=== 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: d1376dc4969b25e1013bb42635e3ab0fe8739442 ################# ARCHITECTURE FILE ################## #!/bin/bash # # CP2K arch file for Cray-XC50 (Piz Daint, CSCS, GPU partition) # # Tested with: GNU 9.3.0, Cray-MPICH 7.7.18, Cray-libsci 20.09.1, Cray-FFTW 3.3.8.10, # COSMA 2.6.6, ELPA 2023.05.001, LIBINT 2.6.0, LIBPEXSI 1.2.0, # LIBXC 6.2.2, LIBVORI 220621, LIBXSMM 1.17, PLUMED 2.8.2, # SIRIUS 7.4.3, SPGLIB 1.16.2 # # Usage: Source this arch file and then run make as instructed. # A full toolchain installation is performed as default. # Replace or adapt the "module add" commands below if needed. # # Last update: 21.06.2023 # # \ if [ "${0}" = "${BASH_SOURCE}" ]; then \ echo "ERROR: Script ${0##*/} must be sourced"; \ echo "Usage: source ${0##*/}"; \ exit 1; \ fi; \ this_file=${BASH_SOURCE##*/}; \ if [ -n "${1}" ]; then \ gcc_version="${1}"; \ else \ gcc_version="9.3.0"; \ fi; \ module add daint-gpu; \ module rm PrgEnv-cray; \ module add PrgEnv-gnu; \ module rm gcc; \ module add gcc/${gcc_version}; \ module add cray-fftw/3.3.8.10; \ module add cudatoolkit; \ echo "Expected setup:"; \ echo " cray-mpich/7.7.18"; \ echo " craype-haswell"; \ echo " daint-gpu/21.09"; \ echo " craype/2.7.10"; \ echo " cray-libsci/20.09.1"; \ echo " PrgEnv-gnu/6.0.10"; \ echo " gcc/${gcc_version}"; \ echo " cray-fftw/3.3.8.10"; \ echo " cudatoolkit/11.0.2_3.38-8.1__g5b73779"; \ module list; \ module -f save cp2k_gpu_gnu_psmp; \ echo "To load the required modules in your batch job script, use:"; \ echo " module restore cp2k_gpu_gnu_psmp"; \ cd tools/toolchain; \ ./install_cp2k_toolchain.sh --enable-cuda=yes --gpu-ver=P100 -j${maxtasks} --no-arch-files --with-gcc=system --with-libvdwxc --with-pexsi --with-plumed; \ cd ../..; \ printf "Sourcing ${PWD}/tools/toolchain/install/setup ... "; \ source ${PWD}/tools/toolchain/install/setup; \ printf "done\n"; \ echo "Check the output above for error messages and consistency!"; \ echo; \ echo "If everything is OK, you can build a CP2K production binary with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.}"; \ echo; \ echo "Alternatively, you can add further checks, e.g. for regression testing, with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} DO_CHECKS=yes"; \ echo "or build CP2K as a library with"; \ echo " make -j ARCH=${this_file%.*} VERSION=${this_file##*.} libcp2k"; \ echo; \ return # Set options DO_CHECKS := no USE_ACC := yes USE_COSMA := 2.6.6 USE_ELPA := 2023.05.001 USE_LIBINT := 2.6.0 USE_LIBPEXSI := 1.2.0 USE_LIBVORI := 220621 USE_LIBXC := 6.2.2 USE_LIBXSMM := 1.17 USE_PLUMED := 2.8.2 #USE_QUIP := 0.9.10 USE_SIRIUS := 7.4.3 USE_SPGLIB := 1.16.2 # Only needed for SIRIUS LIBVDWXC_VER := 0.4.0 SPFFT_VER := 1.0.6 SPLA_VER := 1.5.5 HDF5_VER := 1.12.0 # Only needed for LIBPEXSI SCOTCH_VER := 6.0.0 SUPERLU_VER := 6.1.0 LMAX := 5 MAX_CONTR := 4 GPUVER := P100 OFFLOAD_TARGET := cuda CC := cc CXX := CC OFFLOAD_CC := nvcc FC := ftn LD := ftn AR := ar -r # cc, CC, and ftn include already the proper -march flag CFLAGS := -O2 -fopenmp -fopenmp-simd -ftree-vectorize -funroll-loops -g DFLAGS := -D__parallel DFLAGS += -D__SCALAPACK DFLAGS += -D__FFTW3 DFLAGS += -D__MAX_CONTR=$(strip $(MAX_CONTR)) INSTALL_PATH := $(PWD)/tools/toolchain/install ifeq ($(DO_CHECKS), yes) DFLAGS += -D__CHECK_DIAG endif ifeq ($(USE_ACC), yes) DFLAGS += -D__DBCSR_ACC DFLAGS += -D__OFFLOAD_CUDA # Possibly no performance gain with PW_CUDA currently DFLAGS += -D__NO_OFFLOAD_PW endif ifneq ($(USE_PLUMED),) USE_PLUMED := $(strip $(USE_PLUMED)) PLUMED_LIB := $(INSTALL_PATH)/plumed-$(USE_PLUMED)/lib DFLAGS += -D__PLUMED2 USE_GSL := 2.7 LIBS += $(PLUMED_LIB)/libplumed.a endif ifneq ($(USE_ELPA),) USE_ELPA := $(strip $(USE_ELPA)) TARGET := nvidia ELPA_INC := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/include/elpa-$(USE_ELPA) ELPA_LIB := $(INSTALL_PATH)/elpa-$(USE_ELPA)/$(TARGET)/lib CFLAGS += -I$(ELPA_INC)/elpa -I$(ELPA_INC)/modules DFLAGS += -D__ELPA ifeq ($(TARGET), nvidia) DFLAGS += -D__ELPA_NVIDIA_GPU endif LIBS += $(ELPA_LIB)/libelpa.a endif ifneq ($(USE_QUIP),) USE_QUIP := $(strip $(USE_QUIP)) QUIP_INC := $(INSTALL_PATH)/quip-$(USE_QUIP)/include QUIP_LIB := $(INSTALL_PATH)/quip-$(USE_QUIP)/lib CFLAGS += -I$(QUIP_INC) DFLAGS += -D__QUIP LIBS += $(QUIP_LIB)/libquip_core.a LIBS += $(QUIP_LIB)/libatoms.a LIBS += $(QUIP_LIB)/libFoX_sax.a LIBS += $(QUIP_LIB)/libFoX_common.a LIBS += $(QUIP_LIB)/libFoX_utils.a LIBS += $(QUIP_LIB)/libFoX_fsys.a endif ifneq ($(USE_LIBPEXSI),) USE_LIBPEXSI := $(strip $(USE_LIBPEXSI)) SCOTCH_VER := $(strip $(SCOTCH_VER)) SUPERLU_VER := $(strip $(SUPERLU_VER)) LIBPEXSI_INC := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/include LIBPEXSI_LIB := $(INSTALL_PATH)/pexsi-$(USE_LIBPEXSI)/lib SCOTCH_INC := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/include SCOTCH_LIB := $(INSTALL_PATH)/scotch-$(SCOTCH_VER)/lib SUPERLU_INC := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/include SUPERLU_LIB := $(INSTALL_PATH)/superlu_dist-$(SUPERLU_VER)/lib CFLAGS += -I$(LIBPEXSI_INC) -I$(SCOTCH_INC) -I$(SUPERLU_INC) DFLAGS += -D__LIBPEXSI LIBS += $(LIBPEXSI_LIB)/libpexsi.a LIBS += $(SUPERLU_LIB)/libsuperlu_dist.a LIBS += $(SCOTCH_LIB)/libptscotchparmetis.a LIBS += $(SCOTCH_LIB)/libptscotch.a LIBS += $(SCOTCH_LIB)/libptscotcherr.a LIBS += $(SCOTCH_LIB)/libscotchmetis.a LIBS += $(SCOTCH_LIB)/libscotch.a endif ifneq ($(USE_LIBVORI),) USE_LIBVORI := $(strip $(USE_LIBVORI)) LIBVORI_LIB := $(INSTALL_PATH)/libvori-$(USE_LIBVORI)/lib DFLAGS += -D__LIBVORI LIBS += $(LIBVORI_LIB)/libvori.a endif ifneq ($(USE_LIBXC),) USE_LIBXC := $(strip $(USE_LIBXC)) LIBXC_INC := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/include LIBXC_LIB := $(INSTALL_PATH)/libxc-$(USE_LIBXC)/lib CFLAGS += -I$(LIBXC_INC) DFLAGS += -D__LIBXC LIBS += $(LIBXC_LIB)/libxcf03.a LIBS += $(LIBXC_LIB)/libxc.a endif ifneq ($(USE_LIBINT),) USE_LIBINT := $(strip $(USE_LIBINT)) LMAX := $(strip $(LMAX)) LIBINT_INC := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/include LIBINT_LIB := $(INSTALL_PATH)/libint-v$(USE_LIBINT)-cp2k-lmax-$(LMAX)/lib CFLAGS += -I$(LIBINT_INC) DFLAGS += -D__LIBINT LIBS += $(LIBINT_LIB)/libint2.a endif ifneq ($(USE_SPGLIB),) USE_SPGLIB := $(strip $(USE_SPGLIB)) SPGLIB_INC := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/include SPGLIB_LIB := $(INSTALL_PATH)/spglib-$(USE_SPGLIB)/lib CFLAGS += -I$(SPGLIB_INC) DFLAGS += -D__SPGLIB LIBS += $(SPGLIB_LIB)/libsymspg.a endif ifneq ($(USE_LIBXSMM),) USE_LIBXSMM := $(strip $(USE_LIBXSMM)) LIBXSMM_INC := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/include LIBXSMM_LIB := $(INSTALL_PATH)/libxsmm-$(USE_LIBXSMM)/lib CFLAGS += -I$(LIBXSMM_INC) DFLAGS += -D__LIBXSMM LIBS += $(LIBXSMM_LIB)/libxsmmf.a LIBS += $(LIBXSMM_LIB)/libxsmm.a endif ifneq ($(USE_SIRIUS),) USE_SIRIUS := $(strip $(USE_SIRIUS)) HDF5_VER := $(strip $(HDF5_VER)) HDF5_LIB := $(INSTALL_PATH)/hdf5-$(HDF5_VER)/lib LIBVDWXC_VER := $(strip $(LIBVDWXC_VER)) LIBVDWXC_INC := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/include LIBVDWXC_LIB := $(INSTALL_PATH)/libvdwxc-$(LIBVDWXC_VER)/lib SPFFT_VER := $(strip $(SPFFT_VER)) SPFFT_INC := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/include SPLA_VER := $(strip $(SPLA_VER)) SPLA_INC := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/include/spla ifeq ($(USE_ACC), yes) DFLAGS += -D__OFFLOAD_GEMM SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib/cuda SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib/cuda SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include/cuda SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib/cuda else SPFFT_LIB := $(INSTALL_PATH)/SpFFT-$(SPFFT_VER)/lib SPLA_LIB := $(INSTALL_PATH)/SpLA-$(SPLA_VER)/lib SIRIUS_INC := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/include SIRIUS_LIB := $(INSTALL_PATH)/sirius-$(USE_SIRIUS)/lib endif CFLAGS += -I$(LIBVDWXC_INC) CFLAGS += -I$(SPFFT_INC) CFLAGS += -I$(SPLA_INC) CFLAGS += -I$(SIRIUS_INC) DFLAGS += -D__HDF5 DFLAGS += -D__LIBVDWXC DFLAGS += -D__SPFFT DFLAGS += -D__SPLA DFLAGS += -D__SIRIUS LIBS += $(SIRIUS_LIB)/libsirius.a LIBS += $(SPLA_LIB)/libspla.a LIBS += $(SPFFT_LIB)/libspfft.a LIBS += $(LIBVDWXC_LIB)/libvdwxc.a LIBS += $(HDF5_LIB)/libhdf5.a endif ifneq ($(USE_COSMA),) USE_COSMA := $(strip $(USE_COSMA)) ifeq ($(USE_ACC), yes) USE_COSMA := $(USE_COSMA)-cuda endif COSMA_INC := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/include COSMA_LIB := $(INSTALL_PATH)/COSMA-$(USE_COSMA)/lib CFLAGS += -I$(COSMA_INC) DFLAGS += -D__COSMA LIBS += $(COSMA_LIB)/libcosma_prefixed_pxgemm.a LIBS += $(COSMA_LIB)/libcosma.a LIBS += $(COSMA_LIB)/libcosta.a LIBS += $(COSMA_LIB)/libTiled-MM.a endif ifneq ($(USE_GSL),) USE_GSL := $(strip $(USE_GSL)) GSL_INC := $(INSTALL_PATH)/gsl-$(USE_GSL)/include GSL_LIB := $(INSTALL_PATH)/gsl-$(USE_GSL)/lib CFLAGS += -I$(GSL_INC) DFLAGS += -D__GSL LIBS += $(GSL_LIB)/libgsl.a endif CFLAGS += $(DFLAGS) CXXFLAGS := $(CFLAGS) -std=c++11 OFFLOAD_FLAGS := $(DFLAGS) -O3 -Xcompiler="-fopenmp" -arch sm_60 --std=c++11 FCFLAGS := $(CFLAGS) ifeq ($(shell [ $(shell gcc -dumpversion | cut -d. -f1) -gt 9 ] && echo yes), yes) FCFLAGS += -fallow-argument-mismatch endif FCFLAGS += -fbacktrace FCFLAGS += -ffree-form FCFLAGS += -ffree-line-length-none FCFLAGS += -fno-omit-frame-pointer FCFLAGS += -std=f2008 ifneq ($(CUDA_HOME),) CUDA_LIB := $(CUDA_HOME)/lib64 LDFLAGS := $(FCFLAGS) -L$(CUDA_LIB) -Wl,-rpath=$(CUDA_LIB) else LDFLAGS := $(FCFLAGS) endif LIBS += -lcusolver -lcudart -lnvrtc -lcuda -lcufft -lcublas -lrt LIBS += -lz -ldl -lpthread -lstdc++ # End ############### END ARCHITECTURE FILE ################ ===== TESTS (description) ===== ~~~~~~~~~ TEST ~~~~~~~~~ description: H2O-32 RI-RPA/RI-MP2 correlation energy input file: benchmarks/QS_mp2_rpa/32-H2O/RI-RPA.inp required files: ['benchmarks/QS_mp2_rpa/32-H2O/BASIS_H2O', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32.xyz', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-PBE-TZ.inp', 'benchmarks/QS_mp2_rpa/32-H2O/H2O-32-RI-dRPA-TZ.inp'] output file: result.log # nodes = 8 # ranks/node = 2 # threads/rank = 6 nrepeat = 1 time[min] = 15 run dir: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/01 job id: 48255670 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/02 job id: 48255671 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/03 job id: 48255672 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/04 job id: 48255673 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/05 job id: 48255674 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/06 job id: 48255675 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/07 job id: 48255676 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/08 job id: 48255677 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/09 job id: 48255678 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/10 job id: 48255679 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/11 job id: 48255680 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/12 job id: 48255681 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/13 job id: 48255682 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/14 job id: 48255683 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/15 job id: 48255684 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/16 job id: 48255685 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/17 job id: 48255686 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/18 job id: 48255687 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/19 job id: 48255688 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/20 job id: 48255689 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/21 job id: 48255690 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/22 job id: 48255691 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/23 job id: 48255692 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/24 job id: 48255693 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/25 job id: 48255694 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/26 job id: 48255695 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/27 job id: 48255696 --- 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/01/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 0.000000E+00 0.0% 0.0% 0.0% flops max/rank 0.000000E+00 0.0% 0.0% 0.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 0 0.0% 0.0% 0.0% number of processed stacks 0 0.0% 0.0% 0.0% average stack size 0.0 0.0 0.0 marketing flops 0.000000E+00 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 1 12. MP_Allreduce 19 21. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 15 177869. MP_Allreduce 424 8. MP_Sync 3 MP_comm_split 1 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.024 0.036 135.615 135.616 farming_run 1 2.0 135.110 135.111 135.583 135.587 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.498251E+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 0.007 0.025 115.531 115.531 qs_energies 1 2.0 0.000 0.000 115.321 115.323 mp2_main 1 3.0 0.000 0.000 112.907 112.909 mp2_gpw_main 1 4.0 0.028 0.034 111.748 111.750 mp2_ri_gpw_compute_in 1 5.0 0.171 0.174 92.865 93.031 mp2_ri_gpw_compute_in_loop 1 6.0 0.004 0.005 54.851 55.016 mp2_eri_3c_integrate_gpw 272 7.0 0.151 0.166 41.238 46.075 get_2c_integrals 1 6.0 0.008 0.010 37.261 37.844 integrate_v_rspace 273 8.0 0.439 0.454 24.631 29.133 pw_transfer 6555 10.6 0.373 0.398 27.474 27.759 fft_wrap_pw1pw2 5465 11.4 0.045 0.048 26.029 26.243 grid_integrate_task_list 273 9.0 20.456 25.383 20.456 25.383 fft_wrap_pw1pw2_100 2178 12.4 1.265 1.420 23.545 23.737 compute_2c_integrals 1 7.0 0.003 0.003 19.794 19.796 compute_2c_integrals_loop_lm 1 8.0 0.003 0.004 19.086 19.496 mp2_eri_2c_integrate_gpw 1 9.0 2.333 2.422 19.083 19.492 rpa_ri_compute_en 1 5.0 0.020 0.022 18.763 18.871 cp_fm_cholesky_decompose 12 8.2 17.479 18.051 17.479 18.051 cholesky_decomp 1 7.0 0.000 0.000 16.311 16.880 fft3d_s 5443 13.4 16.149 16.268 16.172 16.292 ao_to_mo_and_store_B_mult_1 272 7.0 10.771 15.329 10.771 15.329 calculate_wavefunction 272 8.0 5.436 5.520 12.519 13.170 rpa_num_int 1 6.0 0.001 0.008 10.734 10.734 rpa_num_int_RPA_matrix_operati 8 7.0 0.000 0.000 10.716 10.718 calc_mat_Q 8 8.0 0.000 0.000 9.506 9.588 contract_S_to_Q 8 9.0 0.000 0.000 8.926 9.008 calc_potential_gpw 544 9.5 0.005 0.006 8.307 8.758 parallel_gemm_fm 14 9.1 0.000 0.000 8.508 8.608 parallel_gemm_fm_cosma 14 10.1 8.508 8.608 8.508 8.608 mp2_eri_2c_integrate_gpw_pot_l 272 10.0 0.001 0.001 8.336 8.590 potential_pw2rs 545 10.0 0.107 0.109 7.756 8.351 collocate_single_gaussian 272 10.0 0.039 0.042 7.598 7.816 create_integ_mat 1 6.0 0.022 0.028 7.769 7.770 array2fm 1 7.0 0.000 0.000 6.722 7.162 pw_scatter_s 2720 13.7 4.373 4.517 4.373 4.517 pw_gather_s 2722 13.2 3.828 4.139 3.828 4.139 array2fm_buffer_send 1 8.0 2.994 3.174 2.994 3.174 pw_poisson_solve 545 10.5 1.106 1.147 2.261 2.452 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="10", plot="h2o_32_ri_rpa_mp2", label="RI-RPA (8n/2r/6t)", y=111.750176, yerr=0.000000 PlotPoint: name="11", plot="h2o_32_ri_rpa_mp2_mem", label="RI-RPA (8n/2r/6t)", y=2806.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 0.030 0.040 422.200 422.202 farming_run 1 2.0 421.261 421.271 422.159 422.163 ------------------------------------------------------------------------------- @@@@@@@@@@ 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 1.241465E+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 26445167. MP_SendRecv 2876 2171486. MP_ISendRecv 1034 172620. MP_Wait 1346 MP_comm_split 7 MP_ISend 264 362227. MP_IRecv 264 362718. