118 9.6 0.003 0.003 110.357 110.780 multiply_cannon 2527 13.6 0.245 0.254 100.746 102.373 multiply_cannon_loop 2527 14.6 2.279 2.370 98.608 100.197 ot_mini 118 10.6 0.001 0.001 62.401 62.852 multiply_cannon_multrec 60648 15.6 33.093 34.798 42.580 44.383 qs_ot_get_derivative 118 11.6 0.001 0.001 37.139 37.579 rebuild_ks_matrix 129 8.3 0.001 0.001 33.717 34.034 qs_ks_build_kohn_sham_matrix 129 9.3 0.015 0.017 33.717 34.034 qs_ks_update_qs_env 129 7.6 0.001 0.001 30.306 30.556 multiply_cannon_sync_h2d 60648 15.6 28.259 30.144 28.259 30.144 mp_waitall_1 269254 16.5 25.434 28.547 25.434 28.547 qs_ot_get_p 129 10.4 0.001 0.001 26.223 26.562 apply_preconditioner_dbcsr 129 12.6 0.000 0.001 24.746 25.896 apply_single 129 13.6 0.001 0.001 24.746 25.896 ot_diis_step 118 11.6 0.008 0.008 25.067 25.068 init_scf_loop 11 6.9 0.000 0.001 20.095 20.096 qs_ot_p2m_diag 83 11.4 0.079 0.091 19.830 19.890 cp_dbcsr_syevd 83 12.4 0.005 0.005 17.958 17.958 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 16.001 16.340 multiply_cannon_metrocomm3 60648 15.6 0.122 0.125 13.952 15.643 prepare_preconditioner 11 7.9 0.000 0.000 15.523 15.585 make_preconditioner 11 8.9 0.000 0.000 15.523 15.585 cp_fm_diag_elpa 83 13.4 0.000 0.000 15.056 15.099 cp_fm_diag_elpa_base 83 14.4 14.990 15.032 15.052 15.097 make_full_inverse_cholesky 11 9.9 0.000 0.000 14.754 14.945 make_m2s 5054 13.6 0.107 0.118 13.773 14.233 make_images 5054 14.6 0.400 0.419 13.587 14.057 sum_up_and_integrate 129 10.3 0.002 0.005 13.916 13.931 integrate_v_rspace 129 11.3 0.003 0.004 13.856 13.875 qs_rho_update_rho_low 129 7.7 0.001 0.001 12.863 12.918 calculate_rho_elec 129 8.7 0.045 0.064 12.862 12.917 init_scf_run 11 5.9 0.000 0.001 12.382 12.383 scf_env_initial_rho_setup 11 6.9 0.002 0.003 12.382 12.382 dbcsr_mm_accdrv_process 125468 16.2 4.269 4.605 8.988 9.535 wfi_extrapolate 11 7.9 0.001 0.001 9.079 9.079 cp_fm_cholesky_invert 11 10.9 8.793 8.801 8.793 8.801 calculate_dm_sparse 129 9.5 0.001 0.001 8.368 8.458 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 7.992 8.138 pw_transfer 1559 11.6 0.074 0.089 7.542 7.719 make_images_data 5054 15.6 0.066 0.072 6.750 7.648 grid_integrate_task_list 129 12.3 7.100 7.516 7.100 7.516 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 7.339 7.516 qs_ot_get_orbitals 118 10.6 0.001 0.001 7.396 7.474 hybrid_alltoall_any 5240 16.5 0.296 2.271 5.919 7.223 density_rs2pw 129 9.7 0.006 0.007 6.623 7.118 multiply_cannon_metrocomm1 60648 15.6 0.097 0.101 5.125 6.783 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.003 6.694 6.708 fft_wrap_pw1pw2_140 527 13.2 1.135 1.181 6.497 6.671 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.822 5.986 mp_sum_l 8010 12.9 4.820 5.888 4.820 5.888 fft3d_ps 1301 14.7 2.211 2.864 5.330 5.664 grid_collocate_task_list 129 9.7 4.876 5.266 4.876 5.266 mp_alltoall_d11v 2423 14.1 4.124 4.901 4.124 4.901 cp_fm_cholesky_decompose 22 10.9 4.555 4.568 4.555 4.568 potential_pw2rs 129 12.3 0.009 0.010 4.365 4.385 mp_sum_d 4499 12.1 3.490 4.072 3.490 4.072 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=199.183000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=595.090909, yerr=4.357951 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420241154048 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528896499712 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514757E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+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 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 840.507392E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834560144 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58214. MP_Allreduce 11083 960. MP_Sync 87 MP_Alltoall 1969 6182355. MP_SendRecv 12032 47072. MP_ISendRecv 12032 47072. MP_Wait 25916 MP_ISend 11748 212467. MP_IRecv 11748 212467. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.019 0.055 188.474 188.485 qs_mol_dyn_low 1 2.0 0.003 0.004 188.103 188.119 qs_forces 11 3.9 0.004 0.005 187.646 187.647 qs_energies 11 4.9 0.002 0.002 180.889 180.901 scf_env_do_scf 11 5.9 0.001 0.001 164.472 164.483 scf_env_do_scf_inner_loop 117 6.6 0.006 0.013 132.050 132.052 velocity_verlet 10 3.0 0.002 0.002 118.678 118.688 dbcsr_multiply_generic 2507 12.6 0.212 0.220 97.415 98.504 qs_scf_new_mos 117 7.6 0.001 0.001 93.634 94.108 qs_scf_loop_do_ot 117 8.6 0.001 0.001 93.633 94.107 ot_scf_mini 117 9.6 0.004 0.004 88.835 89.397 multiply_cannon 2507 13.6 0.514 0.571 77.439 81.892 multiply_cannon_loop 2507 14.6 1.566 1.631 74.034 76.504 ot_mini 117 10.6 0.001 0.001 50.067 50.616 mp_waitall_1 214728 16.6 23.641 38.825 23.641 38.825 multiply_cannon_multrec 30084 15.6 21.217 26.000 31.787 37.042 rebuild_ks_matrix 128 8.3 0.001 0.001 32.234 32.870 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 32.233 32.869 init_scf_loop 11 6.9 0.000 0.000 32.335 32.335 qs_ks_update_qs_env 128 7.6 0.001 0.001 28.980 29.563 multiply_cannon_metrocomm3 30084 15.6 0.096 0.102 15.413 28.757 qs_ot_get_derivative 117 11.6 0.001 0.001 28.146 28.697 prepare_preconditioner 11 7.9 0.000 0.000 28.007 28.081 make_preconditioner 11 8.9 0.000 0.000 28.007 28.081 make_full_inverse_cholesky 11 9.9 0.000 0.000 26.735 27.286 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.066 23.211 apply_single 128 13.6 0.001 0.001 22.065 23.211 ot_diis_step 117 11.6 0.014 0.015 21.747 21.749 qs_ot_get_p 128 10.4 0.008 0.015 20.788 21.442 multiply_cannon_sync_h2d 30084 15.6 18.261 20.711 18.261 20.711 cp_fm_cholesky_invert 11 10.9 16.374 16.386 16.374 16.386 qs_ot_p2m_diag 83 11.4 0.188 0.216 15.844 15.876 make_m2s 5014 13.6 0.090 0.096 14.045 15.754 make_images 5014 14.6 1.173 1.375 13.832 15.542 cp_dbcsr_syevd 83 12.4 0.005 0.006 14.653 14.654 sum_up_and_integrate 128 10.3 0.002 0.004 13.915 13.943 integrate_v_rspace 128 11.3 0.003 0.004 13.855 13.883 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.271 13.299 calculate_rho_elec 128 8.7 0.087 0.104 13.270 13.298 init_scf_run 11 5.9 0.000 0.002 11.580 11.581 scf_env_initial_rho_setup 11 6.9 0.001 0.001 11.580 11.581 