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.013 0.040 217.300 217.301 qs_energies 1 2.0 0.000 0.000 217.002 217.009 scf_env_do_scf 1 3.0 0.000 0.000 114.702 114.703 qs_ks_update_qs_env 5 5.0 0.000 0.000 113.756 113.762 rebuild_ks_matrix 4 6.0 0.000 0.000 113.755 113.761 qs_ks_build_kohn_sham_matrix 4 7.0 0.055 0.061 113.755 113.761 hfx_ks_matrix 4 8.0 0.001 0.001 113.359 113.363 integrate_four_center 4 9.0 0.153 0.466 113.358 113.362 integrate_four_center_main 4 10.0 0.095 0.560 101.941 104.885 integrate_four_center_bin 265 11.0 101.846 104.873 101.846 104.873 mp2_main 1 3.0 0.001 0.008 101.990 101.996 mp2_gpw_main 1 4.0 0.054 0.094 101.072 101.081 init_scf_loop 1 4.0 0.000 0.000 96.633 96.634 mp2_ri_gpw_compute_in 1 5.0 0.064 0.071 74.217 75.288 mp2_ri_gpw_compute_in_loop 1 6.0 0.002 0.003 53.973 55.046 mp2_eri_3c_integrate_gpw 91 7.0 0.144 0.160 41.664 46.512 integrate_v_rspace 95 8.0 0.397 0.564 28.053 32.721 pw_transfer 2240 10.6 0.145 0.162 29.887 30.275 fft_wrap_pw1pw2 1868 11.4 0.018 0.020 28.926 29.319 ao_to_mo_and_store_B_mult_1 91 7.0 10.629 29.079 10.629 29.079 mp2_ri_gpw_compute_en 1 5.0 0.056 0.065 26.642 28.770 grid_integrate_task_list 95 9.0 23.341 28.214 23.341 28.214 fft_wrap_pw1pw2_100 730 12.4 1.238 1.380 26.630 27.090 mp2_ri_gpw_compute_en_RI_loop 1 6.0 1.835 1.877 24.906 24.918 get_2c_integrals 1 6.0 0.000 0.000 20.163 20.183 compute_2c_integrals 1 7.0 0.005 0.014 19.138 19.147 compute_2c_integrals_loop_lm 1 8.0 0.003 0.009 18.764 19.002 mp2_eri_2c_integrate_gpw 1 9.0 1.748 1.806 18.761 18.993 fft3d_s 1823 13.4 18.419 18.908 18.433 18.921 scf_env_do_scf_inner_loop 4 4.0 0.000 0.001 18.067 18.067 calculate_wavefunction 91 8.0 2.017 2.043 9.741 9.942 mp2_ri_gpw_compute_en_expansio 172 7.0 0.558 0.585 8.768 9.220 potential_pw2rs 186 10.0 0.033 0.034 8.621 9.121 local_gemm 172 8.0 8.210 8.665 8.210 8.665 mp2_eri_2c_integrate_gpw_pot_l 91 10.0 0.001 0.001 8.181 8.469 mp2_ri_gpw_compute_en_comm 22 7.0 0.498 0.516 7.919 8.410 collocate_single_gaussian 91 10.0 0.018 0.061 7.842 8.100 calc_potential_gpw 182 9.5 0.002 0.002 7.871 8.077 integrate_four_center_load 4 10.0 0.000 0.000 6.783 6.787 hfx_load_balance 1 11.0 0.000 0.000 6.783 6.787 mp_sendrecv_dm3 2068 8.0 5.956 6.450 5.956 6.450 mp2_ri_gpw_compute_en_ener 172 7.0 6.340 6.405 6.340 6.405 mp_sync 37 10.5 3.355 6.143 3.355 6.143 pw_gather_s 912 13.2 4.942 5.331 4.942 5.331 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="20", plot="h2o_32_ri_rpa_mp2", label="RI-MP2 (8n/6r/2t)", y=101.072515, yerr=0.000000 PlotPoint: name="21", plot="h2o_32_ri_rpa_mp2_mem", label="RI-MP2 (8n/6r/2t)", y=1501.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 452.132864E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 9436608 MPI messages size (bytes): total size 333.233553E+09 min size 0.000000E+00 max size 315.840000E+03 average size 35.312852E+03 MPI breakdown and total messages size (bytes): size <= 128 4913240 0 128 < size <= 8192 1155432 9465298944 8192 < size <= 32768 1984512 54190407680 32768 < size <= 131072 551296 42776657920 131072 < size <= 4194304 832128 226802306368 4194304 < size <= 16777216 0 0 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3683 62379. MP_Allreduce 10329 270. MP_Sync 530 MP_Alltoall 2083 MP_SendRecv 22610 5520. MP_ISendRecv 22610 5520. MP_Wait 37876 MP_comm_split 50 MP_ISend 20771 42672. MP_IRecv 20771 42672. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.020 0.035 61.873 61.875 qs_mol_dyn_low 1 2.0 0.003 0.003 61.602 61.612 qs_forces 11 3.9 0.003 0.003 61.529 61.538 qs_energies 11 4.9 0.001 0.001 59.758 59.773 scf_env_do_scf 11 5.9 0.001 0.002 52.605 52.606 scf_env_do_scf_inner_loop 108 6.5 0.002 0.007 49.225 49.226 qs_scf_new_mos 108 7.5 0.000 0.001 34.942 35.241 qs_scf_loop_do_ot 108 8.5 0.000 0.001 34.942 35.240 dbcsr_multiply_generic 2286 12.5 0.093 0.098 33.624 33.993 ot_scf_mini 108 9.5 0.002 0.002 33.277 33.435 velocity_verlet 10 3.0 0.001 0.002 30.076 30.076 multiply_cannon 2286 13.5 0.193 0.207 25.934 27.306 multiply_cannon_loop 2286 14.5 1.796 1.936 24.535 25.927 ot_mini 108 10.5 0.001 0.001 19.090 19.351 qs_ot_get_derivative 108 11.5 0.001 0.001 16.174 16.336 mp_waitall_1 245248 16.5 8.689 14.271 8.689 14.271 multiply_cannon_metrocomm3 54864 15.5 0.074 0.080 5.123 12.232 multiply_cannon_multrec 54864 15.5 3.760 5.968 7.816 11.333 rebuild_ks_matrix 119 8.3 0.000 0.000 10.244 10.410 qs_ks_build_kohn_sham_matrix 119 9.3 0.011 0.012 10.243 10.409 qs_ot_get_p 119 10.4 0.001 0.001 9.526 9.787 qs_ks_update_qs_env 119 7.6 0.001 0.001 9.034 9.186 qs_rho_update_rho_low 119 7.7 0.000 0.001 6.785 6.860 calculate_rho_elec 119 8.7 0.012 0.017 6.785 6.859 sum_up_and_integrate 119 10.3 0.012 0.014 6.440 6.453 integrate_v_rspace 119 11.3 0.002 0.003 6.428 6.443 qs_ot_p2m_diag 50 11.0 0.004 0.006 6.278 6.325 mp_sum_l 7287 12.8 3.922 5.787 3.922 5.787 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 5.325 5.763 dbcsr_mm_accdrv_process 76910 16.1 1.765 2.871 3.969 5.735 multiply_cannon_sync_h2d 54864 15.5 5.133 5.685 5.133 5.685 init_scf_run 11 5.9 0.000 0.001 5.593 5.593 scf_env_initial_rho_setup 11 6.9 0.000 0.000 5.593 5.593 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 5.431 5.531 cp_dbcsr_syevd 50 12.0 0.002 0.003 5.443 5.445 rs_pw_transfer 974 11.9 0.012 0.013 4.947 5.214 cp_fm_diag_elpa 50 13.0 0.000 0.000 4.918 4.919 cp_fm_redistribute_end 50 14.0 2.491 4.847 2.516 4.856 density_rs2pw 119 9.7 0.004 0.005 4.521 4.684 cp_fm_diag_elpa_base 50 14.0 2.320 4.671 2.333 4.682 pw_transfer 1439 11.6 0.052 0.056 3.901 4.091 fft_wrap_pw1pw2 1201 12.6 0.007 0.007 3.826 4.020 potential_pw2rs 119 12.3 0.004 0.004 3.580 3.689 fft3d_ps 1201 14.6 0.382 0.498 3.477 3.674 make_m2s 4572 13.5 0.053 0.056 3.460 3.644 make_images 4572 14.5 0.134 0.140 3.377 3.561 init_scf_loop 11 6.9 0.000 0.000 3.336 3.337 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.009 3.203 apply_single 119 13.6 0.000 0.000 3.008 3.203 calculate_dm_sparse 119 9.5 0.000 0.001 3.027 3.183 fft_wrap_pw1pw2_140 487 13.2 0.176 0.195 2.979 3.173 multiply_cannon_metrocomm1 54864 15.5 0.058 0.066 1.514 2.847 acc_transpose_blocks 54864 15.5 0.224 0.252 2.196 2.808 jit_kernel_multiply 13 15.8 2.142 2.781 2.142 2.781 wfi_extrapolate 11 7.9 0.001 0.001 2.763 2.763 ot_diis_step 108 11.5 0.006 0.006 2.732 2.733 calculate_first_density_matrix 1 7.0 0.000 0.000 2.718 2.724 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.572 2.576 mp_alltoall_d11v 2130 13.8 2.256 2.502 2.256 2.502 mp_alltoall_z22v 1201 16.6 2.327 2.490 2.327 2.490 qs_ot_get_orbitals 108 10.5 0.000 0.000 2.298 2.362 make_images_sizes 4572 15.5 0.004 0.005 1.786 2.228 mp_alltoall_i44 4572 16.5 1.782 2.224 1.782 2.224 grid_integrate_task_list 119 12.3 1.972 2.128 1.972 2.128 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.022 2.075 prepare_preconditioner 11 7.9 0.000 0.000 1.997 2.027 make_preconditioner 11 8.9 0.000 0.000 1.997 2.027 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.907 1.948 mp_sum_d 4135 12.0 1.165 1.823 1.165 1.823 mp_waitany 12084 13.8 1.600 1.781 1.600 1.781 yz_to_x 249 15.8 0.019 0.028 1.459 1.610 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.573 1.589 make_images_data 4572 15.5 0.041 0.046 0.972 1.548 acc_transpose_blocks_sync 164592 16.5 1.158 1.457 1.158 1.457 grid_collocate_task_list 119 9.7 1.358 1.445 1.358 1.445 mp_allgather_i34 2286 14.5 0.934 1.439 0.934 1.439 hybrid_alltoall_any 4725 16.4 0.043 0.113 0.798 1.364 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.262 1.277 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="100", plot="h2o_64_md", label="(8n/12r/1t)", y=61.875000, yerr=0.000000 PlotPoint: name="101", plot="h2o_64_md_mem", label="(8n/12r/1t)", y=431.000000, yerr=1.414214 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 488.452096E+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 1164706. 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 0.041 0.053 41.130 41.131 qs_mol_dyn_low 1 2.0 0.003 0.003 40.884 40.892 qs_forces 11 3.9 0.002 0.003 40.816 40.818 qs_energies 11 4.9 0.001 0.004 39.033 39.038 scf_env_do_scf 11 5.9 0.001 0.001 33.559 33.559 scf_env_do_scf_inner_loop 108 6.5 0.012 0.018 30.931 30.931 dbcsr_multiply_generic 2286 12.5 0.100 0.103 22.531 22.882 qs_scf_new_mos 108 7.5 0.001 0.001 21.251 21.495 qs_scf_loop_do_ot 108 8.5 0.001 0.001 21.250 21.495 ot_scf_mini 108 9.5 0.002 0.003 20.316 20.493 velocity_verlet 10 3.0 0.001 0.002 19.448 19.449 multiply_cannon 2286 13.5 0.211 0.217 17.155 18.567 multiply_cannon_loop 2286 14.5 1.191 1.245 15.924 17.400 ot_mini 108 10.5 0.001 0.001 12.382 12.626 mp_waitall_1 200699 16.5 5.741 10.984 5.741 10.984 qs_ot_get_derivative 108 11.5 0.001 0.001 9.990 10.169 multiply_cannon_metrocomm3 27432 15.5 0.071 0.075 3.988 9.529 multiply_cannon_multrec 27432 15.5 1.821 4.193 6.346 8.993 rebuild_ks_matrix 119 8.3 0.000 0.000 7.556 7.693 qs_ks_build_kohn_sham_matrix 119 9.3 0.015 0.035 7.555 7.693 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.638 6.763 dbcsr_mm_accdrv_process 47894 16.0 3.723 5.924 4.443 6.521 qs_ot_get_p 119 10.4 0.001 0.001 4.964 5.187 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.709 4.620 sum_up_and_integrate 119 10.3 0.024 0.027 4.426 4.432 integrate_v_rspace 119 11.3 0.002 0.003 4.402 4.411 init_scf_run 11 5.9 0.000 0.001 4.240 4.240 scf_env_initial_rho_setup 11 6.9 0.000 0.001 4.239 4.240 mp_sum_l 7287 12.8 2.139 4.156 2.139 4.156 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.086 4.130 calculate_rho_elec 119 8.7 0.021 0.024 4.085 4.129 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 3.095 4.079 apply_single 119 13.6 0.000 0.000 3.094 4.079 rs_pw_transfer 974 11.9 0.010 0.011 2.900 3.502 qs_ot_p2m_diag 50 11.0 0.009 0.013 3.224 3.242 make_m2s 4572 13.5 0.052 0.054 2.797 3.132 make_images 4572 14.5 0.205 0.246 2.709 3.047 density_rs2pw 119 9.7 0.004 0.004 2.341 2.943 cp_dbcsr_syevd 50 12.0 0.003 0.003 2.811 2.812 calculate_first_density_matrix 1 7.0 0.000 0.003 2.615 2.620 init_scf_loop 11 6.9 0.001 0.006 2.603 2.603 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.404 2.494 cp_fm_diag_elpa 50 13.0 0.000 0.000 2.442 2.443 cp_fm_redistribute_end 50 14.0 1.234 2.406 1.239 2.410 calculate_dm_sparse 119 9.5 0.000 0.001 2.311 2.396 ot_diis_step 108 11.5 0.011 0.011 2.343 2.344 cp_fm_diag_elpa_base 50 14.0 1.134 2.297 1.165 2.344 multiply_cannon_sync_h2d 27432 15.5 1.717 2.240 1.717 2.240 potential_pw2rs 119 12.3 0.006 0.006 2.120 2.135 pw_transfer 1439 11.6 0.065 0.069 2.078 2.112 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.986 2.023 acc_transpose_blocks 27432 15.5 0.115 0.120 1.620 1.957 grid_integrate_task_list 119 12.3 1.838 1.941 1.838 1.941 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.912 1.915 jit_kernel_multiply 9 16.4 0.662 1.750 0.662 1.750 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.684 1.723 prepare_preconditioner 11 7.9 0.000 0.001 1.645 1.674 make_preconditioner 11 8.9 0.000 0.001 1.645 1.674 make_images_data 4572 15.5 0.048 0.054 1.293 1.671 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.545 1.606 fft3d_ps 1201 14.6 0.521 0.577 1.538 1.568 wfi_extrapolate 11 7.9 0.001 0.001 1.557 1.558 fft_wrap_pw1pw2_140 487 13.2 0.201 0.210 1.511 1.549 hybrid_alltoall_any 4725 16.4 0.054 0.117 1.150 1.526 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.494 1.507 grid_collocate_task_list 119 9.7 1.277 1.431 1.277 1.431 mp_alltoall_d11v 2130 13.8 1.291 1.405 1.291 1.405 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.223 1.269 mp_waitany 5720 13.7 0.531 1.162 0.531 1.162 rs_pw_transfer_RS2PW_140 130 11.5 0.139 0.147 0.557 1.143 mp_sum_d 4135 12.0 0.581 1.046 0.581 1.046 mp_allgather_i34 2286 14.5 0.652 1.043 0.652 1.043 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.978 0.990 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 0.947 0.949 acc_transpose_blocks_sync 82296 16.5 0.816 0.940 0.816 0.940 make_images_sizes 4572 15.5 0.005 0.005 0.634 0.905 mp_alltoall_i44 4572 16.5 0.629 0.901 0.629 0.901 acc_transpose_blocks_kernels 27432 16.5 0.189 0.278 0.662 0.872 mp_alltoall_z22v 1201 16.6 0.782 0.869 0.782 0.869 rs_pw_transfer_PW2RS_50 119 14.3 0.590 0.611 0.798 0.860 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="102", plot="h2o_64_md", label="(8n/6r/2t)", y=41.131000, yerr=0.000000 PlotPoint: name="103", plot="h2o_64_md_mem", label="(8n/6r/2t)", y=465.090909, yerr=1.504813 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/05/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 26877100032 0.0% 0.0% 100.0% flops 9 x 9 x 32 44168260608 0.0% 0.0% 100.0% flops 22 x 9 x 32 53835724800 0.0% 0.0% 100.0% flops 9 x 22 x 32 53885500416 0.0% 0.0% 100.0% flops 32 x 32 x 9 63568871424 0.0% 0.0% 100.0% flops 22 x 22 x 32 67007283200 0.0% 0.0% 100.0% flops 32 x 32 x 22 77695287296 0.0% 0.0% 100.0% flops 9 x 32 x 32 78422999040 0.0% 0.0% 100.0% flops 22 x 32 x 32 95850332160 0.0% 0.0% 100.0% flops 9 x 32 x 9 266263676928 0.0% 0.0% 100.0% flops 22 x 32 x 9 326697440256 0.0% 0.0% 100.0% flops 9 x 32 x 22 326697440256 0.0% 0.0% 100.0% flops 22 x 32 x 22 399918497792 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 1.880888E+12 0.0% 0.0% 100.0% flops max/rank 59.051995E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 146984760 0.0% 0.0% 100.0% number of processed stacks 3143552 0.0% 0.0% 100.0% average stack size 0.0 0.0 46.8 marketing flops 2.107587E+12 ------------------------------------------------------------------------------- # multiplications 2286 max memory usage/rank 520.806400E+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 0.031 0.054 34.258 34.263 qs_mol_dyn_low 1 2.0 0.004 0.013 33.824 33.832 qs_forces 11 3.9 0.003 0.007 33.019 33.024 qs_energies 11 4.9 0.006 0.043 31.422 31.428 scf_env_do_scf 11 5.9 0.001 0.002 26.614 26.619 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 23.909 23.910 dbcsr_multiply_generic 2286 12.5 0.094 0.102 17.011 17.083 velocity_verlet 10 3.0 0.001 0.002 16.245 16.270 qs_scf_new_mos 108 7.5 0.001 0.001 15.602 15.617 qs_scf_loop_do_ot 108 8.5 0.001 0.001 15.602 15.616 ot_scf_mini 108 9.5 0.002 0.003 14.846 14.860 multiply_cannon 2286 13.5 0.195 0.199 13.667 14.501 multiply_cannon_loop 2286 14.5 0.850 0.881 12.893 13.734 ot_mini 108 10.5 0.001 0.001 9.159 9.175 qs_ot_get_derivative 108 11.5 0.001 0.001 7.680 7.695 multiply_cannon_multrec 18288 15.5 1.862 2.823 6.960 7.335 rebuild_ks_matrix 119 8.3 0.000 0.000 6.503 6.514 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.019 6.502 6.514 dbcsr_mm_accdrv_process 38222 16.0 4.910 5.738 5.004 5.803 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.741 5.752 sum_up_and_integrate 119 10.3 0.031 0.032 3.997 4.003 integrate_v_rspace 119 11.3 0.002 0.003 3.965 3.976 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.595 3.597 calculate_rho_elec 119 8.7 0.031 0.032 3.594 3.597 init_scf_run 11 5.9 0.000 0.001 3.574 3.574 scf_env_initial_rho_setup 11 6.9 0.000 0.002 3.574 3.574 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.906 3.567 qs_ot_get_p 119 10.4 0.001 0.001 3.450 3.469 mp_waitall_1 158411 16.6 2.651 3.398 2.651 3.398 rs_pw_transfer 974 11.9 0.009 0.010 2.417 2.707 init_scf_loop 11 6.9 0.003 0.026 2.688 2.693 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.064 2.464 apply_single 119 13.6 0.000 0.000 2.064 2.464 density_rs2pw 119 9.7 0.004 0.004 2.076 2.369 qs_ot_p2m_diag 50 11.0 0.012 0.013 2.285 2.290 calculate_first_density_matrix 1 7.0 0.000 0.002 2.283 2.284 multiply_cannon_metrocomm3 18288 15.5 0.047 0.048 1.450 2.181 make_m2s 4572 13.5 0.044 0.046 1.842 1.987 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.985 1.986 