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 11.161 11.545 cp_fm_diag_elpa 83 13.4 0.000 0.000 11.513 11.543 cp_fm_diag_elpa_base 83 14.4 11.261 11.355 11.508 11.536 multiply_cannon_metrocomm4 27577 15.6 0.104 0.118 3.779 11.017 mp_irecv_dv 69486 16.3 3.576 10.612 3.576 10.612 dbcsr_mm_accdrv_process 62242 16.2 5.455 6.181 10.018 10.569 make_images_data 5014 15.6 0.066 0.076 8.255 10.282 hybrid_alltoall_any 5200 16.5 0.351 1.541 7.011 9.794 pw_transfer 1547 11.6 0.084 0.095 8.536 8.603 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.314 8.383 wfi_extrapolate 11 7.9 0.001 0.001 8.319 8.319 grid_integrate_task_list 128 12.3 7.186 7.533 7.186 7.533 fft_wrap_pw1pw2_140 523 13.2 1.209 1.233 7.389 7.467 density_rs2pw 128 9.7 0.006 0.007 6.980 7.431 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 6.238 7.022 cp_fm_cholesky_decompose 22 10.9 6.818 6.897 6.818 6.897 calculate_dm_sparse 128 9.5 0.001 0.001 6.470 6.616 mp_sum_l 7950 12.9 4.110 6.236 4.110 6.236 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.167 6.175 fft3d_ps 1291 14.7 2.841 3.023 5.878 5.945 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.301 5.448 grid_collocate_task_list 128 9.7 5.008 5.430 5.008 5.430 qs_ot_get_orbitals 117 10.6 0.001 0.001 5.357 5.420 mp_allgather_i34 2507 14.6 1.907 4.987 1.907 4.987 mp_alltoall_d11v 2415 14.1 4.056 4.559 4.056 4.559 potential_pw2rs 128 12.3 0.016 0.018 4.402 4.415 mp_sum_d 4470 12.1 2.595 3.980 2.595 3.980 dbcsr_complete_redistribute 395 12.7 0.774 0.857 3.054 3.892 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=188.485000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=801.000000, yerr=2.860388 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430458527744 0.0% 0.0% 100.0% flops 32 x 32 x 32 1958505086976 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986244964352 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992000282624 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753956716544 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613083000832 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239146475520 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239146475520 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911124992000 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.228655E+12 0.0% 0.0% 100.0% flops max/rank 2.950886E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806314816 0.0% 0.0% 100.0% number of processed stacks 4015680 0.0% 0.0% 100.0% average stack size 0.0 0.0 1694.9 marketing flops 145.646636E+12 ------------------------------------------------------------------------------- # multiplications 2527 max memory usage/rank 977.334272E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1051232 MPI messages size (bytes): total size 2.737039E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.603648E+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 284160 37245419520 131072 < size <= 4194304 665408 1004401590272 4194304 < size <= 16777216 66080 937889582784 16777216 < size 28896 757491302400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4030 57812. MP_Allreduce 11152 997. MP_Sync 87 MP_Alltoall 1724 9394160. MP_SendRecv 7998 75008. MP_ISendRecv 7998 75008. MP_Wait 21986 MP_ISend 11836 275177. MP_IRecv 11836 275177. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.025 0.061 176.236 176.246 qs_mol_dyn_low 1 2.0 0.003 0.004 175.808 175.827 qs_forces 11 3.9 0.004 0.005 175.710 175.711 qs_energies 11 4.9 0.001 0.002 169.161 169.171 scf_env_do_scf 11 5.9 0.001 0.001 153.507 153.508 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 118.396 118.396 velocity_verlet 10 3.0 0.001 0.002 113.399 113.402 dbcsr_multiply_generic 2527 12.6 0.181 0.186 82.360 83.546 qs_scf_new_mos 118 7.6 0.001 0.001 81.097 81.433 qs_scf_loop_do_ot 118 8.6 0.001 0.001 81.097 81.432 ot_scf_mini 118 9.6 0.003 0.004 76.800 77.196 multiply_cannon 2527 13.6 0.512 0.535 62.467 67.299 multiply_cannon_loop 2527 14.6 1.132 1.189 59.532 62.313 ot_mini 118 10.6 0.001 0.001 43.042 43.444 init_scf_loop 11 6.9 0.000 0.000 35.011 35.012 mp_waitall_1 171878 16.6 24.518 34.302 24.518 34.302 prepare_preconditioner 11 7.9 0.000 0.000 31.041 31.095 make_preconditioner 11 8.9 0.000 0.000 31.041 31.095 rebuild_ks_matrix 129 8.3 0.001 0.001 30.284 30.768 qs_ks_build_kohn_sham_matrix 129 9.3 0.017 0.024 30.283 30.768 make_full_inverse_cholesky 11 9.9 0.000 0.000 28.656 30.110 qs_ks_update_qs_env 129 7.6 0.001 0.001 27.318 27.751 multiply_cannon_multrec 20216 15.6 13.286 16.716 22.682 26.180 multiply_cannon_metrocomm3 20216 15.6 0.062 0.066 15.183 24.774 qs_ot_get_derivative 118 11.6 0.001 0.002 23.344 23.732 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 19.718 20.768 apply_single 129 13.6 0.001 0.001 19.718 20.768 qs_ot_get_p 129 10.4 0.002 0.006 19.117 19.711 ot_diis_step 118 11.6 0.018 0.020 19.586 19.586 make_m2s 5054 13.6 0.081 0.087 14.755 15.953 multiply_cannon_sync_h2d 20216 15.6 13.795 15.876 13.795 15.876 make_images 5054 14.6 1.169 1.259 14.518 15.712 qs_ot_p2m_diag 83 11.4 0.265 0.272 14.722 14.729 cp_fm_cholesky_invert 11 10.9 14.394 14.403 14.394 14.403 sum_up_and_integrate 129 10.3 0.002 0.003 14.023 14.047 integrate_v_rspace 129 11.3 0.003 0.004 13.963 13.985 qs_rho_update_rho_low 129 7.7 0.001 0.001 13.653 13.688 calculate_rho_elec 129 8.7 0.131 0.146 13.652 13.687 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.658 13.659 init_scf_run 11 5.9 0.000 0.001 10.629 10.629 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.629 10.629 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.470 10.489 cp_fm_diag_elpa_base 83 14.4 10.060 10.214 10.467 10.485 make_images_data 5054 15.6 0.062 0.070 8.735 10.406 hybrid_alltoall_any 5240 16.5 0.453 2.059 7.492 9.939 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.147 9.411 multiply_cannon_metrocomm4 17689 15.6 0.067 0.076 3.449 9.102 dbcsr_mm_accdrv_process 41830 16.2 5.636 5.974 8.845 8.975 pw_transfer 1559 11.6 0.085 0.099 8.739 8.846 mp_irecv_dv 50625 16.2 3.321 8.840 3.321 8.840 fft_wrap_pw1pw2 1301 12.7 0.010 0.011 8.515 8.629 grid_integrate_task_list 129 12.3 7.341 7.709 7.341 7.709 fft_wrap_pw1pw2_140 527 13.2 1.309 1.340 7.558 7.683 wfi_extrapolate 11 7.9 0.001 0.001 7.416 7.416 cp_fm_cholesky_decompose 22 10.9 7.385 7.409 7.385 7.409 density_rs2pw 129 9.7 0.006 0.006 6.934 7.388 cp_fm_upper_to_full 105 14.8 5.813 7.337 5.813 7.337 dbcsr_complete_redistribute 395 12.7 1.170 1.203 4.611 6.371 calculate_dm_sparse 129 9.5 0.001 0.001 5.886 5.990 fft3d_ps 1301 14.7 2.783 3.011 5.876 5.965 grid_collocate_task_list 129 9.7 5.195 5.738 5.195 5.738 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.423 5.429 qs_ot_get_derivative_taylor 41 13.0 0.001 0.001 4.776 5.412 copy_fm_to_dbcsr 209 11.7 0.002 0.002 3.452 5.216 mp_alltoall_d11v 2423 14.1 4.399 4.985 4.399 4.985 mp_sum_l 8010 12.9 3.294 4.874 3.294 4.874 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.523 4.656 mp_allgather_i34 2527 14.6 1.427 4.431 1.427 4.431 potential_pw2rs 129 12.3 0.020 0.022 4.369 4.379 transfer_fm_to_dbcsr 11 9.9 0.019 0.025 2.366 4.121 qs_ot_get_orbitals 118 10.6 0.001 0.001 3.960 3.989 mp_alltoall_i22 718 14.1 1.978 3.865 1.978 3.865 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.737 3.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=176.246000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=926.272727, yerr=16.624685 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 4.353791E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 5977344 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.173955E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1143192 MPI messages size (bytes): total size 2.023815E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.770320E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 319024 36042702848 131072 < size <= 4194304 715736 785529176064 4194304 < size <= 16777216 70320 665379475120 16777216 < size 30720 536870912000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58205. MP_Allreduce 11082 1082. MP_Sync 87 MP_Alltoall 1712 12503107. MP_SendRecv 5888 75008. MP_ISendRecv 5888 75008. MP_Wait 22442 MP_ISend 14952 244818. MP_IRecv 14952 244818. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.062 0.100 189.394 189.404 qs_mol_dyn_low 1 2.0 0.003 0.003 188.600 188.618 qs_forces 11 3.9 0.004 0.004 186.860 186.864 qs_energies 11 4.9 0.004 0.005 179.775 179.784 scf_env_do_scf 11 5.9 0.001 0.001 162.906 162.918 velocity_verlet 10 3.0 0.002 0.002 125.699 125.770 scf_env_do_scf_inner_loop 117 6.6 0.003 0.008 117.037 117.039 dbcsr_multiply_generic 2507 12.6 0.186 0.193 80.651 81.523 qs_scf_new_mos 117 7.6 0.001 0.001 80.873 81.223 qs_scf_loop_do_ot 117 8.6 0.001 0.001 80.872 81.222 ot_scf_mini 117 9.6 0.003 0.003 76.307 76.670 multiply_cannon 2507 13.6 0.553 0.584 55.861 59.171 multiply_cannon_loop 2507 14.6 1.818 1.891 52.151 54.135 init_scf_loop 11 6.9 0.000 0.000 45.744 45.745 ot_mini 117 10.6 0.001 0.001 43.003 43.354 prepare_preconditioner 11 7.9 0.000 0.000 41.666 41.696 make_preconditioner 11 8.9 0.000 0.000 41.666 41.696 make_full_inverse_cholesky 11 9.9 0.011 0.023 35.336 40.283 multiply_cannon_multrec 30084 15.6 13.592 18.820 26.514 31.354 rebuild_ks_matrix 128 8.3 0.001 0.001 29.176 29.481 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.019 29.175 29.480 mp_waitall_1 147882 16.7 17.271 26.716 17.271 26.716 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.341 26.612 qs_ot_get_derivative 117 11.6 0.001 0.002 23.436 23.805 make_m2s 5014 13.6 0.097 0.101 20.340 21.435 make_images 5014 14.6 1.945 2.231 20.027 21.123 qs_ot_get_p 128 10.4 0.001 0.002 19.446 19.805 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 18.954 19.616 apply_single 128 13.6 0.001 0.001 18.954 19.616 ot_diis_step 117 11.6 0.017 0.018 19.449 19.450 cp_fm_upper_to_full 105 14.8 11.322 16.657 11.322 16.657 cp_fm_cholesky_invert 11 10.9 15.863 15.873 15.863 15.873 multiply_cannon_metrocomm3 30084 15.6 0.050 0.054 6.370 15.417 qs_ot_p2m_diag 83 11.4 0.342 0.389 15.203 15.253 sum_up_and_integrate 128 10.3 0.002 0.003 13.913 13.938 integrate_v_rspace 128 11.3 0.003 0.004 13.853 13.880 cp_dbcsr_syevd 83 12.4 0.005 0.006 13.805 13.806 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.705 13.744 calculate_rho_elec 128 8.7 0.173 0.190 13.704 13.744 dbcsr_mm_accdrv_process 62264 16.2 8.451 9.293 12.491 13.027 dbcsr_complete_redistribute 395 12.7 1.498 1.614 9.071 12.758 make_images_data 5014 15.6 0.065 0.072 10.916 12.631 hybrid_alltoall_any 5200 16.5 0.528 2.220 9.685 11.898 copy_fm_to_dbcsr 209 11.7 0.001 0.002 7.692 11.384 multiply_cannon_sync_h2d 30084 15.6 10.431 11.340 10.431 11.340 init_scf_run 11 5.9 0.000 0.002 10.926 10.929 scf_env_initial_rho_setup 11 6.9 0.001 0.001 10.926 10.929 cp_fm_diag_elpa 83 13.4 0.000 0.000 10.592 10.606 cp_fm_diag_elpa_base 83 14.4 9.646 9.967 10.586 10.602 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 9.854 10.116 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.307 9.922 mp_alltoall_i22 716 14.1 5.551 9.345 5.551 9.345 pw_transfer 1547 11.6 0.085 0.100 8.898 8.991 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.673 8.772 grid_integrate_task_list 128 12.3 7.501 7.876 7.501 7.876 fft_wrap_pw1pw2_140 523 13.2 1.443 1.473 7.717 7.823 wfi_extrapolate 11 7.9 0.001 0.001 7.723 7.723 cp_fm_cholesky_decompose 22 10.9 7.463 7.542 7.463 7.542 density_rs2pw 128 9.7 0.006 0.006 6.732 7.135 multiply_cannon_metrocomm4 25070 15.6 0.084 0.093 2.838 7.097 mp_irecv_dv 76098 16.2 2.683 6.823 2.683 6.823 calculate_dm_sparse 128 9.5 0.001 0.001 6.261 6.363 fft3d_ps 1291 14.7 2.845 2.972 5.853 5.939 grid_collocate_task_list 128 9.7 5.288 5.711 5.288 5.711 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.576 5.630 mp_alltoall_d11v 2415 14.1 4.971 5.393 4.971 5.393 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 4.509 4.611 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.473 4.570 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 4.442 4.442 qs_ot_get_orbitals 117 10.6 0.001 0.001 4.215 4.268 potential_pw2rs 128 12.3 0.023 0.023 4.191 4.205 