pw_transfer 1439 11.6 0.065 0.069 1.927 1.935 make_images 4572 14.5 0.194 0.212 1.756 1.901 grid_integrate_task_list 119 12.3 1.803 1.901 1.803 1.901 calculate_dm_sparse 119 9.5 0.000 0.001 1.839 1.846 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.834 1.845 prepare_preconditioner 11 7.9 0.000 0.000 1.753 1.755 make_preconditioner 11 8.9 0.007 0.056 1.753 1.755 potential_pw2rs 119 12.3 0.007 0.008 1.746 1.754 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.734 1.744 cp_fm_diag_elpa_base 50 14.0 1.711 1.723 1.733 1.743 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.714 1.722 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.718 1.719 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.603 1.695 acc_transpose_blocks 18288 15.5 0.079 0.081 1.624 1.671 mp_sum_l 7287 12.8 1.175 1.596 1.175 1.596 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.541 1.547 fft_wrap_pw1pw2_140 487 13.2 0.252 0.258 1.461 1.473 ot_diis_step 108 11.5 0.011 0.011 1.466 1.466 grid_collocate_task_list 119 9.7 1.250 1.413 1.250 1.413 fft3d_ps 1201 14.6 0.530 0.548 1.322 1.334 multiply_cannon_sync_h2d 18288 15.5 1.061 1.220 1.061 1.220 wfi_extrapolate 11 7.9 0.001 0.001 1.197 1.197 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.179 1.182 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 0.999 1.004 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.944 0.962 make_images_data 4572 15.5 0.047 0.050 0.775 0.939 mp_waitany 9880 13.7 0.592 0.892 0.592 0.892 rs_pw_transfer_RS2PW_140 130 11.5 0.119 0.122 0.587 0.878 hybrid_alltoall_any 4725 16.4 0.058 0.115 0.666 0.852 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.835 0.836 acc_transpose_blocks_kernels 18288 16.5 0.218 0.225 0.796 0.806 mp_sum_dm 438 4.9 0.776 0.802 0.776 0.802 mp_alltoall_d11v 2130 13.8 0.672 0.799 0.672 0.799 md_write_output 11 3.9 0.024 0.743 0.026 0.787 md_output 10 3.0 0.000 0.000 0.027 0.786 update_particle_set 20 4.0 0.000 0.000 0.758 0.783 acc_transpose_blocks_sync 54864 16.5 0.729 0.770 0.729 0.770 jit_kernel_multiply 4 15.7 0.036 0.764 0.036 0.764 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.671 0.724 cp_fm_cholesky_invert 11 10.9 0.712 0.716 0.712 0.716 mp_alltoall_z22v 1201 16.6 0.646 0.716 0.646 0.716 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="104", plot="h2o_64_md", label="(8n/4r/3t)", y=34.263000, yerr=0.000000 PlotPoint: name="105", plot="h2o_64_md_mem", label="(8n/4r/3t)", y=496.454545, yerr=0.890724 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 560.558080E+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 0.022 0.038 37.783 37.784 qs_mol_dyn_low 1 2.0 0.003 0.003 37.495 37.504 qs_forces 11 3.9 0.002 0.003 37.120 37.120 qs_energies 11 4.9 0.001 0.003 35.366 35.371 scf_env_do_scf 11 5.9 0.001 0.001 30.194 30.195 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 26.743 26.744 dbcsr_multiply_generic 2286 12.5 0.099 0.108 20.073 20.178 velocity_verlet 10 3.0 0.010 0.010 19.222 19.237 qs_scf_new_mos 108 7.5 0.001 0.001 18.092 18.153 qs_scf_loop_do_ot 108 8.5 0.001 0.001 18.092 18.152 ot_scf_mini 108 9.5 0.002 0.004 17.056 17.115 multiply_cannon 2286 13.5 0.219 0.230 16.179 16.712 multiply_cannon_loop 2286 14.5 1.531 1.593 15.200 15.616 ot_mini 108 10.5 0.001 0.001 10.486 10.553 multiply_cannon_multrec 27432 15.5 2.467 3.144 9.035 9.355 qs_ot_get_derivative 108 11.5 0.001 0.001 8.605 8.668 dbcsr_mm_accdrv_process 47916 15.9 6.055 7.755 6.463 7.820 rebuild_ks_matrix 119 8.3 0.000 0.000 6.712 6.770 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 6.711 6.769 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.952 6.002 sum_up_and_integrate 119 10.3 0.035 0.038 3.883 3.893 integrate_v_rspace 119 11.3 0.003 0.003 3.847 3.858 init_scf_run 11 5.9 0.000 0.001 3.790 3.791 scf_env_initial_rho_setup 11 6.9 0.000 0.001 3.790 3.790 qs_ot_get_p 119 10.4 0.001 0.003 3.704 3.784 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.625 3.658 calculate_rho_elec 119 8.7 0.040 0.046 3.625 3.657 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 3.128 3.599 init_scf_loop 11 6.9 0.001 0.004 3.431 3.432 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.178 2.599 apply_single 119 13.6 0.000 0.000 2.178 2.599 prepare_preconditioner 11 7.9 0.000 0.000 2.564 2.572 make_preconditioner 11 8.9 0.000 0.002 2.563 2.572 acc_transpose_blocks 27432 15.5 0.120 0.122 2.415 2.511 make_full_inverse_cholesky 11 9.9 0.000 0.000 2.178 2.500 make_m2s 4572 13.5 0.054 0.058 2.349 2.458 qs_ot_get_derivative_diag 49 12.0 0.001 0.003 2.415 2.446 calculate_first_density_matrix 1 7.0 0.000 0.002 2.381 2.382 make_images 4572 14.5 0.271 0.332 2.240 2.348 mp_waitall_1 137007 16.6 1.839 2.347 1.839 2.347 rs_pw_transfer 974 11.9 0.009 0.009 2.129 2.319 qs_ot_p2m_diag 50 11.0 0.015 0.023 2.265 2.275 calculate_dm_sparse 119 9.5 0.000 0.000 2.146 2.207 density_rs2pw 119 9.7 0.004 0.004 2.013 2.206 pw_transfer 1439 11.6 0.065 0.070 2.037 2.073 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.944 1.985 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.921 1.922 grid_integrate_task_list 119 12.3 1.818 1.899 1.818 1.899 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.853 1.853 ot_diis_step 108 11.5 0.012 0.012 1.837 1.838 mp_sum_l 7287 12.8 1.140 1.776 1.140 1.776 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.732 1.750 fft_wrap_pw1pw2_140 487 13.2 0.288 0.301 1.591 1.634 potential_pw2rs 119 12.3 0.009 0.009 1.625 1.630 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.598 1.606 cp_fm_diag_elpa_base 50 14.0 1.563 1.580 1.595 1.604 acc_transpose_blocks_sync 82296 16.5 1.436 1.528 1.436 1.528 jit_kernel_multiply 8 16.1 0.342 1.436 0.342 1.436 fft3d_ps 1201 14.6 0.560 0.612 1.380 1.406 grid_collocate_task_list 119 9.7 1.250 1.383 1.250 1.383 wfi_extrapolate 11 7.9 0.001 0.001 1.356 1.356 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.302 1.320 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.239 1.248 multiply_cannon_metrocomm3 27432 15.5 0.040 0.042 0.743 1.216 cp_fm_upper_to_full 72 14.2 0.810 1.149 0.810 1.149 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 1.113 1.114 dbcsr_complete_redistribute 329 12.2 0.124 0.150 0.792 1.060 make_images_data 4572 15.5 0.047 0.050 0.874 1.000 hybrid_alltoall_any 4725 16.4 0.065 0.155 0.760 0.928 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.798 0.874 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.850 0.856 copy_fm_to_dbcsr 176 11.2 0.001 0.001 0.574 0.837 acc_transpose_blocks_kernels 27432 16.5 0.273 0.279 0.830 0.837 mp_alltoall_d11v 2130 13.8 0.749 0.826 0.749 0.826 cp_fm_cholesky_invert 11 10.9 0.755 0.758 0.755 0.758 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="106", plot="h2o_64_md", label="(8n/3r/4t)", y=37.784000, yerr=0.000000 PlotPoint: name="107", plot="h2o_64_md_mem", label="(8n/3r/4t)", y=531.181818, yerr=4.195629 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 621.789184E+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 0.022 0.038 29.288 29.289 qs_mol_dyn_low 1 2.0 0.003 0.004 28.908 28.917 qs_forces 11 3.9 0.031 0.091 28.801 28.805 qs_energies 11 4.9 0.001 0.001 27.047 27.081 scf_env_do_scf 11 5.9 0.000 0.001 22.173 22.174 scf_env_do_scf_inner_loop 108 6.5 0.003 0.006 19.699 19.699 velocity_verlet 10 3.0 0.002 0.002 14.932 14.938 dbcsr_multiply_generic 2286 12.5 0.090 0.093 12.977 13.059 qs_scf_new_mos 108 7.5 0.001 0.001 11.674 11.695 qs_scf_loop_do_ot 108 8.5 0.001 0.001 11.674 11.694 ot_scf_mini 108 9.5 0.002 0.002 10.962 10.984 multiply_cannon 2286 13.5 0.223 0.232 10.397 10.886 multiply_cannon_loop 2286 14.5 0.636 0.657 9.487 9.665 ot_mini 108 10.5 0.001 0.001 6.229 6.257 multiply_cannon_multrec 9144 15.5 1.661 1.857 5.996 6.187 rebuild_ks_matrix 119 8.3 0.000 0.000 5.997 6.022 qs_ks_build_kohn_sham_matrix 119 9.3 0.012 0.013 5.997 6.021 qs_ks_update_qs_env 119 7.6 0.001 0.001 5.334 5.356 qs_ot_get_derivative 108 11.5 0.001 0.001 4.936 4.959 dbcsr_mm_accdrv_process 12550 15.8 3.560 4.223 4.221 4.290 sum_up_and_integrate 119 10.3 0.037 0.041 3.672 3.676 integrate_v_rspace 119 11.3 0.003 0.003 3.635 3.639 qs_rho_update_rho_low 119 7.7 0.001 0.001 3.558 3.565 calculate_rho_elec 119 8.7 0.060 0.061 3.557 3.564 init_scf_run 11 5.9 0.000 0.001 3.346 3.346 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.345 3.345 qs_ot_get_p 119 10.4 0.001 0.002 2.854 2.885 init_scf_loop 11 6.9 0.000 0.000 2.455 2.456 calculate_first_density_matrix 1 7.0 0.000 0.000 2.199 2.200 pw_transfer 1439 11.6 0.066 0.069 2.044 2.053 density_rs2pw 119 9.7 0.004 0.004 1.847 1.985 fft_wrap_pw1pw2 1201 12.6 0.008 0.008 1.950 1.960 make_m2s 4572 13.5 0.034 0.035 1.794 1.945 rs_pw_transfer 974 11.9 0.008 0.008 1.740 1.938 grid_integrate_task_list 119 12.3 1.866 1.932 1.866 1.932 qs_ot_p2m_diag 50 11.0 0.022 0.023 1.855 1.858 make_images 4572 14.5 0.268 0.301 1.705 1.855 mp_waitall_1 115863 16.7 1.316 1.838 1.316 1.838 calculate_dm_sparse 119 9.5 0.000 0.000 1.818 1.835 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 1.701 1.722 prepare_preconditioner 11 7.9 0.000 0.000 1.692 1.699 make_preconditioner 11 8.9 0.000 0.000 1.692 1.699 fft_wrap_pw1pw2_140 487 13.2 0.365 0.372 1.606 1.617 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.609 1.609 make_full_inverse_cholesky 11 9.9 0.000 0.000 1.581 1.608 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.533 1.534 acc_transpose_blocks 9144 15.5 0.042 0.044 1.426 1.468 grid_collocate_task_list 119 9.7 1.299 1.426 1.299 1.426 potential_pw2rs 119 12.3 0.010 0.010 1.390 1.394 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.374 1.385 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 1.320 1.352 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 1.332 1.346 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.317 1.324 cp_fm_diag_elpa_base 50 14.0 1.289 1.305 1.315 1.323 fft3d_ps 1201 14.6 0.562 0.573 1.283 1.293 ot_diis_step 108 11.5 0.012 0.013 1.278 1.278 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 1.217 1.252 apply_single 119 13.6 0.000 0.000 1.217 1.252 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.201 1.207 wfi_extrapolate 11 7.9 0.001 0.001 1.093 1.093 jit_kernel_multiply 6 15.5 0.622 1.033 0.622 1.033 hybrid_alltoall_any 4725 16.4 0.065 0.179 0.773 0.988 make_images_data 4572 15.5 0.042 0.046 0.776 0.941 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.000 0.874 0.924 mp_alltoall_d11v 2130 13.8 0.775 0.870 0.775 0.870 cp_fm_cholesky_invert 11 10.9 0.846 0.849 0.846 0.849 qs_env_update_s_mstruct 11 6.9 0.000 0.000 0.761 0.833 qs_ot_get_orbitals 108 10.5 0.000 0.000 0.821 0.828 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.784 0.787 acc_transpose_blocks_sync 27432 16.5 0.710 0.748 0.710 0.748 mp_waitany 5200 13.7 0.510 0.680 0.510 0.680 rs_pw_transfer_RS2PW_140 130 11.5 0.102 0.106 0.475 0.669 acc_transpose_blocks_kernels 9144 16.5 0.118 0.120 0.658 0.661 mp_allgather_i34 2286 14.5 0.229 0.644 0.229 0.644 mp_alltoall_z22v 1201 16.6 0.596 0.627 0.596 0.627 multiply_cannon_metrocomm3 9144 15.5 0.020 0.020 0.313 0.598 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="108", plot="h2o_64_md", label="(8n/2r/6t)", y=29.289000, yerr=0.000000 PlotPoint: name="109", plot="h2o_64_md_mem", label="(8n/2r/6t)", y=585.909091, yerr=7.609259 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 763.502592E+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 0.019 0.050 43.591 43.593 qs_mol_dyn_low 1 2.0 0.003 0.004 43.368 43.377 qs_forces 11 3.9 0.002 0.002 43.303 43.303 qs_energies 11 4.9 0.001 0.001 41.264 41.267 scf_env_do_scf 11 5.9 0.001 0.001 35.540 35.540 scf_env_do_scf_inner_loop 108 6.5 0.003 0.007 27.544 27.545 velocity_verlet 10 3.0 0.002 0.002 24.506 24.512 dbcsr_multiply_generic 2286 12.5 0.098 0.100 19.262 19.507 qs_scf_new_mos 108 7.5 0.001 0.001 17.665 17.767 qs_scf_loop_do_ot 108 8.5 0.001 0.001 17.664 17.766 ot_scf_mini 108 9.5 0.002 0.002 16.489 16.596 multiply_cannon 2286 13.5 0.299 0.307 15.062 16.221 multiply_cannon_loop 2286 14.5 0.875 0.901 13.791 14.999 ot_mini 108 10.5 0.001 0.001 10.077 10.194 multiply_cannon_multrec 9144 15.5 3.400 4.826 8.830 9.075 qs_ot_get_derivative 108 11.5 0.001 0.001 8.021 8.126 init_scf_loop 11 6.9 0.000 0.000 7.968 7.970 rebuild_ks_matrix 119 8.3 0.000 0.000 7.447 7.591 qs_ks_build_kohn_sham_matrix 119 9.3 0.013 0.013 7.446 7.591 prepare_preconditioner 11 7.9 0.000 0.000 6.926 6.940 make_preconditioner 11 8.9 0.000 0.000 6.926 6.940 qs_ks_update_qs_env 119 7.6 0.001 0.001 6.705 6.835 make_full_inverse_cholesky 11 9.9 0.000 0.000 5.466 6.812 dbcsr_mm_accdrv_process 12550 15.8 4.536 6.496 5.286 6.732 cp_fm_upper_to_full 72 14.2 3.206 4.604 3.206 4.604 qs_rho_update_rho_low 119 7.7 0.001 0.001 4.313 4.316 calculate_rho_elec 119 8.7 0.118 0.121 4.312 4.316 sum_up_and_integrate 119 10.3 0.064 0.066 4.064 4.069 integrate_v_rspace 119 11.3 0.004 0.004 3.999 4.006 init_scf_run 11 5.9 0.000 0.001 3.630 3.630 scf_env_initial_rho_setup 11 6.9 0.000 0.000 3.629 3.630 qs_ot_get_p 119 10.4 0.002 0.004 3.457 3.599 qs_ot_get_derivative_taylor 59 13.0 0.001 0.001 2.850 3.495 mp_waitall_1 94719 16.7 2.268 3.353 2.268 3.353 dbcsr_complete_redistribute 329 12.2 0.287 0.293 2.014 2.872 apply_preconditioner_dbcsr 119 12.6 0.000 0.000 2.446 2.733 apply_single 119 13.6 0.000 0.000 2.446 2.733 pw_transfer 1439 11.6 0.069 0.070 2.683 2.688 fft_wrap_pw1pw2 1201 12.6 0.009 0.009 2.584 2.590 make_m2s 4572 13.5 0.037 0.038 2.387 2.541 copy_fm_to_dbcsr 176 11.2 0.001 0.001 1.697 2.535 make_images 4572 14.5 0.352 0.385 2.266 2.420 mp_alltoall_i22 627 13.8 1.434 2.293 1.434 2.293 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.454 2.284 multiply_cannon_metrocomm3 9144 15.5 0.021 0.021 1.274 2.280 calculate_dm_sparse 119 9.5 0.000 0.000 2.251 2.265 density_rs2pw 119 9.7 0.004 0.004 2.190 2.204 fft_wrap_pw1pw2_140 487 13.2 0.617 0.621 2.177 2.182 qs_ot_get_derivative_diag 49 12.0 0.001 0.001 2.087 2.141 calculate_first_density_matrix 1 7.0 0.000 0.000 2.137 2.140 grid_integrate_task_list 119 12.3 2.084 2.123 2.084 2.123 qs_ot_p2m_diag 50 11.0 0.043 0.044 2.046 2.049 mp_sum_l 7287 12.8 1.269 2.037 1.269 2.037 ot_diis_step 108 11.5 0.014 0.014 2.030 2.030 acc_transpose_blocks 9144 15.5 0.045 0.046 1.843 1.880 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 1.829 1.830 qs_energies_init_hamiltonians 11 5.9 0.040 0.041 1.808 1.809 cp_dbcsr_syevd 50 12.0 0.003 0.003 1.683 1.684 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 1.570 1.611 fft3d_ps 1201 14.6 0.597 0.611 1.601 1.606 grid_collocate_task_list 119 9.7 1.526 1.540 1.526 1.540 rs_pw_transfer 974 11.9 0.009 0.009 1.503 1.525 cp_fm_cholesky_invert 11 10.9 1.485 1.488 1.485 1.488 wfi_extrapolate 11 7.9 0.001 0.001 1.427 1.427 potential_pw2rs 119 12.3 0.014 0.014 1.413 1.415 cp_fm_diag_elpa 50 13.0 0.000 0.000 1.391 1.391 cp_fm_diag_elpa_base 50 14.0 1.242 1.297 1.389 1.389 qs_ot_get_orbitals 108 10.5 0.000 0.000 1.333 1.350 hybrid_alltoall_any 4725 16.4 0.090 0.148 1.072 1.266 jit_kernel_multiply 6 15.9 0.721 1.264 0.721 1.264 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.000 0.000 1.244 1.249 make_images_data 4572 15.5 0.046 0.049 1.010 1.203 acc_transpose_blocks_sync 27432 16.5 1.124 1.161 1.124 1.161 mp_alltoall_d11v 2130 13.8 1.125 1.146 1.125 1.146 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.095 1.112 build_core_hamiltonian_matrix_ 11 4.9 0.000 0.001 0.992 1.032 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 0.953 0.967 qs_create_task_list 11 7.9 0.001 0.001 0.930 0.945 generate_qs_task_list 11 8.9 0.367 0.386 0.929 0.945 mp_alltoall_z22v 1201 16.6 0.868 0.884 0.868 0.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="110", plot="h2o_64_md", label="(8n/1r/12t)", y=43.593000, yerr=0.000000 PlotPoint: name="111", plot="h2o_64_md_mem", label="(8n/1r/12t)", y=713.454545, yerr=18.539215 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/09/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 198.287135E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 8410880 0.0% 0.0% 100.0% average stack size 0.0 0.0 117.0 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 501.137408E+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 496241. 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 0.068 0.151 83.870 83.874 qs_mol_dyn_low 1 2.0 0.003 0.006 83.306 83.317 qs_forces 11 3.9 0.005 0.014 83.232 83.233 qs_energies 11 4.9 0.002 0.007 80.317 80.333 scf_env_do_scf 11 5.9 0.001 0.001 71.033 71.036 scf_env_do_scf_inner_loop 99 6.5 0.004 0.014 65.513 65.514 dbcsr_multiply_generic 2055 12.4 0.106 0.111 51.728 51.984 qs_scf_new_mos 99 7.5 0.000 0.001 47.954 48.039 qs_scf_loop_do_ot 99 8.5 0.001 0.001 47.954 48.039 ot_scf_mini 99 9.5 0.002 0.003 45.569 45.641 velocity_verlet 10 3.0 0.001 0.002 43.924 43.925 multiply_cannon 2055 13.4 0.183 0.198 43.141 43.832 multiply_cannon_loop 2055 14.4 1.804 1.841 42.137 42.893 ot_mini 99 10.5 0.001 0.001 26.911 26.981 qs_ot_get_derivative 99 11.5 0.001 0.001 20.129 20.200 multiply_cannon_multrec 49320 15.4 11.318 11.958 17.306 17.911 rebuild_ks_matrix 110 8.3 0.000 0.000 14.681 14.781 qs_ks_build_kohn_sham_matrix 110 9.3 0.011 0.015 14.680 14.780 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.840 12.931 mp_waitall_1 220248 16.4 11.136 11.909 11.136 11.909 qs_ot_get_p 110 10.4 0.001 0.002 10.121 10.253 multiply_cannon_sync_h2d 49320 15.4 9.539 10.204 9.539 10.204 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 7.782 8.307 multiply_cannon_metrocomm3 49320 15.4 0.084 0.086 6.753 7.854 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 7.171 7.653 apply_single 110 13.6 0.000 0.001 7.171 7.652 sum_up_and_integrate 110 10.3 0.036 0.043 7.180 7.193 init_scf_run 11 5.9 0.000 0.001 7.177 7.178 scf_env_initial_rho_setup 11 6.9 0.000 0.002 7.177 7.177 integrate_v_rspace 110 11.3 0.003 0.004 7.143 7.166 qs_rho_update_rho_low 110 7.6 0.001 0.001 6.767 6.918 calculate_rho_elec 110 8.6 0.021 0.026 6.766 6.918 qs_ot_p2m_diag 48 11.0 0.013 0.019 6.755 6.802 ot_diis_step 99 11.5 0.006 0.006 6.547 6.547 dbcsr_mm_accdrv_process 87628 16.1 3.033 3.137 5.858 6.110 cp_dbcsr_syevd 48 12.0 0.002 0.003 5.844 5.844 init_scf_loop 11 6.9 0.001 0.004 5.491 5.496 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 5.425 5.464 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.257 5.279 cp_fm_diag_elpa_base 48 14.0 5.242 5.265 5.255 5.278 mp_sum_l 6594 12.7 4.269 4.969 4.269 4.969 rs_pw_transfer 902 11.9 0.012 0.013 3.798 4.501 density_rs2pw 110 9.6 0.004 0.005 3.484 4.109 wfi_extrapolate 11 7.9 0.001 0.001 4.104 4.104 make_m2s 4110 13.4 0.060 0.065 3.933 4.066 make_images 4110 14.4 0.177 0.189 3.838 3.973 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.006 3.893 3.898 calculate_dm_sparse 110 9.5 0.001 0.001 3.759 3.896 multiply_cannon_metrocomm1 49320 15.4 0.065 0.068 2.501 3.574 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 3.535 3.565 grid_integrate_task_list 110 12.3 3.274 3.452 3.274 3.452 qs_ot_get_orbitals 99 10.5 0.001 0.001 3.273 3.321 prepare_preconditioner 11 7.9 0.000 0.000 3.284 3.300 make_preconditioner 11 8.9 0.000 0.001 3.284 3.300 pw_transfer 1331 11.6 0.055 0.065 3.211 3.291 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 3.122 3.204 make_full_inverse_cholesky 11 9.9 0.000 0.000 3.082 3.127 calculate_first_density_matrix 1 7.0 0.001 0.002 2.971 2.977 acc_transpose_blocks 49320 15.4 0.216 0.224 2.901 2.971 fft_wrap_pw1pw2_140 451 13.1 0.453 0.500 2.642 2.734 potential_pw2rs 110 12.3 0.006 0.007 2.690 2.716 jit_kernel_multiply 13 15.9 2.548 2.565 2.548 2.565 mp_waitany 14300 13.8 1.840 2.529 1.840 2.529 mp_alltoall_d11v 2046 13.8 2.050 2.492 2.050 2.492 grid_collocate_task_list 110 9.6 2.156 2.394 2.156 2.394 fft3d_ps 1111 14.6 0.798 0.884 2.316 2.369 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.973 1.988 mp_sum_d 3889 11.9 1.410 1.915 1.410 1.915 make_images_data 4110 15.4 0.044 0.048 1.782 1.908 hybrid_alltoall_any 4261 16.3 0.085 0.484 1.538 1.845 cp_fm_cholesky_invert 11 10.9 1.823 1.827 1.823 1.827 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.655 1.682 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="200", plot="h2o_128_md", label="(8n/12r/1t)", y=83.874000, yerr=0.000000 PlotPoint: name="201", plot="h2o_128_md_mem", label="(8n/12r/1t)", y=475.363636, yerr=2.603240 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 592.453632E+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 1945019. 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 0.042 0.068 88.511 88.512 qs_mol_dyn_low 1 2.0 0.003 0.004 88.113 88.123 qs_forces 11 3.9 0.004 0.010 88.036 88.037 qs_energies 11 4.9 0.003 0.015 84.343 84.370 scf_env_do_scf 11 5.9 0.001 0.001 74.786 74.788 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 63.132 63.133 velocity_verlet 10 3.0 0.001 0.002 46.353 46.356 dbcsr_multiply_generic 2055 12.4 0.120 0.125 43.652 45.042 qs_scf_new_mos 99 7.5 0.001 0.001 43.071 43.338 qs_scf_loop_do_ot 99 8.5 0.001 0.001 43.071 43.338 ot_scf_mini 99 9.5 0.003 0.004 40.995 41.265 multiply_cannon 2055 13.4 0.226 0.246 33.155 37.769 multiply_cannon_loop 2055 14.4 1.162 1.205 31.274 35.541 ot_mini 99 10.5 0.001 0.001 23.112 23.226 mp_waitall_1 176588 16.5 10.701 18.798 10.701 18.798 qs_ot_get_derivative 99 11.5 0.001 0.001 16.828 17.061 rebuild_ks_matrix 110 8.3 0.000 0.000 15.871 16.005 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.023 15.871 16.005 multiply_cannon_multrec 24660 15.4 6.711 8.387 13.757 15.624 multiply_cannon_metrocomm3 24660 15.4 0.071 0.076 6.577 14.421 qs_ks_update_qs_env 110 7.6 0.001 0.001 13.968 14.080 init_scf_loop 11 6.9 0.001 0.006 11.595 11.597 qs_ot_get_p 110 10.4 0.001 0.001 10.326 10.493 prepare_preconditioner 11 7.9 0.000 0.001 9.291 9.315 make_preconditioner 11 8.9 0.000 0.001 9.291 9.315 make_full_inverse_cholesky 11 9.9 0.000 0.000 8.777 9.006 sum_up_and_integrate 110 10.3 0.052 0.059 8.487 8.500 integrate_v_rspace 110 11.3 0.003 0.003 8.435 8.449 qs_rho_update_rho_low 110 7.6 0.001 0.001 8.257 8.280 calculate_rho_elec 110 8.6 0.040 0.047 8.256 8.279 qs_ot_p2m_diag 48 11.0 0.029 0.044 7.476 7.546 multiply_cannon_sync_h2d 24660 15.4 6.205 7.365 6.205 7.365 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 6.608 7.285 apply_single 110 13.6 0.000 0.001 6.608 7.284 dbcsr_mm_accdrv_process 52282 16.1 5.392 6.477 6.884 7.203 init_scf_run 11 5.9 0.000 0.001 6.879 6.880 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.879 6.880 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.817 6.818 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 5.938 6.685 rs_pw_transfer 902 11.9 0.012 0.014 5.528 6.510 make_m2s 4110 13.4 0.057 0.059 5.647 6.064 ot_diis_step 99 11.5 0.010 0.010 6.003 6.003 density_rs2pw 110 9.6 0.004 0.005 5.229 5.951 make_images 4110 14.4 0.404 0.457 5.536 5.947 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.850 5.866 cp_fm_diag_elpa_base 48 14.0 5.785 5.802 5.847 5.863 mp_sum_l 6594 12.7 4.223 5.524 4.223 5.524 cp_fm_cholesky_invert 11 10.9 5.063 5.087 5.063 5.087 pw_transfer 1331 11.6 0.067 0.077 4.803 5.083 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.923 5.052 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.695 4.976 potential_pw2rs 110 12.3 0.008 0.009 4.081 4.203 wfi_extrapolate 11 7.9 0.001 0.001 4.191 4.191 fft3d_ps 1111 14.6 1.109 1.337 3.677 3.979 fft_wrap_pw1pw2_140 451 13.1 0.512 0.533 3.440 3.725 make_images_data 4110 15.4 0.049 0.054 2.759 3.459 calculate_dm_sparse 110 9.5 0.001 0.001 3.411 3.445 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.002 3.400 3.405 grid_integrate_task_list 110 12.3 3.107 3.395 3.107 3.395 hybrid_alltoall_any 4261 16.3 0.107 0.453 2.377 3.326 mp_alltoall_d11v 2046 13.8 2.443 3.084 2.443 3.084 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.927 2.976 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.740 2.906 calculate_first_density_matrix 1 7.0 0.001 0.007 2.556 2.560 grid_collocate_task_list 110 9.6 2.097 2.530 2.097 2.530 cp_fm_cholesky_decompose 22 10.9 2.459 2.466 2.459 2.466 mp_alltoall_z22v 1111 16.6 2.219 2.438 2.219 2.438 multiply_cannon_metrocomm4 22605 15.4 0.080 0.086 0.795 2.351 mp_waitany 10164 13.8 1.648 2.330 1.648 2.330 mp_allgather_i34 2055 14.4 1.188 2.314 1.188 2.314 jit_kernel_multiply 11 16.2 1.131 2.124 1.131 2.124 mp_irecv_dv 57340 16.2 0.665 2.117 0.665 2.117 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.077 2.094 acc_transpose_blocks 24660 15.4 0.117 0.122 1.982 2.015 qs_energies_init_hamiltonians 11 5.9 0.013 0.019 2.011 2.012 mp_sum_d 3889 11.9 1.538 1.924 1.538 1.924 dbcsr_complete_redistribute 325 12.2 0.252 0.307 1.589 1.885 rs_pw_transfer_RS2PW_140 121 11.5 0.207 0.218 1.108 1.860 make_images_sizes 4110 15.4 0.005 0.005 1.425 1.812 mp_alltoall_i44 4110 16.4 1.420 1.807 1.420 1.807 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="202", plot="h2o_128_md", label="(8n/6r/2t)", y=88.512000, yerr=0.000000 PlotPoint: name="203", plot="h2o_128_md_mem", label="(8n/6r/2t)", y=559.727273, yerr=5.658315 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 667.729920E+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 0.037 0.078 78.929 78.930 qs_mol_dyn_low 1 2.0 0.003 0.004 78.447 78.481 qs_forces 11 3.9 0.005 0.011 78.136 78.137 qs_energies 11 4.9 0.002 0.007 74.618 74.639 scf_env_do_scf 11 5.9 0.001 0.001 65.381 65.381 scf_env_do_scf_inner_loop 99 6.5 0.003 0.006 52.683 52.685 velocity_verlet 10 3.0 0.001 0.002 42.580 42.589 dbcsr_multiply_generic 2055 12.4 0.113 0.119 34.493 35.937 qs_scf_new_mos 99 7.5 0.001 0.001 34.683 34.831 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.683 34.831 ot_scf_mini 99 9.5 0.002 0.004 33.084 33.231 multiply_cannon 2055 13.4 0.213 0.222 24.697 29.479 multiply_cannon_loop 2055 14.4 0.813 0.846 22.941 26.896 ot_mini 99 10.5 0.001 0.001 18.026 18.241 mp_waitall_1 139946 16.5 9.885 16.641 9.885 16.641 rebuild_ks_matrix 110 8.3 0.000 0.000 14.123 14.600 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.015 14.123 14.599 qs_ot_get_derivative 99 11.5 0.001 0.001 12.996 13.150 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.432 12.863 init_scf_loop 11 6.9 0.001 0.004 12.649 12.652 multiply_cannon_metrocomm3 16440 15.4 0.045 0.048 5.720 11.920 prepare_preconditioner 11 7.9 0.000 0.000 10.623 10.651 make_preconditioner 11 8.9 0.000 0.001 10.623 10.650 multiply_cannon_multrec 16440 15.4 3.436 4.369 9.787 10.650 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.891 10.272 qs_ot_get_p 110 10.4 0.001 0.002 9.391 9.614 sum_up_and_integrate 110 10.3 0.060 0.061 7.968 7.983 integrate_v_rspace 110 11.3 0.003 0.004 7.908 7.923 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.565 7.582 calculate_rho_elec 110 8.6 0.059 0.060 7.564 7.581 qs_ot_p2m_diag 48 11.0 0.042 0.044 6.748 6.800 init_scf_run 11 5.9 0.000 0.001 6.590 6.590 scf_env_initial_rho_setup 11 6.9 0.000 0.002 6.590 6.590 dbcsr_mm_accdrv_process 34862 16.1 5.471 5.840 6.205 6.302 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.216 6.218 rs_pw_transfer 902 11.9 0.010 0.011 4.781 6.007 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.100 5.864 apply_single 110 13.6 0.000 0.000 5.100 5.864 density_rs2pw 110 9.6 0.004 0.005 4.512 5.809 make_m2s 4110 13.4 0.049 0.052 5.369 5.786 make_images 4110 14.4 0.397 0.519 5.252 5.668 mp_sum_l 6594 12.7 3.819 5.416 3.819 5.416 cp_fm_cholesky_invert 11 10.9 5.333 5.346 5.333 5.346 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.324 5.331 cp_fm_diag_elpa_base 48 14.0 5.240 5.266 5.322 5.328 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.320 4.972 ot_diis_step 99 11.5 0.010 0.011 4.815 4.816 pw_transfer 1331 11.6 0.066 0.072 4.359 4.423 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.252 4.316 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.855 3.951 multiply_cannon_sync_h2d 16440 15.4 3.176 3.786 3.176 3.786 wfi_extrapolate 11 7.9 0.001 0.001 3.747 3.747 make_images_data 4110 15.4 0.045 0.050 2.781 3.704 potential_pw2rs 110 12.3 0.011 0.011 3.562 3.624 hybrid_alltoall_any 4261 16.3 0.109 0.381 2.430 3.490 fft_wrap_pw1pw2_140 451 13.1 0.635 0.646 3.419 3.487 grid_integrate_task_list 110 12.3 3.161 3.474 3.161 3.474 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.169 3.171 fft3d_ps 1111 14.6 1.092 1.104 3.077 3.136 calculate_dm_sparse 110 9.5 0.001 0.001 3.043 3.072 mp_alltoall_d11v 2046 13.8 2.298 2.963 2.298 2.963 mp_waitany 17072 13.8 1.581 2.904 1.581 2.904 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.628 2.784 calculate_first_density_matrix 1 7.0 0.001 0.004 2.724 2.726 multiply_cannon_metrocomm4 14385 15.4 0.048 0.054 0.932 2.692 cp_fm_cholesky_decompose 22 10.9 2.663 2.685 2.663 2.685 mp_irecv_dv 48980 15.7 0.858 2.551 0.858 2.551 grid_collocate_task_list 110 9.6 2.123 2.521 2.123 2.521 dbcsr_complete_redistribute 325 12.2 0.321 0.369 1.818 2.325 rs_pw_transfer_RS2PW_140 121 11.5 0.178 0.183 1.020 2.305 mp_allgather_i34 2055 14.4 1.039 2.161 1.039 2.161 qs_energies_init_hamiltonians 11 5.9 0.001 0.004 2.059 2.061 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.946 2.058 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.889 1.934 mp_alltoall_z22v 1111 16.6 1.776 1.840 1.776 1.840 cp_fm_upper_to_full 70 14.2 1.415 1.832 1.415 1.832 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.665 1.708 make_images_sizes 4110 15.4 0.005 0.005 1.152 1.672 mp_alltoall_i44 4110 16.4 1.147 1.667 1.147 1.667 copy_fm_to_dbcsr 174 11.2 0.001 0.002 1.184 1.665 acc_transpose_blocks 16440 15.4 0.079 0.082 1.592 1.633 rs_gather_matrices 110 12.3 0.234 0.258 1.095 1.628 mp_sum_d 3887 11.9 1.240 1.628 1.240 1.628 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="204", plot="h2o_128_md", label="(8n/4r/3t)", y=78.930000, yerr=0.000000 PlotPoint: name="205", plot="h2o_128_md_mem", label="(8n/4r/3t)", y=631.727273, yerr=9.630358 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/12/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 32 x 32 x 32 184415158272 0.0% 0.0% 100.0% flops 9 x 9 x 32 269180485632 0.0% 0.0% 100.0% flops 9 x 22 x 32 349395425280 0.0% 0.0% 100.0% flops 22 x 9 x 32 350042406912 0.0% 0.0% 100.0% flops 22 x 22 x 32 453581815808 0.0% 0.0% 100.0% flops 32 x 32 x 9 465064427520 0.0% 0.0% 100.0% flops 32 x 32 x 22 568412078080 0.0% 0.0% 100.0% flops 9 x 32 x 32 572195340288 0.0% 0.0% 100.0% flops 22 x 32 x 32 699349860352 0.0% 0.0% 100.0% flops 9 x 32 x 9 1735942275072 0.0% 0.0% 100.0% flops 22 x 32 x 9 2216407818240 0.0% 0.0% 100.0% flops 9 x 32 x 22 2216407818240 0.0% 0.0% 100.0% flops 22 x 32 x 22 2803661053952 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 12.884056E+12 0.0% 0.0% 100.0% flops max/rank 601.317074E+09 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 984178160 0.0% 0.0% 100.0% number of processed stacks 4916280 0.0% 0.0% 100.0% average stack size 0.0 0.0 200.2 marketing flops 15.646302E+12 ------------------------------------------------------------------------------- # multiplications 2055 max memory usage/rank 732.512256E+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 0.039 82.778 82.780 qs_mol_dyn_low 1 2.0 0.003 0.004 82.071 82.082 qs_forces 11 3.9 0.005 0.015 81.990 81.992 qs_energies 11 4.9 0.002 0.006 78.288 78.303 scf_env_do_scf 11 5.9 0.001 0.001 68.894 68.897 scf_env_do_scf_inner_loop 99 6.5 0.004 0.018 52.703 52.704 