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=189.404000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1111.090909, yerr=17.427536 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420242647040 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528903135232 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514766E+12 0.0% 0.0% 100.0% flops max/rank 5.865089E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755941440 0.0% 0.0% 100.0% number of processed stacks 1960712 0.0% 0.0% 100.0% average stack size 0.0 0.0 3445.7 marketing flops 144.579337E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 1.540915E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 240672 MPI messages size (bytes): total size 1.331455E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.532238E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 113904 59718500352 4194304 < size <= 16777216 104976 550376570880 16777216 < size 20208 721350232272 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8931 51. MP_Alltoall 9654 799394. MP_ISend 40068 2102573. MP_IRecv 40068 2101676. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4002 58203. MP_Allreduce 11082 1166. MP_Sync 87 MP_Alltoall 1712 18838222. MP_SendRecv 3840 122880. MP_ISendRecv 3840 122880. MP_Wait 16122 MP_ISend 10680 423556. MP_IRecv 10680 423556. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.046 0.073 168.577 168.580 qs_mol_dyn_low 1 2.0 0.003 0.003 168.087 168.118 qs_forces 11 3.9 0.004 0.004 167.251 167.253 qs_energies 11 4.9 0.002 0.002 159.874 159.881 scf_env_do_scf 11 5.9 0.001 0.001 142.833 142.846 velocity_verlet 10 3.0 0.021 0.022 111.012 111.062 scf_env_do_scf_inner_loop 117 6.6 0.005 0.008 107.586 107.588 dbcsr_multiply_generic 2507 12.6 0.181 0.191 71.589 71.963 qs_scf_new_mos 117 7.6 0.001 0.001 71.140 71.227 qs_scf_loop_do_ot 117 8.6 0.001 0.001 71.139 71.226 ot_scf_mini 117 9.6 0.003 0.004 66.731 66.795 multiply_cannon 2507 13.6 0.560 0.588 53.125 56.489 multiply_cannon_loop 2507 14.6 0.804 0.840 50.217 50.974 ot_mini 117 10.6 0.001 0.001 37.315 37.370 init_scf_loop 11 6.9 0.000 0.000 35.097 35.099 prepare_preconditioner 11 7.9 0.000 0.000 31.237 31.258 make_preconditioner 11 8.9 0.000 0.000 31.237 31.258 mp_waitall_1 125778 16.7 24.497 30.490 24.497 30.490 make_full_inverse_cholesky 11 9.9 0.015 0.028 29.103 29.396 rebuild_ks_matrix 128 8.3 0.001 0.001 28.839 28.927 qs_ks_build_kohn_sham_matrix 128 9.3 0.017 0.018 28.838 28.927 qs_ks_update_qs_env 128 7.6 0.001 0.001 26.310 26.395 multiply_cannon_multrec 10028 15.6 10.392 14.549 17.969 21.210 qs_ot_get_derivative 117 11.6 0.001 0.001 20.331 20.396 multiply_cannon_metrocomm3 10028 15.6 0.025 0.026 12.450 19.582 cp_fm_cholesky_invert 11 10.9 17.750 17.755 17.750 17.755 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 16.996 17.138 apply_single 128 13.6 0.001 0.001 16.996 17.137 ot_diis_step 117 11.6 0.020 0.020 16.911 16.912 qs_ot_get_p 128 10.4 0.012 0.012 16.776 16.844 make_m2s 5014 13.6 0.066 0.069 14.660 15.497 make_images 5014 14.6 2.176 2.605 14.352 15.186 sum_up_and_integrate 128 10.3 0.002 0.002 13.998 14.050 integrate_v_rspace 128 11.3 0.004 0.004 13.937 13.992 qs_rho_update_rho_low 128 7.7 0.001 0.001 13.871 13.909 calculate_rho_elec 128 8.7 0.256 0.267 13.870 13.909 qs_ot_p2m_diag 83 11.4 0.495 0.501 13.043 13.061 cp_dbcsr_syevd 83 12.4 0.005 0.006 11.889 11.890 multiply_cannon_sync_h2d 10028 15.6 10.898 11.358 10.898 11.358 init_scf_run 11 5.9 0.000 0.002 10.243 10.243 scf_env_initial_rho_setup 11 6.9 0.011 0.020 10.242 10.243 hybrid_alltoall_any 5200 16.5 0.844 3.799 8.304 9.713 make_images_data 5014 15.6 0.054 0.064 8.434 9.693 pw_transfer 1547 11.6 0.084 0.093 9.146 9.169 cp_fm_diag_elpa 83 13.4 0.000 0.000 9.040 9.053 cp_fm_diag_elpa_base 83 14.4 8.800 8.882 9.037 9.049 fft_wrap_pw1pw2 1291 12.7 0.010 0.011 8.924 8.955 grid_integrate_task_list 128 12.3 7.765 8.174 7.765 8.174 qs_ot_get_derivative_diag 77 12.4 0.002 0.003 7.960 8.010 fft_wrap_pw1pw2_140 523 13.2 1.802 1.834 7.930 7.963 dbcsr_mm_accdrv_process 20762 16.1 3.229 4.434 7.221 7.920 cp_fm_cholesky_decompose 22 10.9 7.758 7.869 7.758 7.869 wfi_extrapolate 11 7.9 0.001 0.001 7.275 7.275 multiply_cannon_metrocomm1 10028 15.6 0.030 0.031 4.413 7.195 density_rs2pw 128 9.7 0.005 0.006 6.607 6.939 calculate_dm_sparse 128 9.5 0.001 0.001 6.047 6.117 grid_collocate_task_list 128 9.7 5.557 5.870 5.557 5.870 fft3d_ps 1291 14.7 2.763 2.835 5.652 5.677 dbcsr_complete_redistribute 395 12.7 2.108 2.160 5.196 5.627 qs_energies_init_hamiltonians 11 5.9 0.006 0.009 5.249 5.250 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.096 5.103 mp_alltoall_d11v 2415 14.1 4.503 4.960 4.503 4.960 mp_allgather_i34 2507 14.6 1.103 4.548 1.103 4.548 potential_pw2rs 128 12.3 0.026 0.027 4.170 4.174 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.026 4.073 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 3.595 3.906 multiply_cannon_metrocomm4 7521 15.6 0.026 0.028 1.724 3.904 mp_irecv_dv 28860 15.9 1.684 3.823 1.684 3.823 copy_fm_to_dbcsr 209 11.7 0.001 0.002 3.364 3.692 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.577 3.636 qs_ot_get_orbitals 117 10.6 0.001 0.001 3.551 3.587 copy_dbcsr_to_fm 186 11.8 0.004 0.004 3.425 3.578 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 3.430 3.438 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=168.580000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1454.454545, yerr=11.113985 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410025439232 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 1957874282496 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963550527488 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 11444716634112 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019206377472 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019206377472 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 92.796636E+12 0.0% 0.0% 100.0% flops max/rank 11.606419E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705506624 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 3.141509E+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 716109171264 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4047 58501. MP_Allreduce 11195 1501. MP_Sync 86 MP_Alltoall 1700 36954361. 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 0.028 0.054 286.160 286.161 qs_mol_dyn_low 1 2.0 0.003 0.005 285.596 285.629 qs_forces 11 3.9 0.005 0.005 285.500 285.503 qs_energies 11 4.9 0.017 0.018 276.676 276.687 scf_env_do_scf 11 5.9 0.001 0.001 254.233 254.244 velocity_verlet 10 3.0 0.013 0.014 207.053 207.061 scf_env_do_scf_inner_loop 116 6.6 0.004 0.009 130.735 130.737 init_scf_loop 11 6.9 0.000 0.000 123.229 123.231 prepare_preconditioner 11 7.9 0.000 0.000 118.448 118.467 make_preconditioner 11 8.9 0.000 0.000 118.448 118.467 make_full_inverse_cholesky 11 9.9 0.037 0.039 94.847 115.640 qs_scf_new_mos 116 7.6 0.001 0.001 88.059 88.105 qs_scf_loop_do_ot 116 8.6 0.001 0.001 88.059 88.104 ot_scf_mini 116 9.6 0.004 0.004 83.258 83.264 dbcsr_multiply_generic 2485 12.5 0.211 0.225 81.036 81.450 cp_fm_upper_to_full 104 14.8 52.713 75.301 52.713 75.301 multiply_cannon 2485 13.5 0.692 0.724 58.269 59.923 multiply_cannon_loop 2485 14.5 1.028 1.059 54.259 55.373 ot_mini 116 10.6 0.001 0.001 43.483 43.495 dbcsr_complete_redistribute 393 12.7 3.998 4.046 29.494 42.295 copy_fm_to_dbcsr 208 11.6 0.001 0.002 26.012 38.782 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 23.558 36.122 mp_alltoall_i22 712 14.1 21.327 34.171 21.327 34.171 rebuild_ks_matrix 127 8.3 0.001 0.001 33.427 33.437 qs_ks_build_kohn_sham_matrix 127 9.3 0.018 0.018 33.426 33.436 cp_fm_cholesky_invert 11 10.9 32.987 32.993 32.987 32.993 mp_waitall_1 102768 16.8 27.479 31.907 27.479 31.907 qs_ks_update_qs_env 127 7.6 0.001 0.001 31.008 31.036 qs_ot_get_p 127 10.4 0.002 0.002 24.786 24.829 qs_ot_get_derivative 116 11.6 0.002 0.002 24.332 24.340 qs_ot_p2m_diag 82 11.4 0.868 0.876 20.803 20.833 make_m2s 4970 13.5 0.075 0.077 18.286 19.778 multiply_cannon_metrocomm3 9940 15.5 0.024 0.027 18.532 19.663 make_images 4970 14.5 3.048 3.223 17.808 19.301 ot_diis_step 116 11.6 0.022 0.022 19.128 19.128 cp_dbcsr_syevd 82 12.4 0.006 0.006 19.038 19.041 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 18.480 18.557 apply_single 127 13.6 0.001 0.001 18.480 18.557 multiply_cannon_multrec 9940 15.5 10.231 12.112 17.949 18.044 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.197 16.209 calculate_rho_elec 127 8.7 0.478 0.479 16.196 16.209 cp_fm_diag_elpa 82 13.4 0.000 0.000 15.923 15.924 cp_fm_diag_elpa_base 82 14.4 11.621 13.163 15.920 15.920 sum_up_and_integrate 127 10.3 0.002 0.002 15.790 15.880 integrate_v_rspace 127 11.3 0.004 0.005 15.728 15.818 multiply_cannon_sync_h2d 9940 15.5 14.204 14.234 14.204 14.234 hybrid_alltoall_any 5155 16.4 1.297 3.039 10.556 12.879 make_images_data 4970 15.5 0.063 0.068 10.264 12.444 init_scf_run 11 5.9 0.000 0.001 12.164 12.164 scf_env_initial_rho_setup 11 6.9 0.001 0.001 12.163 12.164 pw_transfer 1535 11.6 0.093 0.094 11.070 11.074 fft_wrap_pw1pw2 1281 12.7 0.011 0.011 10.834 10.838 fft_wrap_pw1pw2_140 519 13.2 3.017 3.088 9.683 9.690 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.504 9.509 dbcsr_mm_accdrv_process 20590 16.0 4.200 5.951 7.472 9.355 wfi_extrapolate 11 7.9 0.001 0.001 8.921 8.921 cp_fm_cholesky_decompose 22 10.9 8.883 8.919 8.883 8.919 grid_integrate_task_list 127 12.3 8.504 8.686 8.504 8.686 qs_energies_init_hamiltonians 11 5.9 0.001 0.001 8.129 8.130 density_rs2pw 127 9.7 0.005 0.005 7.286 7.326 mp_alltoall_d11v 2401 14.1 6.784 6.945 6.784 6.945 calculate_dm_sparse 127 9.5 0.001 0.001 6.666 6.729 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.429 6.474 grid_collocate_task_list 127 9.7 6.372 6.430 6.372 6.430 copy_dbcsr_to_fm 185 11.7 0.004 0.004 6.121 6.211 fft3d_ps 1281 14.7 2.822 2.834 6.122 6.172 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=286.161000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2830.181818, yerr=155.010743 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 1.260151E+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 0.013 0.077 85.151 85.166 qs_energies 1 2.0 0.000 0.000 84.615 84.631 ls_scf 1 3.0 0.000 0.000 83.710 83.726 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.767 72.944 multiply_cannon 111 7.7 0.017 0.021 56.083 57.110 multiply_cannon_loop 111 8.7 0.211 0.232 52.516 53.764 ls_scf_main 1 4.0 0.000 0.000 52.274 52.286 density_matrix_trs4 2 5.0 0.002 0.003 46.736 46.825 ls_scf_init_scf 1 4.0 0.000 0.000 28.372 28.373 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.246 27.301 mp_waitall_1 11031 10.9 22.635 25.773 22.635 25.773 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 25.098 25.126 multiply_cannon_multrec 2664 9.7 8.120 8.839 15.561 17.300 multiply_cannon_sync_h2d 2664 9.7 13.456 15.511 13.456 15.511 make_m2s 222 7.7 0.009 0.011 13.044 13.641 make_images 222 8.7 0.099 0.107 13.022 13.622 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.836 12.119 multiply_cannon_metrocomm3 2664 9.7 0.009 0.011 5.525 8.270 make_images_data 222 9.7 0.004 0.005 7.609 8.084 dbcsr_mm_accdrv_process 4760 10.4 0.582 0.711 7.057 8.058 hybrid_alltoall_any 227 10.6 0.215 1.831 6.514 7.791 dbcsr_mm_accdrv_process_sort 4760 11.4 6.276 7.153 6.276 7.153 calculate_norms 4752 9.8 5.480 6.121 5.480 6.121 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.076 5.160 mp_sum_l 887 5.1 3.131 4.723 3.131 4.723 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.044 3.608 mp_irecv_dv 6231 10.9 2.028 3.586 2.028 3.586 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.288 3.404 make_images_sizes 222 9.7 0.000 0.000 0.698 3.323 mp_alltoall_i44 222 10.7 0.698 3.323 0.698 3.323 arnoldi_extremal 4 6.8 0.000 0.000 3.189 3.216 arnoldi_normal_ev 4 7.8 0.001 0.002 3.189 3.216 build_subspace 16 8.4 0.009 0.012 3.089 3.091 ls_scf_post 1 4.0 0.000 0.000 3.064 3.080 ls_scf_store_result 1 5.0 0.000 0.000 2.875 2.922 dbcsr_special_finalize 555 9.7 0.005 0.006 2.384 2.800 dbcsr_merge_single_wm 555 10.7 0.491 0.630 2.375 2.791 make_images_pack 222 9.7 2.206 2.620 2.208 2.622 