velocity_verlet 10 3.0 0.001 0.002 46.152 46.155 qs_scf_new_mos 99 7.5 0.001 0.001 34.863 34.965 qs_scf_loop_do_ot 99 8.5 0.001 0.001 34.863 34.965 dbcsr_multiply_generic 2055 12.4 0.118 0.128 34.088 34.743 ot_scf_mini 99 9.5 0.003 0.003 32.978 33.099 multiply_cannon 2055 13.4 0.239 0.260 24.701 26.891 multiply_cannon_loop 2055 14.4 1.418 1.485 22.825 24.236 ot_mini 99 10.5 0.001 0.001 17.713 17.829 init_scf_loop 11 6.9 0.001 0.003 16.127 16.128 prepare_preconditioner 11 7.9 0.000 0.000 14.182 14.200 make_preconditioner 11 8.9 0.002 0.011 14.182 14.200 multiply_cannon_multrec 24660 15.4 4.062 6.414 12.988 14.040 rebuild_ks_matrix 110 8.3 0.000 0.000 13.843 13.971 qs_ks_build_kohn_sham_matrix 110 9.3 0.012 0.014 13.843 13.970 make_full_inverse_cholesky 11 9.9 0.000 0.000 12.376 13.838 qs_ot_get_derivative 99 11.5 0.001 0.001 13.155 13.272 qs_ks_update_qs_env 110 7.6 0.001 0.001 12.223 12.335 qs_ot_get_p 110 10.4 0.001 0.002 9.609 9.789 dbcsr_mm_accdrv_process 52304 16.0 7.619 9.033 8.771 9.717 mp_waitall_1 121746 16.5 6.333 9.671 6.333 9.671 sum_up_and_integrate 110 10.3 0.067 0.071 7.871 7.886 integrate_v_rspace 110 11.3 0.003 0.004 7.804 7.823 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.692 7.709 calculate_rho_elec 110 8.6 0.078 0.082 7.691 7.709 make_m2s 4110 13.4 0.060 0.063 6.966 7.216 qs_ot_p2m_diag 48 11.0 0.058 0.084 7.079 7.099 make_images 4110 14.4 0.572 0.695 6.823 7.068 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.407 6.413 init_scf_run 11 5.9 0.000 0.001 6.344 6.344 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.344 6.344 cp_fm_cholesky_invert 11 10.9 5.680 5.698 5.680 5.698 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.408 5.419 cp_fm_diag_elpa_base 48 14.0 5.229 5.283 5.405 5.417 multiply_cannon_metrocomm3 24660 15.4 0.038 0.039 1.951 5.322 density_rs2pw 110 9.6 0.004 0.004 4.458 4.970 cp_fm_upper_to_full 70 14.2 3.435 4.951 3.435 4.951 rs_pw_transfer 902 11.9 0.010 0.011 4.369 4.863 pw_transfer 1331 11.6 0.066 0.076 4.609 4.704 fft_wrap_pw1pw2 1111 12.6 0.008 0.008 4.502 4.602 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 4.195 4.427 apply_single 110 13.6 0.000 0.000 4.195 4.427 ot_diis_step 99 11.5 0.011 0.011 4.414 4.415 dbcsr_complete_redistribute 325 12.2 0.409 0.485 3.035 4.276 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 4.193 4.245 make_images_data 4110 15.4 0.048 0.052 3.341 4.207 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 4.139 4.203 fft_wrap_pw1pw2_140 451 13.1 0.665 0.689 3.644 3.743 wfi_extrapolate 11 7.9 0.001 0.001 3.701 3.702 copy_fm_to_dbcsr 174 11.2 0.001 0.001 2.327 3.561 potential_pw2rs 110 12.3 0.012 0.013 3.385 3.450 grid_integrate_task_list 110 12.3 3.241 3.408 3.241 3.408 fft3d_ps 1111 14.6 1.086 1.118 3.268 3.346 hybrid_alltoall_any 4261 16.3 0.123 0.462 2.767 3.332 calculate_dm_sparse 110 9.5 0.001 0.001 3.209 3.235 mp_alltoall_d11v 2046 13.8 2.355 3.079 2.355 3.079 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.019 3.021 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 1.795 3.015 mp_alltoall_i22 605 13.7 1.731 3.006 1.731 3.006 cp_fm_cholesky_decompose 22 10.9 2.872 2.920 2.872 2.920 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.532 2.581 acc_transpose_blocks 24660 15.4 0.112 0.115 2.420 2.547 calculate_first_density_matrix 1 7.0 0.001 0.004 2.519 2.523 multiply_cannon_sync_h2d 24660 15.4 2.359 2.506 2.359 2.506 grid_collocate_task_list 110 9.6 2.222 2.464 2.222 2.464 qs_energies_init_hamiltonians 11 5.9 0.006 0.038 2.366 2.370 qs_ot_get_orbitals 99 10.5 0.001 0.001 2.233 2.322 mp_sum_l 6594 12.7 1.751 2.289 1.751 2.289 mp_alltoall_z22v 1111 16.6 1.966 2.078 1.966 2.078 mp_waitany 13376 13.8 1.477 1.981 1.477 1.981 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.867 1.881 jit_kernel_multiply 10 15.7 0.809 1.860 0.809 1.860 make_images_sizes 4110 15.4 0.005 0.005 1.274 1.792 mp_alltoall_i44 4110 16.4 1.270 1.788 1.270 1.788 rs_gather_matrices 110 12.3 0.267 0.293 1.086 1.742 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.608 1.707 mp_allgather_i34 2055 14.4 0.963 1.705 0.963 1.705 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="206", plot="h2o_128_md", label="(8n/3r/4t)", y=82.780000, yerr=0.000000 PlotPoint: name="207", plot="h2o_128_md_mem", label="(8n/3r/4t)", y=692.090909, yerr=9.394917 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 863.162368E+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 0.025 0.044 71.389 71.394 qs_mol_dyn_low 1 2.0 0.003 0.004 71.108 71.120 qs_forces 11 3.9 0.003 0.003 71.027 71.030 qs_energies 11 4.9 0.001 0.001 67.049 67.057 scf_env_do_scf 11 5.9 0.000 0.001 57.831 57.831 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 45.989 45.991 velocity_verlet 10 3.0 0.002 0.002 39.818 39.823 qs_scf_new_mos 99 7.5 0.001 0.001 28.233 28.346 qs_scf_loop_do_ot 99 8.5 0.001 0.001 28.233 28.346 dbcsr_multiply_generic 2055 12.4 0.108 0.112 27.019 27.557 ot_scf_mini 99 9.5 0.002 0.002 26.771 26.858 multiply_cannon 2055 13.4 0.240 0.251 19.286 22.027 multiply_cannon_loop 2055 14.4 0.600 0.629 17.699 19.246 ot_mini 99 10.5 0.001 0.001 13.588 13.657 rebuild_ks_matrix 110 8.3 0.000 0.000 13.520 13.655 qs_ks_build_kohn_sham_matrix 110 9.3 0.013 0.013 13.520 13.655 qs_ks_update_qs_env 110 7.6 0.001 0.001 11.994 12.114 init_scf_loop 11 6.9 0.000 0.000 11.779 11.781 mp_waitall_1 103326 16.6 8.143 11.135 8.143 11.135 prepare_preconditioner 11 7.9 0.000 0.000 9.900 9.910 make_preconditioner 11 8.9 0.000 0.000 9.900 9.910 qs_ot_get_derivative 99 11.5 0.001 0.001 9.537 9.619 make_full_inverse_cholesky 11 9.9 0.000 0.000 9.366 9.546 qs_ot_get_p 110 10.4 0.002 0.003 8.698 8.792 multiply_cannon_multrec 8220 15.4 3.170 4.571 7.577 8.602 qs_rho_update_rho_low 110 7.6 0.001 0.001 7.824 7.850 calculate_rho_elec 110 8.6 0.114 0.115 7.824 7.850 sum_up_and_integrate 110 10.3 0.079 0.080 7.791 7.803 integrate_v_rspace 110 11.3 0.003 0.003 7.712 7.724 qs_ot_p2m_diag 48 11.0 0.081 0.085 6.584 6.603 multiply_cannon_metrocomm3 8220 15.4 0.019 0.019 3.599 6.485 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.091 6.094 init_scf_run 11 5.9 0.000 0.001 5.866 5.867 scf_env_initial_rho_setup 11 6.9 0.000 0.001 5.866 5.866 make_m2s 4110 13.4 0.038 0.039 5.340 5.755 cp_fm_cholesky_invert 11 10.9 5.716 5.725 5.716 5.725 make_images 4110 14.4 0.647 0.706 5.208 5.626 dbcsr_mm_accdrv_process 17442 15.9 3.057 3.996 4.265 5.206 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.190 5.198 cp_fm_diag_elpa_base 48 14.0 5.130 5.156 5.188 5.197 pw_transfer 1331 11.6 0.067 0.073 4.934 4.977 fft_wrap_pw1pw2 1111 12.6 0.008 0.009 4.826 4.874 density_rs2pw 110 9.6 0.004 0.005 4.341 4.716 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 3.921 4.114 apply_single 110 13.6 0.000 0.000 3.921 4.113 rs_pw_transfer 902 11.9 0.010 0.010 3.693 4.053 fft_wrap_pw1pw2_140 451 13.1 0.830 0.848 3.949 4.001 ot_diis_step 99 11.5 0.012 0.012 3.959 3.960 grid_integrate_task_list 110 12.3 3.384 3.593 3.384 3.593 wfi_extrapolate 11 7.9 0.001 0.001 3.452 3.452 make_images_data 4110 15.4 0.041 0.046 2.613 3.446 hybrid_alltoall_any 4261 16.3 0.200 0.851 2.547 3.432 fft3d_ps 1111 14.6 1.145 1.183 3.374 3.423 potential_pw2rs 110 12.3 0.015 0.015 3.088 3.123 multiply_cannon_metrocomm1 8220 15.4 0.022 0.023 1.521 2.975 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 2.889 2.941 mp_alltoall_d11v 2046 13.8 2.425 2.887 2.425 2.887 calculate_dm_sparse 110 9.5 0.001 0.001 2.716 2.768 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 2.727 2.767 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 2.728 2.732 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 2.687 2.689 grid_collocate_task_list 110 9.6 2.322 2.584 2.322 2.584 cp_fm_cholesky_decompose 22 10.9 2.547 2.570 2.547 2.570 multiply_cannon_sync_h2d 8220 15.4 2.355 2.478 2.355 2.478 mp_sum_l 6594 12.7 1.651 2.361 1.651 2.361 calculate_first_density_matrix 1 7.0 0.000 0.000 2.297 2.299 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.188 2.232 mp_alltoall_z22v 1111 16.6 1.990 2.040 1.990 2.040 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 1.786 2.014 dbcsr_complete_redistribute 325 12.2 0.554 0.589 1.880 1.976 mp_allgather_i34 2055 14.4 0.758 1.957 0.758 1.957 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 1.863 1.877 mp_waitany 9240 13.8 1.405 1.794 1.405 1.794 qs_env_update_s_mstruct 11 6.9 0.000 0.000 1.548 1.678 make_images_sizes 4110 15.4 0.005 0.005 1.121 1.657 mp_alltoall_i44 4110 16.4 1.117 1.652 1.117 1.652 jit_kernel_multiply 9 15.6 0.896 1.548 0.896 1.548 cp_dbcsr_plus_fm_fm_t_native 22 8.9 0.001 0.001 1.500 1.519 rs_gather_matrices 110 12.3 0.327 0.374 1.137 1.515 copy_dbcsr_to_fm 151 11.3 0.003 0.003 1.477 1.509 qs_ot_get_orbitals 99 10.5 0.000 0.001 1.408 1.470 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="208", plot="h2o_128_md", label="(8n/2r/6t)", y=71.394000, yerr=0.000000 PlotPoint: name="209", plot="h2o_128_md_mem", label="(8n/2r/6t)", y=815.181818, yerr=11.621439 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 1.414681E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 82200 MPI messages size (bytes): total size 297.640985E+09 min size 0.000000E+00 max size 26.214400E+06 average size 3.620936E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 44 1441792 32768 < size <= 131072 18560 2432696320 131072 < size <= 4194304 54216 84915781632 4194304 < size <= 16777216 0 0 16777216 < size 8808 210291069504 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3462 67098. MP_Allreduce 9752 812. MP_Sync 52 MP_Alltoall 1474 16505187. MP_SendRecv 2310 360267. MP_ISendRecv 2310 360267. MP_Wait 5214 MP_ISend 2420 1187840. MP_IRecv 2420 1187840. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.023 0.045 104.587 104.589 qs_mol_dyn_low 1 2.0 0.003 0.003 104.180 104.195 qs_forces 11 3.9 0.003 0.003 104.102 104.108 qs_energies 11 4.9 0.001 0.001 99.565 99.581 scf_env_do_scf 11 5.9 0.000 0.001 88.583 88.583 velocity_verlet 10 3.0 0.002 0.002 66.057 66.064 scf_env_do_scf_inner_loop 99 6.5 0.003 0.007 54.464 54.465 init_scf_loop 11 6.9 0.000 0.000 34.026 34.029 qs_scf_new_mos 99 7.5 0.001 0.001 33.201 33.239 qs_scf_loop_do_ot 99 8.5 0.001 0.001 33.201 33.239 dbcsr_multiply_generic 2055 12.4 0.122 0.123 32.505 32.988 prepare_preconditioner 11 7.9 0.000 0.000 31.700 31.704 make_preconditioner 11 8.9 0.000 0.000 31.700 31.704 ot_scf_mini 99 9.5 0.002 0.002 31.189 31.237 make_full_inverse_cholesky 11 9.9 0.000 0.000 25.675 31.113 multiply_cannon 2055 13.4 0.336 0.358 23.986 26.059 multiply_cannon_loop 2055 14.4 0.829 0.841 22.132 23.534 cp_fm_upper_to_full 70 14.2 12.741 18.424 12.741 18.424 ot_mini 99 10.5 0.001 0.001 16.390 16.446 rebuild_ks_matrix 110 8.3 0.000 0.001 16.004 16.113 qs_ks_build_kohn_sham_matrix 110 9.3 0.014 0.014 16.004 16.112 qs_ks_update_qs_env 110 7.6 0.001 0.001 14.446 14.544 qs_ot_get_derivative 99 11.5 0.001 0.001 11.300 11.345 dbcsr_complete_redistribute 325 12.2 1.024 1.044 7.939 11.155 mp_waitall_1 84994 16.7 10.064 10.765 10.064 10.765 copy_fm_to_dbcsr 174 11.2 0.001 0.001 6.722 9.942 multiply_cannon_multrec 8220 15.4 4.082 4.269 9.779 9.872 qs_rho_update_rho_low 110 7.6 0.000 0.001 9.305 9.340 calculate_rho_elec 110 8.6 0.225 0.225 9.305 9.340 transfer_fm_to_dbcsr 11 9.9 0.000 0.000 6.008 9.211 qs_ot_get_p 110 10.4 0.009 0.010 9.058 9.115 cp_fm_cholesky_invert 11 10.9 9.105 9.112 9.105 9.112 mp_alltoall_i22 605 13.7 5.667 8.923 5.667 8.923 sum_up_and_integrate 110 10.3 0.149 0.151 8.681 8.689 integrate_v_rspace 110 11.3 0.003 0.003 8.531 8.540 multiply_cannon_metrocomm3 8220 15.4 0.019 0.020 6.282 7.735 qs_ot_p2m_diag 48 11.0 0.151 0.156 6.857 6.870 init_scf_run 11 5.9 0.000 0.001 6.518 6.518 scf_env_initial_rho_setup 11 6.9 0.000 0.001 6.518 6.518 pw_transfer 1331 11.6 0.075 0.076 6.291 6.331 make_m2s 4110 13.4 0.043 0.044 5.847 6.303 cp_dbcsr_syevd 48 12.0 0.003 0.003 6.252 6.253 fft_wrap_pw1pw2 1111 12.6 0.009 0.009 6.173 6.213 make_images 4110 14.4 0.877 0.919 5.657 6.114 dbcsr_mm_accdrv_process 11614 15.7 4.106 5.056 5.547 5.753 apply_preconditioner_dbcsr 110 12.6 0.000 0.000 5.076 5.479 apply_single 110 13.6 0.000 0.000 5.076 5.479 cp_fm_diag_elpa 48 13.0 0.000 0.000 5.381 5.382 cp_fm_diag_elpa_base 48 14.0 4.822 4.987 5.379 5.379 fft_wrap_pw1pw2_140 451 13.1 1.337 1.343 5.194 5.230 density_rs2pw 110 9.6 0.004 0.005 5.111 5.146 ot_diis_step 99 11.5 0.016 0.016 4.975 4.976 fft3d_ps 1111 14.6 1.307 1.314 4.102 4.135 wfi_extrapolate 11 7.9 0.001 0.001 4.044 4.045 qs_ot_get_derivative_taylor 52 13.0 0.001 0.001 3.529 4.001 rs_pw_transfer 902 11.9 0.010 0.011 3.848 3.920 make_images_data 4110 15.4 0.044 0.047 3.170 3.819 grid_integrate_task_list 110 12.3 3.719 3.781 3.719 3.781 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 3.719 3.723 hybrid_alltoall_any 4261 16.3 0.264 0.575 3.079 3.680 potential_pw2rs 110 12.3 0.021 0.021 3.576 3.604 calculate_dm_sparse 110 9.5 0.001 0.001 3.417 3.473 cp_fm_cholesky_decompose 22 10.9 3.310 3.333 3.310 3.333 qs_ot_get_derivative_diag 47 12.0 0.001 0.001 3.228 3.254 multiply_cannon_sync_h2d 8220 15.4 3.123 3.143 3.123 3.143 cp_dbcsr_sm_fm_multiply 37 9.5 0.001 0.001 3.101 3.106 mp_sum_l 6594 12.7 1.776 2.790 1.776 2.790 grid_collocate_task_list 110 9.6 2.705 2.768 2.705 2.768 mp_alltoall_d11v 2046 13.8 2.628 2.694 2.628 2.694 mp_alltoall_z22v 1111 16.6 2.542 2.575 2.542 2.575 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 2.468 2.503 calculate_first_density_matrix 1 7.0 0.000 0.000 2.320 2.322 qs_env_update_s_mstruct 11 6.9 0.000 0.000 2.233 2.303 build_core_hamiltonian_matrix_ 11 4.9 0.001 0.001 2.103 2.221 qs_ks_update_qs_env_forces 11 4.9 0.000 0.000 2.162 2.173 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="210", plot="h2o_128_md", label="(8n/1r/12t)", y=104.589000, yerr=0.000000 PlotPoint: name="211", plot="h2o_128_md_mem", label="(8n/1r/12t)", y=1275.000000, yerr=53.421140 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/15/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 1.094965E+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 11950464 0.0% 0.0% 100.0% average stack size 0.0 0.0 565.3 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 631.296000E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 10348896 MPI messages size (bytes): total size 4.491514E+12 min size 0.000000E+00 max size 4.537280E+06 average size 434.009000E+03 MPI breakdown and total messages size (bytes): size <= 128 65736 0 128 < size <= 8192 1232 10092544 8192 < size <= 32768 3576680 95640223744 32768 < size <= 131072 1294784 74079797248 131072 < size <= 4194304 5148576 3175955383376 4194304 < size <= 16777216 261888 1145794321408 16777216 < size 0 0 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4020 57505. MP_Allreduce 11129 795. MP_Sync 87 MP_Alltoall 2226 2380771. MP_SendRecv 24320 18752. MP_ISendRecv 24320 18752. MP_Wait 42476 MP_ISend 16020 108028. MP_IRecv 16020 108028. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.028 0.066 234.957 234.961 qs_mol_dyn_low 1 2.0 0.005 0.022 234.317 234.331 qs_forces 11 3.9 0.005 0.008 234.180 234.182 qs_energies 11 4.9 0.001 0.001 228.134 228.156 scf_env_do_scf 11 5.9 0.001 0.001 210.270 210.274 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 182.639 182.641 velocity_verlet 10 3.0 0.003 0.011 141.031 141.033 qs_scf_new_mos 117 7.6 0.001 0.001 136.675 136.993 qs_scf_loop_do_ot 117 8.6 0.001 0.001 136.674 136.993 ot_scf_mini 117 9.6 0.003 0.005 129.933 130.255 dbcsr_multiply_generic 2507 12.6 0.181 0.185 127.880 128.503 multiply_cannon 2507 13.6 0.243 0.250 102.471 104.669 multiply_cannon_loop 2507 14.6 2.383 2.480 99.445 101.152 ot_mini 117 10.6 0.001 0.004 68.120 68.428 qs_ot_get_derivative 117 11.6 0.001 0.002 43.054 43.358 multiply_cannon_multrec 60168 15.6 31.891 34.041 41.484 43.271 qs_ot_get_p 128 10.4 0.001 0.002 39.288 39.620 rebuild_ks_matrix 128 8.3 0.001 0.001 36.893 37.208 qs_ks_build_kohn_sham_matrix 128 9.3 0.016 0.018 36.892 37.208 mp_waitall_1 267128 16.5 31.408 35.303 31.408 35.303 qs_ks_update_qs_env 128 7.6 0.001 0.001 33.112 33.388 qs_ot_p2m_diag 83 11.4 0.079 0.092 31.945 32.035 cp_dbcsr_syevd 83 12.4 0.005 0.005 28.940 28.944 multiply_cannon_sync_h2d 60168 15.6 26.466 28.795 26.466 28.795 init_scf_loop 11 6.9 0.000 0.001 27.533 27.535 apply_preconditioner_dbcsr 128 12.6 0.000 0.001 24.478 25.216 apply_single 128 13.6 0.001 0.001 24.478 25.216 ot_diis_step 117 11.6 0.008 0.011 24.774 24.775 cp_fm_diag_elpa 83 13.4 0.000 0.000 24.709 24.743 cp_fm_diag_elpa_base 83 14.4 24.567 24.616 24.704 24.737 prepare_preconditioner 11 7.9 0.000 0.000 22.627 22.680 make_preconditioner 11 8.9 0.000 0.000 22.627 22.680 make_full_inverse_cholesky 11 9.9 0.000 0.000 21.729 21.967 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 20.228 20.429 multiply_cannon_metrocomm3 60168 15.6 0.119 0.126 15.583 17.536 qs_rho_update_rho_low 128 7.7 0.002 0.009 17.324 17.430 calculate_rho_elec 128 8.7 0.046 0.067 17.322 17.430 sum_up_and_integrate 128 10.3 0.089 0.107 16.995 17.013 integrate_v_rspace 128 11.3 0.004 0.005 16.906 16.927 make_m2s 5014 13.6 0.102 0.110 15.662 16.116 make_images 5014 14.6 0.400 0.420 15.481 15.943 density_rs2pw 128 9.7 0.006 0.008 10.486 13.784 cp_fm_cholesky_invert 11 10.9 13.648 13.658 13.648 13.658 rs_pw_transfer 1046 11.9 0.016 0.019 9.872 13.526 init_scf_run 11 5.9 0.000 0.001 13.393 13.393 scf_env_initial_rho_setup 11 6.9 0.001 0.012 13.392 13.393 pw_transfer 1547 11.6 0.076 0.088 9.737 10.184 wfi_extrapolate 11 7.9 0.001 0.001 10.060 10.060 fft_wrap_pw1pw2 1291 12.7 0.011 0.013 9.533 9.983 mp_sum_l 7950 12.9 8.523 9.970 8.523 9.970 dbcsr_mm_accdrv_process 124484 16.2 4.568 4.879 9.150 9.732 multiply_cannon_metrocomm1 60168 15.6 0.096 0.102 6.878 9.608 make_images_data 5014 15.6 0.068 0.073 7.447 8.901 calculate_dm_sparse 128 9.5 0.001 0.001 8.587 8.697 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 8.454 8.614 qs_ot_get_orbitals 117 10.6 0.001 0.001 8.245 8.359 hybrid_alltoall_any 5200 16.5 0.294 2.262 6.385 8.300 fft3d_ps 1291 14.7 2.149 2.783 7.331 7.928 grid_integrate_task_list 128 12.3 6.993 7.731 6.993 7.731 fft_wrap_pw1pw2_140 523 13.2 1.243 1.303 7.237 7.689 potential_pw2rs 128 12.3 0.009 0.011 7.115 7.342 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.005 6.808 6.822 cp_fm_cholesky_decompose 22 10.9 6.638 6.653 6.638 6.653 mp_waitany 16020 13.9 3.027 6.402 3.027 6.402 mp_alltoall_d11v 2415 14.1 5.032 6.396 5.032 6.396 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.869 5.983 grid_collocate_task_list 128 9.7 4.673 5.963 4.673 5.963 rs_pw_transfer_RS2PW_140 139 11.5 0.283 0.309 2.260 5.632 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="400", plot="h2o_256_md", label="(8n/12r/1t)", y=234.961000, yerr=0.000000 PlotPoint: name="401", plot="h2o_256_md_mem", label="(8n/12r/1t)", y=597.909091, yerr=6.721127 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/16/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1420239992832 0.0% 0.0% 100.0% flops 32 x 32 x 32 1943472701440 0.0% 0.0% 100.0% flops 22 x 9 x 32 1972057190400 0.0% 0.0% 100.0% flops 9 x 22 x 32 1977770336256 0.0% 0.0% 100.0% flops 22 x 22 x 32 2734287699968 0.0% 0.0% 100.0% flops 32 x 32 x 9 4416300122112 0.0% 0.0% 100.0% flops 32 x 32 x 22 5397700149248 0.0% 0.0% 100.0% flops 9 x 32 x 32 5443971710976 0.0% 0.0% 100.0% flops 22 x 32 x 32 6653743202304 0.0% 0.0% 100.0% flops 9 x 32 x 9 11528891191296 0.0% 0.0% 100.0% flops 22 x 32 x 9 15129160814592 0.0% 0.0% 100.0% flops 9 x 32 x 22 15129160814592 0.0% 0.0% 100.0% flops 22 x 32 x 22 19767995056128 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 93.514751E+12 0.0% 0.0% 100.0% flops max/rank 2.183246E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6755938624 0.0% 0.0% 100.0% number of processed stacks 5975232 0.0% 0.0% 100.0% average stack size 0.0 0.0 1130.7 marketing flops 144.580175E+12 ------------------------------------------------------------------------------- # multiplications 2507 max memory usage/rank 844.488704E+06 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 2406720 MPI messages size (bytes): total size 4.100942E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.703955E+06 MPI breakdown and total messages size (bytes): size <= 128 14916 0 128 < size <= 8192 0 0 8192 < size <= 32768 70860 2317615104 32768 < size <= 131072 722992 55511613440 131072 < size <= 4194304 1375664 1398181724160 4194304 < size <= 16777216 154704 1463834332048 16777216 < size 67584 1181116006400 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3992 58357. MP_Allreduce 11058 960. MP_Sync 87 MP_Alltoall 1969 6453752. 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 0.032 0.068 238.670 238.672 qs_mol_dyn_low 1 2.0 0.003 0.006 238.129 238.145 qs_forces 11 3.9 0.005 0.005 237.971 237.972 qs_energies 11 4.9 0.003 0.007 230.598 230.632 scf_env_do_scf 11 5.9 0.001 0.002 212.076 212.087 scf_env_do_scf_inner_loop 117 6.6 0.003 0.009 168.059 168.061 velocity_verlet 10 3.0 0.002 0.003 148.398 148.402 qs_scf_new_mos 117 7.6 0.001 0.001 121.776 122.473 qs_scf_loop_do_ot 117 8.6 0.001 0.002 121.776 122.472 ot_scf_mini 117 9.6 0.004 0.004 116.584 117.207 dbcsr_multiply_generic 2507 12.6 0.194 0.200 106.963 110.140 multiply_cannon 2507 13.6 0.512 0.592 80.214 90.186 multiply_cannon_loop 2507 14.6 1.564 1.637 76.088 85.191 ot_mini 117 10.6 0.001 0.001 58.682 59.430 mp_waitall_1 214728 16.6 31.047 49.945 31.047 49.945 init_scf_loop 11 6.9 0.005 0.024 43.884 43.886 multiply_cannon_metrocomm3 30084 15.6 0.097 0.107 19.166 39.854 qs_ot_get_p 128 10.4 0.002 0.004 38.851 39.637 prepare_preconditioner 11 7.9 0.001 0.002 38.963 39.051 make_preconditioner 11 8.9 0.005 0.043 38.963 39.051 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.455 38.134 rebuild_ks_matrix 128 8.3 0.001 0.001 36.803 37.612 qs_ks_build_kohn_sham_matrix 128 9.3 0.028 0.070 36.803 37.612 qs_ot_get_derivative 117 11.6 0.001 0.002 35.477 36.115 multiply_cannon_multrec 30084 15.6 20.048 24.629 30.555 35.445 qs_ks_update_qs_env 128 7.6 0.001 0.001 33.079 33.827 qs_ot_p2m_diag 83 11.4 0.188 0.217 32.299 32.485 cp_dbcsr_syevd 83 12.4 0.006 0.006 30.541 30.546 cp_fm_diag_elpa 83 13.4 0.000 0.001 26.147 26.185 cp_fm_diag_elpa_base 83 14.4 25.825 25.938 26.139 26.172 apply_preconditioner_dbcsr 128 12.6 0.000 0.000 22.710 24.729 apply_single 128 13.6 0.001 0.001 22.710 24.728 cp_fm_cholesky_invert 11 10.9 23.645 23.663 23.645 23.663 ot_diis_step 117 11.6 0.014 0.015 22.828 22.831 multiply_cannon_sync_h2d 30084 15.6 17.212 21.069 17.212 21.069 make_m2s 5014 13.6 0.088 0.092 16.977 18.596 make_images 5014 14.6 1.171 1.378 16.769 18.389 sum_up_and_integrate 128 10.3 0.115 0.131 17.924 17.959 integrate_v_rspace 128 11.3 0.004 0.004 17.808 17.848 qs_rho_update_rho_low 128 7.7 0.001 0.001 17.670 17.731 calculate_rho_elec 128 8.7 0.088 0.104 17.669 17.730 qs_ot_get_derivative_diag 77 12.4 0.002 0.002 14.777 15.247 init_scf_run 11 5.9 0.000 0.001 13.192 13.194 scf_env_initial_rho_setup 11 6.9 0.001 0.003 13.192 13.194 density_rs2pw 128 9.7 0.006 0.007 10.638 13.145 rs_pw_transfer 1046 11.9 0.014 0.016 10.217 12.624 multiply_cannon_metrocomm4 27577 15.6 0.109 0.128 3.981 11.472 make_images_data 5014 15.6 0.069 0.078 9.510 11.455 mp_sum_l 7950 12.9 8.027 11.363 8.027 11.363 mp_irecv_dv 69486 16.3 3.773 11.049 3.773 11.049 pw_transfer 1547 11.6 0.087 0.099 10.786 10.972 fft_wrap_pw1pw2 1291 12.7 0.011 0.011 10.560 10.747 dbcsr_mm_accdrv_process 62242 16.2 5.395 6.279 10.008 10.589 hybrid_alltoall_any 5200 16.5 0.354 1.549 7.937 10.324 cp_fm_cholesky_decompose 22 10.9 10.186 10.267 10.186 10.267 wfi_extrapolate 11 7.9 0.001 0.001 9.654 9.655 fft_wrap_pw1pw2_140 523 13.2 1.324 1.340 8.521 8.717 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 7.646 8.451 fft3d_ps 1291 14.7 2.810 2.994 7.880 8.044 potential_pw2rs 128 12.3 0.015 0.018 7.879 8.034 grid_integrate_task_list 128 12.3 7.070 7.771 7.070 7.771 mp_alltoall_d11v 2415 14.1 5.765 7.160 5.765 7.160 calculate_dm_sparse 128 9.5 0.001 0.001 6.921 7.118 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.662 6.669 qs_ot_get_orbitals 117 10.6 0.001 0.001 6.255 6.375 mp_allgather_i34 2507 14.6 2.638 6.081 2.638 6.081 grid_collocate_task_list 128 9.7 4.783 5.953 4.783 5.953 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 5.427 5.686 mp_waitany 11748 13.9 3.115 5.621 3.115 5.621 mp_sum_d 4465 12.1 3.636 5.282 3.636 5.282 dbcsr_complete_redistribute 395 12.7 0.782 0.866 4.202 5.122 rs_pw_transfer_RS2PW_140 139 11.5 0.355 0.389 2.322 4.842 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="402", plot="h2o_256_md", label="(8n/6r/2t)", y=238.672000, yerr=0.000000 PlotPoint: name="403", plot="h2o_256_md_mem", label="(8n/6r/2t)", y=802.727273, yerr=3.077511 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/17/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 2.906045E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 3951168 0.0% 0.0% 100.0% average stack size 0.0 0.0 1697.1 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 965.332992E+06 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 1033760 MPI messages size (bytes): total size 2.695213E+12 min size 0.000000E+00 max size 26.214400E+06 average size 2.607194E+06 MPI breakdown and total messages size (bytes): size <= 128 6424 0 128 < size <= 8192 0 0 8192 < size <= 32768 264 8650752 32768 < size <= 131072 279168 36591108096 131072 < size <= 4194304 654272 987691483136 4194304 < size <= 16777216 65184 925172905552 16777216 < size 28448 745747251200 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4003 58192. MP_Allreduce 11085 1000. MP_Sync 86 MP_Alltoall 1700 9383497. MP_SendRecv 7874 75008. MP_ISendRecv 7874 75008. MP_Wait 21654 MP_ISend 11660 275234. MP_IRecv 11660 275234. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.038 0.067 209.507 209.509 qs_mol_dyn_low 1 2.0 0.003 0.003 208.970 208.983 qs_forces 11 3.9 0.011 0.042 208.833 208.837 qs_energies 11 4.9 0.004 0.018 201.641 201.664 scf_env_do_scf 11 5.9 0.001 0.004 184.739 184.741 scf_env_do_scf_inner_loop 116 6.6 0.004 0.010 140.040 140.048 velocity_verlet 10 3.0 0.002 0.002 133.201 133.205 qs_scf_new_mos 116 7.6 0.001 0.001 97.442 97.786 qs_scf_loop_do_ot 116 8.6 0.001 0.002 97.441 97.786 ot_scf_mini 116 9.6 0.003 0.004 93.051 93.442 dbcsr_multiply_generic 2485 12.5 0.183 0.188 87.557 89.920 multiply_cannon 2485 13.5 0.504 0.526 64.332 73.061 multiply_cannon_loop 2485 14.5 1.117 1.179 60.651 66.751 ot_mini 116 10.6 0.001 0.001 47.759 48.198 init_scf_loop 11 6.9 0.001 0.006 44.577 44.597 mp_waitall_1 169034 16.6 28.932 40.312 28.932 40.312 prepare_preconditioner 11 7.9 0.000 0.000 40.188 40.240 make_preconditioner 11 8.9 0.000 0.002 40.188 40.240 make_full_inverse_cholesky 11 9.9 0.000 0.000 37.653 39.151 rebuild_ks_matrix 127 8.3 0.001 0.001 33.252 33.937 qs_ks_build_kohn_sham_matrix 127 9.3 0.016 0.019 33.251 33.937 qs_ot_get_p 127 10.4 0.001 0.002 29.996 30.507 multiply_cannon_metrocomm3 19880 15.5 0.061 0.068 17.429 30.445 qs_ks_update_qs_env 127 7.6 0.001 0.001 29.819 30.443 qs_ot_get_derivative 116 11.6 0.001 0.002 27.373 27.762 multiply_cannon_multrec 19880 15.5 12.634 15.943 21.858 25.164 qs_ot_p2m_diag 82 11.4 0.262 0.269 24.856 24.969 cp_dbcsr_syevd 82 12.4 0.005 0.005 23.444 23.449 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 20.089 21.636 apply_single 127 13.6 0.001 0.001 20.089 21.636 cp_fm_cholesky_invert 11 10.9 20.995 21.014 20.995 21.014 ot_diis_step 116 11.6 0.017 0.018 20.120 20.122 cp_fm_diag_elpa 82 13.4 0.000 0.000 19.188 19.208 cp_fm_diag_elpa_base 82 14.4 18.694 18.822 19.183 19.203 make_m2s 4970 13.5 0.078 0.082 15.951 17.198 make_images 4970 14.5 1.140 1.243 15.721 16.965 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.883 16.918 calculate_rho_elec 127 8.7 0.129 0.144 16.882 16.918 sum_up_and_integrate 127 10.3 0.133 0.144 16.752 16.799 integrate_v_rspace 127 11.3 0.004 0.006 16.619 16.661 multiply_cannon_sync_h2d 19880 15.5 13.201 15.107 13.201 15.107 init_scf_run 11 5.9 0.000 0.001 11.599 11.600 scf_env_initial_rho_setup 11 6.9 0.001 0.004 11.599 11.599 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.222 11.507 density_rs2pw 127 9.7 0.006 0.006 9.430 11.359 make_images_data 4970 15.5 0.062 0.072 9.347 10.932 pw_transfer 1535 11.6 0.086 0.104 10.088 10.319 rs_pw_transfer 1038 11.9 0.013 0.014 8.319 10.147 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 9.864 10.103 hybrid_alltoall_any 5155 16.4 0.447 2.032 8.000 10.092 multiply_cannon_metrocomm4 17395 15.5 0.066 0.077 3.573 10.008 cp_fm_cholesky_decompose 22 10.9 9.840 9.877 9.840 9.877 mp_irecv_dv 49801 16.2 3.446 9.739 3.446 9.739 dbcsr_mm_accdrv_process 41158 16.2 5.767 5.974 8.721 8.856 fft_wrap_pw1pw2_140 519 13.2 1.397 1.425 8.301 8.550 wfi_extrapolate 11 7.9 0.001 0.001 8.412 8.412 grid_integrate_task_list 127 12.3 7.161 7.857 7.161 7.857 mp_sum_l 7884 12.9 5.473 7.818 5.473 7.818 cp_fm_upper_to_full 104 14.8 5.816 7.469 5.816 7.469 fft3d_ps 1281 14.7 2.722 2.956 7.013 7.224 dbcsr_complete_redistribute 393 12.7 1.177 1.205 5.235 7.081 mp_alltoall_d11v 2401 14.1 5.397 6.835 5.397 6.835 potential_pw2rs 127 12.3 0.020 0.022 6.675 6.809 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.474 6.122 calculate_dm_sparse 127 9.5 0.001 0.001 6.028 6.109 grid_collocate_task_list 127 9.7 4.939 5.863 4.939 5.863 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.749 5.760 copy_fm_to_dbcsr 208 11.6 0.002 0.002 3.836 5.676 mp_allgather_i34 2485 14.5 2.197 5.367 2.197 5.367 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.647 4.858 mp_waitany 11660 13.9 2.762 4.770 2.762 4.770 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.552 4.612 mp_alltoall_i22 712 14.1 2.329 4.364 2.329 4.364 transfer_fm_to_dbcsr 11 9.9 0.017 0.021 2.515 4.305 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="404", plot="h2o_256_md", label="(8n/4r/3t)", y=209.509000, yerr=0.000000 PlotPoint: name="405", plot="h2o_256_md_mem", label="(8n/4r/3t)", y=912.272727, yerr=9.696212 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/18/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410023282688 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963544850432 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444707676160 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019188129792 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019188129792 0.0% 0.0% 100.0% flops 22 x 32 x 22 19624853225472 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 92.796579E+12 0.0% 0.0% 100.0% flops max/rank 4.320339E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705500928 0.0% 0.0% 100.0% number of processed stacks 5927808 0.0% 0.0% 100.0% average stack size 0.0 0.0 1131.2 marketing flops 143.508480E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.144508E+09 # max total images/rank 3 # max 3D layers 1 # MPI messages exchanged 1133160 MPI messages size (bytes): total size 2.008142E+12 min size 0.000000E+00 max size 17.653760E+06 average size 1.772161E+06 MPI breakdown and total messages size (bytes): size <= 128 6996 0 128 < size <= 8192 0 0 8192 < size <= 32768 396 8650752 32768 < size <= 131072 315952 35695099904 131072 < size <= 4194304 709496 778939400192 4194304 < size <= 16777216 69840 660837789680 16777216 < size 30480 532676608000 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3997 58275. MP_Allreduce 11070 1083. MP_Sync 86 MP_Alltoall 1700 12496381. MP_SendRecv 5842 75008. MP_ISendRecv 5842 75008. MP_Wait 22272 MP_ISend 14840 244848. MP_IRecv 14840 244848. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.031 0.070 221.373 221.376 qs_mol_dyn_low 1 2.0 0.003 0.004 220.819 220.833 qs_forces 11 3.9 0.004 0.006 220.694 220.701 qs_energies 11 4.9 0.003 0.008 213.140 213.150 scf_env_do_scf 11 5.9 0.001 0.001 194.837 194.845 velocity_verlet 10 3.0 0.002 0.002 145.214 145.219 scf_env_do_scf_inner_loop 116 6.6 0.003 0.009 137.929 137.931 qs_scf_new_mos 116 7.6 0.001 0.001 97.026 97.365 qs_scf_loop_do_ot 116 8.6 0.001 0.001 97.026 97.364 ot_scf_mini 116 9.6 0.003 0.004 92.376 92.699 dbcsr_multiply_generic 2485 12.5 0.189 0.193 84.422 85.431 multiply_cannon 2485 13.5 0.542 0.573 56.949 61.407 init_scf_loop 11 6.9 0.001 0.005 56.768 56.770 multiply_cannon_loop 2485 14.5 1.813 1.864 52.664 55.767 prepare_preconditioner 11 7.9 0.000 0.001 52.379 52.400 make_preconditioner 11 8.9 0.001 0.002 52.379 52.400 make_full_inverse_cholesky 11 9.9 0.013 0.024 46.042 50.945 ot_mini 116 10.6 0.001 0.001 46.440 46.785 rebuild_ks_matrix 127 8.3 0.001 0.001 31.949 32.273 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.022 31.948 32.272 qs_ot_get_p 127 10.4 0.001 0.002 31.616 31.984 multiply_cannon_multrec 29820 15.5 13.290 18.253 26.160 30.883 mp_waitall_1 146592 16.7 20.253 29.726 20.253 29.726 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.726 29.005 qs_ot_get_derivative 116 11.6 0.001 0.002 26.468 26.797 qs_ot_p2m_diag 82 11.4 0.339 0.385 26.691 26.783 cp_dbcsr_syevd 82 12.4 0.005 0.005 25.032 25.037 cp_fm_cholesky_invert 11 10.9 23.312 23.328 23.312 23.328 make_m2s 4970 13.5 0.094 0.098 22.036 23.052 make_images 4970 14.5 1.928 2.217 21.730 22.741 cp_fm_diag_elpa 82 13.4 0.000 0.000 20.772 20.783 cp_fm_diag_elpa_base 82 14.4 19.722 20.009 20.765 20.774 apply_preconditioner_dbcsr 127 12.6 0.000 0.001 19.320 19.962 apply_single 127 13.6 0.001 0.001 19.320 19.962 ot_diis_step 116 11.6 0.017 0.019 19.800 19.802 multiply_cannon_metrocomm3 29820 15.5 0.050 0.053 7.182 18.056 cp_fm_upper_to_full 104 14.8 11.592 16.888 11.592 16.888 qs_rho_update_rho_low 127 7.7 0.001 0.001 16.568 16.610 calculate_rho_elec 127 8.7 0.172 0.187 16.567 16.609 sum_up_and_integrate 127 10.3 0.139 0.151 16.504 16.528 integrate_v_rspace 127 11.3 0.004 0.005 16.364 16.393 dbcsr_complete_redistribute 393 12.7 1.527 1.650 9.613 13.501 make_images_data 4970 15.5 0.066 0.071 11.829 13.249 dbcsr_mm_accdrv_process 61748 16.2 8.398 9.274 12.447 12.930 init_scf_run 11 5.9 0.000 0.001 12.131 12.132 scf_env_initial_rho_setup 11 6.9 0.001 0.002 12.131 12.132 hybrid_alltoall_any 5155 16.4 0.529 2.218 10.486 11.895 copy_fm_to_dbcsr 208 11.6 0.002 0.003 7.969 11.844 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 11.268 11.491 multiply_cannon_sync_h2d 29820 15.5 10.256 11.115 10.256 11.115 cp_fm_cholesky_decompose 22 10.9 10.483 10.586 10.483 10.586 pw_transfer 1535 11.6 0.087 0.100 10.311 10.460 density_rs2pw 127 9.7 0.006 0.006 8.817 10.336 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.085 10.237 transfer_fm_to_dbcsr 11 9.9 0.001 0.005 6.314 10.116 mp_alltoall_i22 712 14.1 5.598 9.444 5.598 9.444 fft_wrap_pw1pw2_140 519 13.2 1.546 1.578 8.645 8.808 rs_pw_transfer 1038 11.9 0.013 0.014 7.131 8.793 wfi_extrapolate 11 7.9 0.001 0.001 8.711 8.711 grid_integrate_task_list 127 12.3 7.438 8.028 7.438 8.028 multiply_cannon_metrocomm4 24850 15.5 0.083 0.095 2.832 7.446 mp_irecv_dv 75445 16.2 2.680 7.174 2.680 7.174 fft3d_ps 1281 14.7 2.812 2.866 7.033 7.141 mp_alltoall_d11v 2401 14.1 5.995 6.946 5.995 6.946 calculate_dm_sparse 127 9.5 0.001 0.001 6.442 6.539 potential_pw2rs 127 12.3 0.022 0.023 6.110 6.200 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.850 5.884 grid_collocate_task_list 127 9.7 5.111 5.819 5.111 5.819 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 5.060 5.160 mp_sum_l 7884 12.9 3.538 5.011 3.538 5.011 qs_ot_get_orbitals 116 10.6 0.001 0.001 4.679 4.734 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.551 4.649 qs_energies_init_hamiltonians 11 5.9 0.001 0.003 4.552 4.554 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="406", plot="h2o_256_md", label="(8n/3r/4t)", y=221.376000, yerr=0.000000 PlotPoint: name="407", plot="h2o_256_md_mem", label="(8n/3r/4t)", y=1080.818182, yerr=17.943437 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/19/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1410022950912 0.0% 0.0% 100.0% flops 32 x 32 x 32 1924145348608 0.0% 0.0% 100.0% flops 22 x 9 x 32 1957871443968 0.0% 0.0% 100.0% flops 9 x 22 x 32 1963542011904 0.0% 0.0% 100.0% flops 22 x 22 x 32 2714615709696 0.0% 0.0% 100.0% flops 32 x 32 x 9 4377645416448 0.0% 0.0% 100.0% flops 32 x 32 x 22 5350455508992 0.0% 0.0% 100.0% flops 9 x 32 x 32 5395653328896 0.0% 0.0% 100.0% flops 22 x 32 x 32 6594687401984 0.0% 0.0% 100.0% flops 9 x 32 x 9 11444706349056 0.0% 0.0% 100.0% flops 22 x 32 x 9 15019182452736 0.0% 0.0% 100.0% flops 9 x 32 x 22 15019182452736 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 92.796564E+12 0.0% 0.0% 100.0% flops max/rank 5.820057E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6705499488 0.0% 0.0% 100.0% number of processed stacks 1944496 0.0% 0.0% 100.0% average stack size 0.0 0.0 3448.5 marketing flops 143.507742E+12 ------------------------------------------------------------------------------- # multiplications 2485 max memory usage/rank 1.555120E+09 # max total images/rank 1 # max 3D layers 1 # MPI messages exchanged 238560 MPI messages size (bytes): total size 1.321104E+12 min size 0.000000E+00 max size 52.428800E+06 average size 5.537828E+06 MPI breakdown and total messages size (bytes): size <= 128 1452 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 132 8650752 131072 < size <= 4194304 112800 59139686400 4194304 < size <= 16777216 104112 545846722560 16777216 < size 20064 716108490000 ------------------------------------------------------------------------------- - - - DBCSR MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Bcast 14 12. MP_Allreduce 8852 52. MP_Alltoall 9584 804353. MP_ISend 39716 2104723. MP_IRecv 39716 2103824. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 3995 58302. MP_Allreduce 11065 1167. MP_Sync 86 MP_Alltoall 1700 18828174. MP_SendRecv 3810 122880. MP_ISendRecv 3810 122880. MP_Wait 16000 MP_ISend 10600 423612. MP_IRecv 10600 423612. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.030 0.049 197.712 197.719 qs_mol_dyn_low 1 2.0 0.003 0.004 197.256 197.287 qs_forces 11 3.9 0.004 0.006 197.118 197.122 qs_energies 11 4.9 0.001 0.001 189.161 189.170 scf_env_do_scf 11 5.9 0.001 0.001 170.817 170.819 velocity_verlet 10 3.0 0.019 0.037 128.287 128.291 scf_env_do_scf_inner_loop 116 6.6 0.003 0.008 127.184 127.186 qs_scf_new_mos 116 7.6 0.001 0.001 85.708 85.845 qs_scf_loop_do_ot 116 8.6 0.001 0.001 85.707 85.845 ot_scf_mini 116 9.6 0.003 0.004 81.123 81.192 dbcsr_multiply_generic 2485 12.5 0.181 0.190 74.693 75.227 multiply_cannon 2485 13.5 0.560 0.580 54.072 59.540 multiply_cannon_loop 2485 14.5 0.802 0.845 50.957 52.825 init_scf_loop 11 6.9 0.000 0.000 43.470 43.472 ot_mini 116 10.6 0.001 0.001 40.018 40.105 prepare_preconditioner 11 7.9 0.000 0.000 39.413 39.444 make_preconditioner 11 8.9 0.000 0.000 39.413 39.444 make_full_inverse_cholesky 11 9.9 0.015 0.027 37.071 37.519 mp_waitall_1 124680 16.7 26.823 32.762 26.823 32.762 rebuild_ks_matrix 127 8.3 0.001 0.001 31.546 31.694 qs_ks_build_kohn_sham_matrix 127 9.3 0.017 0.018 31.546 31.693 qs_ks_update_qs_env 127 7.6 0.001 0.001 28.543 28.681 qs_ot_get_p 127 10.4 0.002 0.002 28.107 28.196 cp_fm_cholesky_invert 11 10.9 23.857 23.869 23.857 23.869 qs_ot_p2m_diag 82 11.4 0.489 0.495 23.689 23.720 qs_ot_get_derivative 116 11.6 0.001 0.002 22.880 22.938 cp_dbcsr_syevd 82 12.4 0.005 0.006 22.345 22.347 multiply_cannon_metrocomm3 9940 15.5 0.025 0.026 12.936 21.291 multiply_cannon_multrec 9940 15.5 10.290 14.094 17.776 20.550 cp_fm_diag_elpa 82 13.4 0.000 0.000 18.584 18.592 cp_fm_diag_elpa_base 82 14.4 18.319 18.404 18.579 18.587 apply_preconditioner_dbcsr 127 12.6 0.000 0.000 16.984 17.492 apply_single 127 13.6 0.001 0.001 16.984 17.491 qs_rho_update_rho_low 127 7.7 0.001 0.001 17.114 17.150 calculate_rho_elec 127 8.7 0.254 0.265 17.113 17.150 ot_diis_step 116 11.6 0.020 0.020 16.978 16.978 make_m2s 4970 13.5 0.065 0.069 15.922 16.897 make_images 4970 14.5 2.148 2.575 15.618 16.589 sum_up_and_integrate 127 10.3 0.178 0.187 16.451 16.500 integrate_v_rspace 127 11.3 0.004 0.004 16.273 16.330 init_scf_run 11 5.9 0.000 0.001 11.338 11.339 scf_env_initial_rho_setup 11 6.9 0.000 0.001 11.338 11.339 multiply_cannon_sync_h2d 9940 15.5 10.706 10.964 10.706 10.964 pw_transfer 1535 11.6 0.086 0.095 10.575 10.660 fft_wrap_pw1pw2 1281 12.7 0.010 0.011 10.353 10.449 make_images_data 4970 15.5 0.055 0.064 8.944 10.446 hybrid_alltoall_any 5155 16.4 0.837 3.756 8.668 10.107 cp_fm_cholesky_decompose 22 10.9 9.577 9.699 9.577 9.699 density_rs2pw 127 9.7 0.006 0.006 8.367 9.273 qs_ot_get_derivative_diag 76 12.4 0.002 0.002 9.131 9.188 fft_wrap_pw1pw2_140 519 13.2 1.894 1.929 8.856 8.963 multiply_cannon_metrocomm1 9940 15.5 0.030 0.031 4.981 8.333 wfi_extrapolate 11 7.9 0.001 0.001 8.232 8.232 grid_integrate_task_list 127 12.3 7.698 8.214 7.698 8.214 dbcsr_mm_accdrv_process 20590 16.1 3.266 4.450 7.143 7.820 rs_pw_transfer 1038 11.9 0.013 0.013 5.914 6.966 fft3d_ps 1281 14.7 2.750 2.818 6.858 6.937 mp_alltoall_d11v 2401 14.1 5.925 6.585 5.925 6.585 calculate_dm_sparse 127 9.5 0.001 0.001 6.260 6.341 dbcsr_complete_redistribute 393 12.7 2.104 2.169 5.824 6.119 grid_collocate_task_list 127 9.7 5.431 6.098 5.431 6.098 potential_pw2rs 127 12.3 0.026 0.026 5.648 5.721 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 5.337 5.338 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 5.291 5.301 mp_allgather_i34 2485 14.5 1.326 4.907 1.326 4.907 cp_dbcsr_sm_fm_multiply_core 37 10.5 0.000 0.000 4.096 4.156 qs_ot_get_derivative_taylor 40 13.0 0.001 0.001 3.988 4.067 copy_fm_to_dbcsr 208 11.6 0.001 0.001 3.751 4.067 qs_ot_get_orbitals 116 10.6 0.001 0.001 3.854 3.965 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="408", plot="h2o_256_md", label="(8n/2r/6t)", y=197.719000, yerr=0.000000 PlotPoint: name="409", plot="h2o_256_md_mem", label="(8n/2r/6t)", y=1469.363636, yerr=27.116827 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_performance_tests/20/result.log @@@@@@@@@@ Run number: 1 @@@@@@@@@@ ------------------------------------------------------------------------------- - - - DBCSR STATISTICS - - - ------------------------------------------------------------------------------- COUNTER TOTAL BLAS SMM ACC flops 9 x 9 x 32 1430456039424 0.0% 0.0% 100.0% flops 32 x 32 x 32 1962800054272 0.0% 0.0% 100.0% flops 22 x 9 x 32 1986255912960 0.0% 0.0% 100.0% flops 9 x 22 x 32 1992003932160 0.0% 0.0% 100.0% flops 22 x 22 x 32 2753958699008 0.0% 0.0% 100.0% flops 32 x 32 x 9 4454954827776 0.0% 0.0% 100.0% flops 32 x 32 x 22 5444944789504 0.0% 0.0% 100.0% flops 9 x 32 x 32 5492290093056 0.0% 0.0% 100.0% flops 22 x 32 x 32 6712799002624 0.0% 0.0% 100.0% flops 9 x 32 x 9 11613072052224 0.0% 0.0% 100.0% flops 22 x 32 x 9 15239176077312 0.0% 0.0% 100.0% flops 9 x 32 x 22 15239176077312 0.0% 0.0% 100.0% flops 22 x 32 x 22 19911132921856 0.0% 0.0% 100.0% flops inhomo. stacks 0 0.0% 0.0% 0.0% flops total 94.233020E+12 0.0% 0.0% 100.0% flops max/rank 11.786061E+12 0.0% 0.0% 100.0% matmuls inhomo. stacks 0 0.0% 0.0% 0.0% matmuls total 6806383904 0.0% 0.0% 100.0% number of processed stacks 1980288 0.0% 0.0% 100.0% average stack size 0.0 0.0 3437.1 marketing flops 145.650931E+12 ------------------------------------------------------------------------------- # multiplications 2529 max memory usage/rank 3.193389E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 101160 MPI messages size (bytes): total size 1.144970E+12 min size 0.000000E+00 max size 104.857600E+06 average size 11.318403E+06 MPI breakdown and total messages size (bytes): size <= 128 572 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 44 2883584 131072 < size <= 4194304 45648 35433480192 4194304 < size <= 16777216 44720 382939955200 16777216 < size 10176 726592466352 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 4043 58558. MP_Allreduce 11184 1500. MP_Sync 88 MP_Alltoall 1724 36993632. MP_SendRecv 1806 218624. MP_ISendRecv 1806 218624. MP_Wait 9876 MP_ISend 6456 1080169. MP_IRecv 6456 1080169. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.080 0.145 335.227 335.234 qs_mol_dyn_low 1 2.0 0.003 0.003 334.543 334.557 qs_forces 11 3.9 0.004 0.005 334.438 334.443 qs_energies 11 4.9 0.001 0.001 324.565 324.581 scf_env_do_scf 11 5.9 0.001 0.001 300.326 300.339 velocity_verlet 10 3.0 0.002 0.002 239.944 239.952 scf_env_do_scf_inner_loop 118 6.6 0.003 0.009 162.543 162.545 init_scf_loop 11 6.9 0.000 0.000 137.479 137.481 prepare_preconditioner 11 7.9 0.000 0.000 132.069 132.090 make_preconditioner 11 8.9 0.000 0.000 132.069 132.090 make_full_inverse_cholesky 11 9.9 0.036 0.039 107.796 129.083 qs_scf_new_mos 118 7.6 0.001 0.001 106.989 107.120 qs_scf_loop_do_ot 118 8.6 0.001 0.001 106.988 107.119 ot_scf_mini 118 9.6 0.004 0.004 101.861 101.977 dbcsr_multiply_generic 2529 12.6 0.213 0.225 88.026 88.376 cp_fm_upper_to_full 106 14.8 53.654 76.716 53.654 76.716 multiply_cannon 2529 13.6 0.688 0.738 62.707 65.972 multiply_cannon_loop 2529 14.6 1.074 1.094 58.425 59.496 ot_mini 118 10.6 0.001 0.001 48.513 48.700 dbcsr_complete_redistribute 397 12.7 4.005 4.062 30.800 43.662 cp_fm_cholesky_invert 11 10.9 41.897 41.907 41.897 41.907 rebuild_ks_matrix 129 8.3 0.001 0.001 41.140 41.393 qs_ks_build_kohn_sham_matrix 129 9.3 0.018 0.018 41.139 41.392 copy_fm_to_dbcsr 210 11.7 0.001 0.002 26.962 39.824 qs_ks_update_qs_env 129 7.6 0.001 0.001 37.759 37.994 qs_ot_get_p 129 10.4 0.002 0.002 37.374 37.611 transfer_fm_to_dbcsr 11 9.9 0.030 0.030 24.230 36.984 mp_waitall_1 104580 16.8 32.676 35.779 32.676 35.779 mp_alltoall_i22 720 14.1 21.916 34.838 21.916 34.838 qs_ot_p2m_diag 84 11.4 0.891 0.895 32.330 32.349 cp_dbcsr_syevd 84 12.4 0.006 0.006 30.330 30.332 qs_ot_get_derivative 118 11.6 0.002 0.002 27.145 27.265 cp_fm_diag_elpa 84 13.4 0.000 0.000 26.103 26.104 cp_fm_diag_elpa_base 84 14.4 21.549 23.011 26.098 26.098 qs_rho_update_rho_low 129 7.7 0.001 0.001 23.344 23.363 calculate_rho_elec 129 8.7 0.483 0.484 23.343 23.363 multiply_cannon_metrocomm3 10116 15.6 0.025 0.027 22.028 23.126 sum_up_and_integrate 129 10.3 0.324 0.327 21.861 21.946 integrate_v_rspace 129 11.3 0.004 0.004 21.536 21.620 make_m2s 5058 13.6 0.076 0.079 19.810 21.280 ot_diis_step 118 11.6 0.022 0.023 21.261 21.261 make_images 5058 14.6 3.085 3.286 19.326 20.795 apply_preconditioner_dbcsr 129 12.6 0.000 0.000 20.348 20.788 apply_single 129 13.6 0.001 0.001 20.348 20.788 multiply_cannon_multrec 10116 15.6 10.386 12.221 18.268 18.406 multiply_cannon_sync_h2d 10116 15.6 14.397 14.460 14.397 14.460 hybrid_alltoall_any 5245 16.5 1.324 3.096 11.613 13.970 pw_transfer 1559 11.6 0.094 0.096 13.838 13.883 make_images_data 5058 15.6 0.064 0.069 11.466 13.735 fft_wrap_pw1pw2 1301 12.7 0.011 0.011 13.599 13.646 init_scf_run 11 5.9 0.000 0.001 13.618 13.620 scf_env_initial_rho_setup 11 6.9 0.000 0.000 13.618 13.619 cp_fm_cholesky_decompose 22 10.9 12.185 12.248 12.185 12.248 fft_wrap_pw1pw2_140 527 13.2 3.125 3.212 11.872 11.911 mp_alltoall_d11v 2429 14.1 10.262 11.175 10.262 11.175 qs_ot_get_derivative_diag 78 12.4 0.002 0.002 