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.320 2.555 dbcsr_sort_data 658 11.4 2.142 2.470 2.142 2.470 dbcsr_matrix_vector_mult_local 304 10.0 2.071 2.456 2.073 2.458 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.274 2.335 buffer_matrices_ensure_size 222 8.7 1.833 2.166 1.833 2.166 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.758 1.762 rebuild_ks_matrix 3 7.3 0.000 0.000 1.749 1.752 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.749 1.752 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.166000, 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/b1e573e5365af53cd5b64772cda048b47a415cd7_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 2.167513E+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 0.023 0.040 90.053 90.055 qs_energies 1 2.0 0.000 0.000 89.598 89.602 ls_scf 1 3.0 0.000 0.000 88.267 88.270 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.364 74.706 multiply_cannon 111 7.7 0.030 0.044 52.577 57.421 ls_scf_main 1 4.0 0.000 0.000 54.630 54.638 multiply_cannon_loop 111 8.7 0.135 0.145 49.795 53.624 density_matrix_trs4 2 5.0 0.002 0.003 48.959 49.170 mp_waitall_1 9105 10.9 20.494 29.916 20.494 29.916 ls_scf_init_scf 1 4.0 0.000 0.000 29.903 29.907 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.700 28.798 multiply_cannon_multrec 1332 9.7 13.372 16.776 22.720 27.268 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.346 26.355 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.246 20.874 make_m2s 222 7.7 0.006 0.008 15.131 15.698 make_images 222 8.7 1.373 1.700 15.101 15.669 dbcsr_mm_accdrv_process 4041 10.4 0.348 0.538 8.943 10.505 dbcsr_mm_accdrv_process_sort 4041 11.4 8.463 9.967 8.463 9.967 hybrid_alltoall_any 227 10.6 0.540 2.542 8.073 9.705 make_images_data 222 9.7 0.004 0.005 8.768 9.556 mp_sum_l 887 5.1 5.184 8.315 5.184 8.315 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.195 7.666 mp_irecv_dv 3311 11.0 3.175 7.612 3.175 7.612 calculate_norms 2376 9.8 6.081 6.650 6.081 6.650 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.908 6.631 multiply_cannon_sync_h2d 1332 9.7 4.940 6.072 4.940 6.072 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.924 5.137 arnoldi_extremal 4 6.8 0.000 0.000 4.657 4.677 arnoldi_normal_ev 4 7.8 0.001 0.005 4.656 4.677 build_subspace 16 8.4 0.014 0.021 4.399 4.402 ls_scf_post 1 4.0 0.000 0.000 3.734 3.738 ls_scf_store_result 1 5.0 0.000 0.000 3.411 3.558 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.179 3.396 dbcsr_matrix_vector_mult_local 304 10.0 2.773 3.256 2.775 3.258 mp_allgather_i34 111 8.7 0.775 3.001 0.775 3.001 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.617 2.717 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.124 2.568 dbcsr_data_new 4174 10.1 2.104 2.407 2.104 2.407 make_images_pack 222 9.7 1.818 2.101 1.820 2.103 dbcsr_sort_data 436 11.2 1.824 2.101 1.824 2.101 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.867 1.870 rebuild_ks_matrix 3 7.3 0.000 0.000 1.854 1.858 qs_ks_build_kohn_sham_matrix 3 8.3 0.008 0.011 1.854 1.858 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.055000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1774.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 2.843320E+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 0.045 0.126 92.721 92.725 qs_energies 1 2.0 0.002 0.020 92.001 92.021 ls_scf 1 3.0 0.000 0.001 90.498 90.501 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.693 75.020 multiply_cannon 111 7.7 0.053 0.140 51.940 56.561 ls_scf_main 1 4.0 0.001 0.010 56.273 56.277 multiply_cannon_loop 111 8.7 0.116 0.130 49.181 52.666 density_matrix_trs4 2 5.0 0.003 0.014 50.332 50.477 mp_waitall_1 7281 11.0 23.489 32.658 23.489 32.658 ls_scf_init_scf 1 4.0 0.009 0.072 30.514 30.517 ls_scf_init_matrix_S 1 5.0 0.000 0.001 29.015 29.111 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.569 26.579 multiply_cannon_multrec 888 9.7 12.617 15.107 21.247 24.341 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 11.239 22.152 make_m2s 222 7.7 0.006 0.007 16.462 17.161 make_images 222 8.7 1.585 1.854 16.424 17.123 make_images_data 222 9.7 0.004 0.004 9.584 10.622 hybrid_alltoall_any 227 10.6 0.641 2.962 9.034 10.277 dbcsr_mm_accdrv_process 3754 10.4 0.298 0.490 8.160 9.443 dbcsr_mm_accdrv_process_sort 3754 11.4 7.726 8.952 7.726 8.952 mp_sum_l 887 5.1 4.646 7.808 4.646 7.808 multiply_cannon_sync_h2d 888 9.7 6.116 7.501 6.116 7.501 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.471 7.088 mp_irecv_dv 2335 11.1 2.454 7.023 2.454 7.023 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.308 6.272 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.452 6.115 arnoldi_extremal 4 6.8 0.000 0.000 5.167 5.187 arnoldi_normal_ev 4 7.8 0.001 0.005 5.167 5.187 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.790 5.002 build_subspace 16 8.4 0.014 0.020 4.856 4.866 calculate_norms 1584 9.8 4.338 4.824 4.338 4.824 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.501 3.837 mp_allgather_i34 111 8.7 0.861 3.783 0.861 3.783 ls_scf_post 1 4.0 0.001 0.005 3.711 3.716 dbcsr_matrix_vector_mult_local 304 10.0 3.054 3.624 3.056 3.626 ls_scf_store_result 1 5.0 0.000 0.000 3.441 3.534 ls_scf_dm_to_ks 2 5.0 0.000 0.006 2.858 2.983 dbcsr_data_new 4116 9.9 2.093 2.438 2.093 2.438 make_images_sizes 222 9.7 0.000 0.000 1.040 2.185 mp_alltoall_i44 222 10.7 1.040 2.184 1.040 2.184 dbcsr_sort_data 325 11.1 1.883 2.172 1.883 2.172 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.078 2.080 rebuild_ks_matrix 3 7.3 0.000 0.000 2.059 2.061 qs_ks_build_kohn_sham_matrix 3 8.3 0.002 0.014 2.059 2.061 make_images_pack 222 9.7 1.623 1.880 1.626 1.883 dbcsr_finalize 304 7.8 0.026 0.032 1.603 1.868 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.725000, yerr=0.000000 PlotPoint: name="505", plot="h2o_32_nrep3_ls_mem", label="(8n/4r/3t)", y=2246.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 3.386434E+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 0.067 0.106 97.465 97.466 qs_energies 1 2.0 0.000 0.000 96.303 96.309 ls_scf 1 3.0 0.000 0.000 94.652 94.657 dbcsr_multiply_generic 111 6.7 0.017 0.017 78.337 78.632 ls_scf_main 1 4.0 0.000 0.000 58.783 58.785 multiply_cannon 111 7.7 0.052 0.126 51.636 56.392 density_matrix_trs4 2 5.0 0.002 0.003 52.636 52.740 multiply_cannon_loop 111 8.7 0.152 0.168 46.571 49.727 ls_scf_init_scf 1 4.0 0.000 0.000 32.627 32.631 ls_scf_init_matrix_S 1 5.0 0.000 0.000 31.420 31.492 mp_waitall_1 6369 11.0 22.583 29.334 22.583 29.334 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 28.966 28.980 multiply_cannon_multrec 1332 9.7 14.159 17.033 22.133 25.069 make_m2s 222 7.7 0.006 0.008 21.084 22.552 make_images 222 8.7 3.140 3.620 21.034 22.504 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.163 17.035 make_images_data 222 9.7 0.004 0.004 11.784 13.379 hybrid_alltoall_any 227 10.6 0.796 3.760 11.138 12.896 dbcsr_mm_accdrv_process 3641 10.4 0.311 0.482 7.604 9.112 dbcsr_mm_accdrv_process_sort 3641 11.4 7.164 8.615 7.164 8.615 mp_sum_l 887 5.1 4.033 7.327 4.033 7.327 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.087 6.182 mp_irecv_dv 3229 10.9 2.062 6.101 2.062 6.101 multiply_cannon_sync_h2d 1332 9.7 5.443 6.047 5.443 6.047 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.124 5.740 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.465 5.316 arnoldi_extremal 4 6.8 0.000 0.000 5.277 5.297 arnoldi_normal_ev 4 7.8 0.001 0.005 5.277 5.297 build_subspace 16 8.4 0.014 0.021 4.937 4.945 mp_allgather_i34 111 8.7 2.199 4.790 2.199 4.790 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.558 4.698 calculate_norms 2376 9.8 4.183 4.491 4.183 4.491 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.642 3.925 dbcsr_matrix_vector_mult_local 304 10.0 3.228 3.729 3.230 3.730 dbcsr_sort_data 658 11.4 3.122 3.429 3.122 3.429 ls_scf_post 1 4.0 0.000 0.000 3.241 3.246 dbcsr_special_finalize 555 9.7 0.006 0.007 2.873 3.224 dbcsr_merge_single_wm 555 10.7 0.541 0.673 2.864 3.215 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.009 3.069 ls_scf_store_result 1 5.0 0.000 0.000 2.976 3.046 dbcsr_data_release 10477 10.7 1.580 2.430 1.580 2.430 dbcsr_finalize 304 7.8 0.049 0.061 1.802 1.975 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=97.466000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2728.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 4.733174E+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 0.107 0.128 92.798 92.800 qs_energies 1 2.0 0.000 0.000 91.382 91.387 ls_scf 1 3.0 0.000 0.000 89.445 89.450 dbcsr_multiply_generic 111 6.7 0.017 0.018 70.948 71.093 ls_scf_main 1 4.0 0.000 0.000 56.576 56.577 multiply_cannon 111 7.7 0.088 0.160 52.561 56.222 multiply_cannon_loop 111 8.7 0.089 0.099 49.968 51.231 density_matrix_trs4 2 5.0 0.002 0.003 49.626 49.711 ls_scf_init_scf 1 4.0 0.000 0.000 29.448 29.449 mp_waitall_1 5436 11.0 24.392 28.773 24.392 28.773 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.222 28.250 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.001 26.085 26.096 multiply_cannon_multrec 444 9.7 13.751 16.160 20.879 23.081 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.408 15.470 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.463 15.005 make_m2s 222 7.7 0.005 0.005 13.618 14.564 make_images 222 8.7 2.042 2.476 13.551 14.494 hybrid_alltoall_any 227 10.6 0.802 3.842 8.152 9.839 make_images_data 222 9.7 0.003 0.004 8.336 9.595 multiply_cannon_sync_h2d 444 9.7 6.730 8.129 6.730 8.129 dbcsr_mm_accdrv_process 3003 10.4 0.365 0.410 6.822 7.928 dbcsr_mm_accdrv_process_sort 3003 11.4 6.458 7.535 6.458 7.535 arnoldi_extremal 4 6.8 0.000 0.000 5.913 5.924 arnoldi_normal_ev 4 7.8 0.002 0.005 5.913 5.924 build_subspace 16 8.4 0.015 0.019 5.517 5.526 mp_sum_l 887 5.1 2.769 4.912 2.769 4.912 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.460 4.662 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.231 4.447 dbcsr_matrix_vector_mult_local 304 10.0 3.739 4.203 3.741 4.205 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.533 3.887 mp_irecv_dv 1241 11.2 1.519 3.852 1.519 3.852 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.944 3.816 calculate_norms 792 9.8 3.609 3.760 3.609 3.760 mp_allgather_i34 111 8.7 1.159 3.664 1.159 3.664 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.510 3.567 ls_scf_post 1 4.0 0.000 0.000 3.421 3.426 make_images_sizes 222 9.7 0.000 0.000 0.879 3.336 mp_alltoall_i44 222 10.7 0.878 3.336 0.878 3.336 ls_scf_store_result 1 5.0 0.000 0.000 3.215 3.245 dbcsr_finalize 304 7.8 0.062 0.077 2.190 2.283 dbcsr_data_new 4608 9.7 1.793 2.267 1.793 2.267 dbcsr_merge_all 275 8.9 0.479 0.520 2.049 2.129 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.026 2.027 rebuild_ks_matrix 3 7.3 0.000 0.000 1.994 1.995 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.994 1.995 qs_energies_init_hamiltonians 1 3.0 0.000 0.000 1.921 1.921 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.800000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3736.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 8.879555E+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 0.093 0.129 105.938 105.949 qs_energies 1 2.0 0.000 0.000 104.581 104.584 ls_scf 1 3.0 0.000 0.000 101.629 101.633 dbcsr_multiply_generic 111 6.7 0.024 0.026 75.222 75.343 ls_scf_main 1 4.0 0.000 0.000 64.097 64.101 density_matrix_trs4 2 5.0 0.002 0.003 55.125 55.203 multiply_cannon 111 7.7 0.149 0.201 48.891 51.146 multiply_cannon_loop 111 8.7 0.098 0.100 45.886 46.253 ls_scf_init_scf 1 4.0 0.000 0.000 33.730 33.732 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.214 32.232 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.492 29.500 mp_waitall_1 4527 11.1 22.381 26.192 22.381 26.192 make_m2s 222 7.7 0.005 0.005 22.724 23.812 make_images 222 8.7 3.574 3.873 22.617 23.704 multiply_cannon_multrec 444 9.7 17.817 18.392 22.488 23.103 hybrid_alltoall_any 227 10.6 1.651 3.613 12.821 15.813 make_images_data 222 9.7 0.003 0.004 13.100 15.245 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.684 11.116 multiply_cannon_sync_h2d 444 9.7 8.789 8.843 8.789 8.843 arnoldi_extremal 4 6.8 0.000 0.000 7.496 7.511 arnoldi_normal_ev 4 7.8 0.003 0.009 7.496 7.511 build_subspace 16 8.4 0.026 0.036 6.939 6.949 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.553 5.702 dbcsr_matrix_vector_mult_local 304 10.0 5.103 5.391 5.105 5.394 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.086 5.178 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.891 5.135 dbcsr_mm_accdrv_process 1814 10.4 0.265 0.341 4.487 4.615 dbcsr_mm_accdrv_process_sort 1814 11.4 4.154 4.292 4.154 4.292 ls_scf_post 1 4.0 0.000 0.000 3.802 3.805 make_images_sizes 222 9.7 0.000 0.000 1.481 3.735 mp_alltoall_i44 222 10.7 1.480 3.735 1.480 3.735 mp_allgather_i34 111 8.7 1.102 3.600 1.102 3.600 ls_scf_store_result 1 5.0 0.000 0.000 3.503 3.540 calculate_norms 792 9.8 3.234 3.274 3.234 3.274 dbcsr_finalize 304 7.8 0.083 0.089 3.083 3.178 dbcsr_merge_all 275 8.9 0.888 0.929 2.858 2.945 qs_energies_init_hamiltonians 1 3.0 0.000 0.001 2.920 2.920 dbcsr_complete_redistribute 5 7.6 1.432 1.476 2.744 2.871 matrix_ls_to_qs 2 6.0 0.000 0.000 2.402 2.535 dbcsr_sort_data 325 11.1 2.437 2.485 2.437 2.485 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.401 2.403 rebuild_ks_matrix 3 7.3 0.000 0.000 2.335 2.337 qs_ks_build_kohn_sham_matrix 3 8.3 0.006 0.007 2.335 2.337 dbcsr_data_new 6591 9.6 1.885 2.319 1.885 2.319 dbcsr_new_transposed 4 7.5 0.241 0.252 2.287 2.297 dbcsr_frobenius_norm 74 6.6 2.059 2.140 2.192 2.226 dbcsr_add_d 103 6.2 0.000 0.000 2.138 2.223 dbcsr_add_anytype 103 7.2 0.859 0.891 2.138 2.222 dbcsr_data_release 12724 10.6 1.981 2.171 1.981 2.171 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=105.949000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=7057.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/b1e573e5365af53cd5b64772cda048b47a415cd7_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 593.457152E+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 4805178. 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 0.025 0.092 232.051 232.052 qs_mol_dyn_low 1 2.0 0.003 0.014 230.986 231.033 qs_forces 5 3.8 0.005 0.041 230.889 230.893 qs_energies 5 4.8 0.003 0.064 227.907 227.951 scf_env_do_scf 5 5.8 0.001 0.063 212.243 212.246 scf_env_do_scf_inner_loop 105 6.6 0.004 0.075 185.313 185.329 qs_scf_new_mos 105 7.6 0.000 0.001 144.571 144.747 qs_scf_loop_do_ot 105 8.6 0.001 0.001 144.570 144.747 dbcsr_multiply_generic 1445 12.2 0.127 0.136 135.176 135.588 ot_scf_mini 105 9.6 0.003 0.016 134.457 134.632 multiply_cannon 1445 13.2 0.275 0.291 115.797 117.605 multiply_cannon_loop 1445 14.2 2.686 2.828 114.153 115.522 velocity_verlet 4 3.0 0.002 0.008 107.504 107.505 ot_mini 105 10.6 0.001 0.035 61.090 61.204 mp_waitall_1 488190 16.1 35.661 45.194 35.661 45.194 multiply_cannon_multrec 69360 15.2 30.188 35.097 39.487 44.599 qs_ot_get_p 112 10.4 0.001 0.003 43.291 43.557 qs_ot_get_derivative 55 11.6 0.001 0.001 39.310 39.453 multiply_cannon_metrocomm3 69360 15.2 0.206 0.218 26.379 36.472 multiply_cannon_sync_h2d 69360 15.2 29.847 34.052 29.847 34.052 qs_ot_p2m_diag 40 11.0 0.020 0.031 31.991 32.102 rebuild_ks_matrix 110 8.4 0.000 0.000 29.111 29.276 qs_ks_build_kohn_sham_matrix 110 9.4 0.013 0.093 29.110 29.275 cp_dbcsr_syevd 40 12.0 0.002 0.002 28.744 28.745 qs_ks_update_qs_env 112 7.6 0.001 0.001 26.835 26.978 init_scf_loop 7 6.6 0.002 0.095 26.896 26.898 cp_fm_syevd 40 13.0 0.000 0.000 23.587 23.721 apply_preconditioner_dbcsr 62 12.6 0.000 0.001 23.018 23.237 apply_single 62 13.6 0.000 0.000 23.017 23.237 prepare_preconditioner 7 7.6 0.000 0.001 21.807 21.841 make_preconditioner 7 8.6 0.001 0.051 21.807 21.841 ot_new_cg_direction 55 11.6 0.001 0.003 21.009 21.010 cp_fm_redistribute_end 40 14.0 9.431 18.820 9.436 18.821 cp_fm_syevd_base 40 14.0 9.378 18.768 9.378 18.768 qs_rho_update_rho_low 110 7.6 0.001 0.022 17.567 17.967 calculate_rho_elec 110 8.6 0.030 0.033 17.566 17.967 qs_ot_get_orbitals 105 10.6 0.001 0.001 14.833 15.001 make_full_inverse_cholesky 7 9.6 0.000 0.000 14.670 14.737 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 14.247 14.355 mp_sum_l 4764 12.2 12.557 13.761 12.557 13.761 init_scf_run 5 5.8 0.000 0.001 12.886 12.887 scf_env_initial_rho_setup 5 6.8 0.001 0.013 12.886 12.887 density_rs2pw 110 9.6 0.006 0.016 11.721 12.166 pw_transfer 1645 12.4 0.078 0.100 11.820 12.020 calculate_dm_sparse 110 9.5 0.000 0.001 11.711 11.928 fft_wrap_pw1pw2 1425 13.5 0.012 0.015 11.682 11.884 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 10.612 10.704 dbcsr_mm_accdrv_process 154766 15.8 5.620 5.976 9.165 10.633 qs_vxc_create 110 10.4 0.003 0.027 10.467 10.501 fft_wrap_pw1pw2_240 915 15.0 1.150 1.221 10.276 10.484 cp_fm_cholesky_invert 7 10.6 10.320 10.327 10.320 10.327 check_diag 80 13.5 8.687 8.936 9.555 9.684 transfer_rs2pw 445 10.6 0.007 0.008 8.335 8.783 fft3d_pb 915 16.0 2.394 2.612 8.471 8.685 sum_up_and_integrate 60 10.3 0.001 0.010 8.564 8.573 integrate_v_rspace 60 11.3 0.002 0.028 8.546 8.557 acc_transpose_blocks 69360 15.2 0.319 0.351 7.796 8.448 multiply_cannon_metrocomm1 69360 15.2 0.097 0.104 4.827 8.447 cp_dbcsr_sm_fm_multiply 15 9.3 0.001 0.001 8.032 8.050 calculate_first_density_matrix 1 7.0 0.001 0.021 7.944 7.960 cp_dbcsr_sm_fm_multiply_core 15 10.3 0.000 0.000 7.614 7.670 xc_rho_set_and_dset_create 110 12.4 0.077 0.099 7.388 7.649 xc_vxc_pw_create 60 11.3 0.039 0.068 6.997 7.031 make_m2s 2890 13.2 0.079 0.089 6.332 6.916 make_full_single_inverse 7 9.6 0.001 0.002 6.841 6.876 make_images 2890 14.2 0.239 0.260 6.225 6.809 xc_pw_derive 510 13.4 0.005 0.007 6.233 6.295 mp_alltoall_z22v 2340 17.7 5.483 5.804 5.483 5.804 acc_transpose_blocks_kernels 69360 16.2 0.790 0.845 5.336 5.796 mp_waitany 7680 13.5 4.500 5.030 4.500 5.030 jit_kernel_transpose 5 15.0 4.546 4.970 4.546 4.970 potential_pw2rs 60 12.3 0.003 0.003 4.749 4.773 multiply_cannon_metrocomm4 67915 15.2 0.190 0.207 2.013 4.656 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=232.052000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=563.000000, yerr=3.224903 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: b1e573e5365af53cd5b64772cda048b47a415cd7 Summary: empty Status: OK