10.795 10.873 wfi_extrapolate 11 7.9 0.001 0.001 10.314 10.315 density_rs2pw 129 9.7 0.006 0.006 10.070 10.225 dbcsr_mm_accdrv_process 20934 16.1 4.328 6.132 7.627 9.517 grid_integrate_task_list 129 12.3 8.675 8.872 8.675 8.872 fft3d_ps 1301 14.7 2.879 2.893 8.607 8.668 qs_energies_init_hamiltonians 11 5.9 0.001 0.002 8.296 8.297 calculate_dm_sparse 129 9.5 0.001 0.001 7.090 7.127 rs_scatter_matrices 140 9.7 3.934 4.870 6.804 7.000 cp_dbcsr_sm_fm_multiply 37 9.5 0.002 0.002 6.743 6.817 copy_dbcsr_to_fm 187 11.8 0.004 0.004 6.607 6.738 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="410", plot="h2o_256_md", label="(8n/1r/12t)", y=335.234000, yerr=0.000000 PlotPoint: name="411", plot="h2o_256_md_mem", label="(8n/1r/12t)", y=2893.909091, yerr=153.450291 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 1.259704E+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 0.025 0.042 85.589 85.590 qs_energies 1 2.0 0.000 0.000 84.974 84.981 ls_scf 1 3.0 0.000 0.000 83.975 83.982 dbcsr_multiply_generic 111 6.7 0.014 0.015 72.485 72.703 multiply_cannon 111 7.7 0.018 0.020 55.756 57.042 multiply_cannon_loop 111 8.7 0.226 0.240 52.369 53.808 ls_scf_main 1 4.0 0.000 0.000 52.503 52.503 density_matrix_trs4 2 5.0 0.002 0.003 46.830 46.894 ls_scf_init_scf 1 4.0 0.000 0.001 28.448 28.450 ls_scf_init_matrix_S 1 5.0 0.000 0.000 27.262 27.332 mp_waitall_1 11031 10.9 22.373 25.333 22.373 25.333 matrix_sqrt_Newton_Schulz 2 6.5 0.005 0.013 25.157 25.181 multiply_cannon_multrec 2664 9.7 8.204 8.986 15.571 17.335 multiply_cannon_sync_h2d 2664 9.7 13.506 15.419 13.506 15.419 make_m2s 222 7.7 0.009 0.011 13.074 13.493 make_images 222 8.7 0.099 0.109 13.052 13.470 multiply_cannon_metrocomm1 2664 9.7 0.010 0.011 9.704 12.886 make_images_data 222 9.7 0.004 0.005 7.610 8.219 dbcsr_mm_accdrv_process 4760 10.4 0.584 0.692 6.984 7.987 hybrid_alltoall_any 227 10.6 0.216 1.839 6.580 7.939 multiply_cannon_metrocomm3 2664 9.7 0.009 0.010 5.361 7.391 dbcsr_mm_accdrv_process_sort 4760 11.4 6.201 7.114 6.201 7.114 calculate_norms 4752 9.8 5.509 6.264 5.509 6.264 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.040 5.178 mp_sum_l 887 5.1 3.132 5.089 3.132 5.089 make_images_sizes 222 9.7 0.000 0.000 0.737 3.789 mp_alltoall_i44 222 10.7 0.737 3.788 0.737 3.788 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 2.339 3.690 arnoldi_extremal 4 6.8 0.000 0.000 3.562 3.596 arnoldi_normal_ev 4 7.8 0.007 0.017 3.562 3.596 multiply_cannon_metrocomm4 2442 9.7 0.012 0.015 2.059 3.585 mp_irecv_dv 6231 10.9 2.042 3.562 2.042 3.562 build_subspace 16 8.4 0.009 0.012 3.437 3.442 ls_scf_post 1 4.0 0.000 0.000 3.024 3.031 ls_scf_store_result 1 5.0 0.000 0.000 2.847 2.884 dbcsr_special_finalize 555 9.7 0.005 0.006 2.381 2.808 dbcsr_merge_single_wm 555 10.7 0.458 0.583 2.372 2.799 dbcsr_matrix_vector_mult 304 9.0 0.006 0.013 2.415 2.642 make_images_pack 222 9.7 2.198 2.623 2.200 2.624 dbcsr_sort_data 658 11.4 2.173 2.552 2.173 2.552 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.421 2.518 dbcsr_matrix_vector_mult_local 304 10.0 2.046 2.434 2.048 2.436 buffer_matrices_ensure_size 222 8.7 1.750 2.067 1.750 2.067 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.835 1.836 rebuild_ks_matrix 3 7.3 0.000 0.000 1.825 1.826 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 1.825 1.826 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="500", plot="h2o_32_nrep3_ls", label="(8n/12r/1t)", y=85.590000, yerr=0.000000 PlotPoint: name="501", plot="h2o_32_nrep3_ls_mem", label="(8n/12r/1t)", y=1140.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 2.172293E+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 0.032 0.057 90.617 90.619 qs_energies 1 2.0 0.000 0.000 90.097 90.101 ls_scf 1 3.0 0.000 0.000 88.766 88.770 dbcsr_multiply_generic 111 6.7 0.015 0.016 74.528 74.848 multiply_cannon 111 7.7 0.026 0.035 52.678 57.533 ls_scf_main 1 4.0 0.000 0.000 54.793 54.800 multiply_cannon_loop 111 8.7 0.135 0.146 49.848 53.666 density_matrix_trs4 2 5.0 0.002 0.003 49.010 49.223 ls_scf_init_scf 1 4.0 0.001 0.002 30.317 30.318 mp_waitall_1 9105 10.9 21.001 29.845 21.001 29.845 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.096 29.182 multiply_cannon_multrec 1332 9.7 13.274 16.794 22.531 27.211 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.003 26.607 26.623 multiply_cannon_metrocomm3 1332 9.7 0.007 0.008 11.514 21.252 make_m2s 222 7.7 0.006 0.008 15.385 15.949 make_images 222 8.7 1.365 1.699 15.354 15.919 dbcsr_mm_accdrv_process 4041 10.4 0.326 0.539 8.857 10.444 dbcsr_mm_accdrv_process_sort 4041 11.4 8.380 9.906 8.380 9.906 make_images_data 222 9.7 0.004 0.005 8.967 9.851 hybrid_alltoall_any 227 10.6 0.543 2.573 8.240 9.803 mp_sum_l 887 5.1 5.010 8.382 5.010 8.382 multiply_cannon_metrocomm4 1221 9.7 0.007 0.009 3.177 7.807 mp_irecv_dv 3311 11.0 3.156 7.746 3.156 7.746 calculate_norms 2376 9.8 6.073 6.706 6.073 6.706 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.806 6.669 multiply_cannon_sync_h2d 1332 9.7 4.888 6.397 4.888 6.397 apply_matrix_preconditioner 6 5.3 0.000 0.000 5.068 5.250 arnoldi_extremal 4 6.8 0.000 0.000 4.788 4.815 arnoldi_normal_ev 4 7.8 0.001 0.005 4.788 4.815 build_subspace 16 8.4 0.014 0.021 4.528 4.531 ls_scf_post 1 4.0 0.000 0.000 3.656 3.660 ls_scf_store_result 1 5.0 0.000 0.000 3.382 3.476 dbcsr_matrix_vector_mult 304 9.0 0.010 0.022 3.169 3.424 dbcsr_matrix_vector_mult_local 304 10.0 2.737 3.226 2.739 3.228 mp_allgather_i34 111 8.7 0.826 3.218 0.826 3.218 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.670 2.742 multiply_cannon_metrocomm1 1332 9.7 0.003 0.004 1.167 2.629 dbcsr_data_new 4174 10.1 2.109 2.421 2.109 2.421 make_images_pack 222 9.7 1.862 2.206 1.865 2.208 dbcsr_sort_data 436 11.2 1.841 2.113 1.841 2.113 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.897 1.899 rebuild_ks_matrix 3 7.3 0.000 0.000 1.882 1.884 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 1.881 1.884 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="502", plot="h2o_32_nrep3_ls", label="(8n/6r/2t)", y=90.619000, yerr=0.000000 PlotPoint: name="503", plot="h2o_32_nrep3_ls_mem", label="(8n/6r/2t)", y=1778.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 2.886173E+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 0.032 0.056 92.864 92.867 qs_energies 1 2.0 0.000 0.000 92.251 92.256 ls_scf 1 3.0 0.000 0.000 90.826 90.831 dbcsr_multiply_generic 111 6.7 0.016 0.016 74.887 75.217 ls_scf_main 1 4.0 0.000 0.000 56.482 56.487 multiply_cannon 111 7.7 0.033 0.090 51.808 56.087 multiply_cannon_loop 111 8.7 0.116 0.130 49.031 53.426 density_matrix_trs4 2 5.0 0.002 0.003 50.459 50.667 mp_waitall_1 7281 11.0 23.167 33.250 23.167 33.250 ls_scf_init_scf 1 4.0 0.000 0.001 30.700 30.702 ls_scf_init_matrix_S 1 5.0 0.000 0.000 29.430 29.509 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 27.035 27.049 multiply_cannon_multrec 888 9.7 12.646 15.240 21.214 24.309 multiply_cannon_metrocomm3 888 9.7 0.004 0.004 10.856 23.036 make_m2s 222 7.7 0.006 0.007 16.175 16.933 make_images 222 8.7 1.582 1.849 16.137 16.894 make_images_data 222 9.7 0.004 0.004 9.364 10.526 hybrid_alltoall_any 227 10.6 0.643 2.955 8.879 10.316 dbcsr_mm_accdrv_process 3754 10.4 0.309 0.488 8.078 9.312 mp_sum_l 887 5.1 5.235 8.857 5.235 8.857 dbcsr_mm_accdrv_process_sort 3754 11.4 7.644 8.824 7.644 8.824 multiply_cannon_sync_h2d 888 9.7 6.078 7.467 6.078 7.467 multiply_cannon_metrocomm4 777 9.7 0.004 0.005 2.475 7.343 mp_irecv_dv 2335 11.1 2.459 7.278 2.459 7.278 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.980 6.905 multiply_cannon_metrocomm1 888 9.7 0.003 0.003 3.580 6.884 arnoldi_extremal 4 6.8 0.000 0.000 5.465 5.485 arnoldi_normal_ev 4 7.8 0.001 0.005 5.465 5.485 build_subspace 16 8.4 0.014 0.020 5.151 5.159 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.794 5.051 calculate_norms 1584 9.8 4.363 4.796 4.363 4.796 dbcsr_matrix_vector_mult 304 9.0 0.010 0.020 3.578 3.861 mp_allgather_i34 111 8.7 0.892 3.693 0.892 3.693 ls_scf_post 1 4.0 0.000 0.000 3.644 3.651 dbcsr_matrix_vector_mult_local 304 10.0 3.013 3.587 3.015 3.589 ls_scf_store_result 1 5.0 0.000 0.000 3.404 3.468 ls_scf_dm_to_ks 2 5.0 0.000 0.000 2.951 3.079 dbcsr_data_new 4116 9.9 2.109 2.446 2.109 2.446 dbcsr_sort_data 325 11.1 1.878 2.085 1.878 2.085 make_images_sizes 222 9.7 0.000 0.000 0.968 2.019 mp_alltoall_i44 222 10.7 0.967 2.019 0.967 2.019 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.972 1.974 rebuild_ks_matrix 3 7.3 0.000 0.000 1.952 1.954 qs_ks_build_kohn_sham_matrix 3 8.3 0.004 0.032 1.952 1.954 make_images_pack 222 9.7 1.634 1.880 1.637 1.882 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="504", plot="h2o_32_nrep3_ls", label="(8n/4r/3t)", y=92.867000, 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/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 3.404718E+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 0.039 0.052 98.969 98.970 qs_energies 1 2.0 0.000 0.000 98.351 98.355 ls_scf 1 3.0 0.000 0.000 96.693 96.699 dbcsr_multiply_generic 111 6.7 0.017 0.018 79.604 79.887 ls_scf_main 1 4.0 0.000 0.000 60.101 60.101 multiply_cannon 111 7.7 0.043 0.098 52.149 56.537 density_matrix_trs4 2 5.0 0.002 0.003 53.834 53.953 multiply_cannon_loop 111 8.7 0.152 0.164 47.100 50.780 ls_scf_init_scf 1 4.0 0.000 0.001 33.319 33.321 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.042 32.133 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.464 29.472 mp_waitall_1 6369 11.0 23.360 29.097 23.360 29.097 multiply_cannon_multrec 1332 9.7 14.155 17.351 22.043 25.096 make_m2s 222 7.7 0.007 0.008 21.512 22.956 make_images 222 8.7 3.137 3.589 21.462 22.908 multiply_cannon_metrocomm3 1332 9.7 0.003 0.004 9.451 17.760 make_images_data 222 9.7 0.004 0.004 11.980 13.537 hybrid_alltoall_any 227 10.6 0.797 3.749 11.363 13.030 dbcsr_mm_accdrv_process 3641 10.4 0.294 0.475 7.527 9.045 mp_sum_l 887 5.1 4.361 8.662 4.361 8.662 dbcsr_mm_accdrv_process_sort 3641 11.4 7.087 8.554 7.087 8.554 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 3.353 6.892 multiply_cannon_metrocomm4 1110 9.7 0.005 0.007 2.098 6.112 multiply_cannon_sync_h2d 1332 9.7 5.444 6.062 5.444 6.062 mp_irecv_dv 3229 10.9 2.073 6.026 2.073 6.026 arnoldi_extremal 4 6.8 0.000 0.000 5.738 5.749 arnoldi_normal_ev 4 7.8 0.001 0.005 5.738 5.748 multiply_cannon_metrocomm1 1332 9.7 0.003 0.003 2.754 5.409 build_subspace 16 8.4 0.014 0.021 5.398 5.406 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.642 4.828 mp_allgather_i34 111 8.7 2.164 4.523 2.164 4.523 calculate_norms 2376 9.8 4.180 4.518 4.180 4.518 dbcsr_matrix_vector_mult 304 9.0 0.010 0.021 3.771 4.085 dbcsr_matrix_vector_mult_local 304 10.0 3.188 3.693 3.190 3.695 dbcsr_sort_data 658 11.4 3.088 3.444 3.088 3.444 ls_scf_post 1 4.0 0.000 0.000 3.274 3.279 dbcsr_special_finalize 555 9.7 0.006 0.007 2.829 3.218 dbcsr_merge_single_wm 555 10.7 0.537 0.646 2.820 3.209 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.131 3.191 ls_scf_store_result 1 5.0 0.000 0.000 3.021 3.075 dbcsr_data_release 10477 10.7 1.586 2.399 1.586 2.399 make_images_pack 222 9.7 1.629 1.981 1.631 1.984 qs_ks_update_qs_env 3 6.3 0.000 0.000 1.980 1.982 dbcsr_finalize 304 7.8 0.049 0.061 1.806 1.980 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="506", plot="h2o_32_nrep3_ls", label="(8n/3r/4t)", y=98.970000, yerr=0.000000 PlotPoint: name="507", plot="h2o_32_nrep3_ls_mem", label="(8n/3r/4t)", y=2802.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 4.735533E+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 0.044 0.059 92.692 92.692 qs_energies 1 2.0 0.000 0.000 91.900 91.905 ls_scf 1 3.0 0.000 0.000 89.927 89.932 dbcsr_multiply_generic 111 6.7 0.018 0.018 70.640 70.880 ls_scf_main 1 4.0 0.000 0.000 56.833 56.833 multiply_cannon 111 7.7 0.079 0.128 52.561 56.145 multiply_cannon_loop 111 8.7 0.087 0.095 49.942 51.277 density_matrix_trs4 2 5.0 0.002 0.003 49.646 49.729 ls_scf_init_scf 1 4.0 0.000 0.001 29.693 29.695 ls_scf_init_matrix_S 1 5.0 0.000 0.000 28.372 28.403 mp_waitall_1 5436 11.0 24.140 28.056 24.140 28.056 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 26.266 26.283 multiply_cannon_multrec 444 9.7 13.734 16.159 20.799 22.110 multiply_cannon_metrocomm1 444 9.7 0.002 0.002 10.878 15.901 make_m2s 222 7.7 0.005 0.005 13.375 14.325 make_images 222 8.7 2.037 2.475 13.308 14.256 multiply_cannon_metrocomm3 444 9.7 0.001 0.002 6.010 13.666 hybrid_alltoall_any 227 10.6 0.801 3.811 7.954 9.521 make_images_data 222 9.7 0.003 0.004 8.062 9.450 multiply_cannon_sync_h2d 444 9.7 6.679 8.625 6.679 8.625 dbcsr_mm_accdrv_process 3003 10.4 0.351 0.410 6.761 7.856 dbcsr_mm_accdrv_process_sort 3003 11.4 6.398 7.465 6.398 7.465 arnoldi_extremal 4 6.8 0.000 0.000 6.330 6.346 arnoldi_normal_ev 4 7.8 0.002 0.005 6.330 6.346 build_subspace 16 8.4 0.015 0.020 5.920 5.931 mp_sum_l 887 5.1 2.709 4.649 2.709 4.649 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.426 4.642 dbcsr_matrix_vector_mult 304 9.0 0.011 0.021 4.367 4.526 dbcsr_matrix_vector_mult_local 304 10.0 3.698 4.173 3.700 4.175 ls_scf_dm_to_ks 2 5.0 0.000 0.000 3.710 3.806 multiply_cannon_metrocomm4 333 9.7 0.001 0.002 1.605 3.780 mp_irecv_dv 1241 11.2 1.591 3.758 1.591 3.758 calculate_norms 792 9.8 3.627 3.729 3.627 3.729 dbcsr_multiply_generic_mpsum_f 86 7.8 0.000 0.000 1.889 3.726 mp_allgather_i34 111 8.7 1.192 3.612 1.192 3.612 ls_scf_post 1 4.0 0.000 0.000 3.402 3.407 ls_scf_store_result 1 5.0 0.000 0.000 3.186 3.223 make_images_sizes 222 9.7 0.000 0.000 0.932 3.220 mp_alltoall_i44 222 10.7 0.932 3.220 0.932 3.220 dbcsr_finalize 304 7.8 0.062 0.077 2.192 2.306 dbcsr_data_new 4608 9.7 1.798 2.185 1.798 2.185 dbcsr_merge_all 275 8.9 0.477 0.522 2.051 2.147 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.136 2.137 rebuild_ks_matrix 3 7.3 0.000 0.000 2.105 2.105 qs_ks_build_kohn_sham_matrix 3 8.3 0.000 0.001 2.105 2.105 qs_energies_init_hamiltonians 1 3.0 0.001 0.003 1.956 1.956 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="508", plot="h2o_32_nrep3_ls", label="(8n/2r/6t)", y=92.692000, yerr=0.000000 PlotPoint: name="509", plot="h2o_32_nrep3_ls_mem", label="(8n/2r/6t)", y=3720.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 8.888685E+09 # max total images/rank 2 # max 3D layers 1 # MPI messages exchanged 4440 MPI messages size (bytes): total size 770.525954E+09 min size 0.000000E+00 max size 399.069120E+06 average size 173.541888E+06 MPI breakdown and total messages size (bytes): size <= 128 640 0 128 < size <= 8192 0 0 8192 < size <= 32768 0 0 32768 < size <= 131072 0 0 131072 < size <= 4194304 640 468025344 4194304 < size <= 16777216 0 0 16777216 < size 3160 770057961712 ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - MESSAGE PASSING PERFORMANCE - - - ------------------------------------------------------------------------------- ROUTINE CALLS AVE VOLUME [Bytes] MP_Group 4 MP_Bcast 1026 284089. MP_Allreduce 3123 21388. MP_Sync 4 MP_Alltoall 47 88727262. MP_SendRecv 42 732600. MP_ISendRecv 42 732600. MP_Wait 267 MP_ISend 180 3337386. MP_IRecv 180 3339494. ------------------------------------------------------------------------------- ------------------------------------------------------------------------------- - - - T I M I N G - - - ------------------------------------------------------------------------------- SUBROUTINE CALLS ASD SELF TIME TOTAL TIME MAXIMUM AVERAGE MAXIMUM AVERAGE MAXIMUM CP2K 1 1.0 0.079 0.098 106.990 106.991 qs_energies 1 2.0 0.000 0.000 105.568 105.576 ls_scf 1 3.0 0.000 0.000 102.607 102.615 dbcsr_multiply_generic 111 6.7 0.024 0.027 75.478 75.628 ls_scf_main 1 4.0 0.000 0.000 64.727 64.728 density_matrix_trs4 2 5.0 0.002 0.003 55.495 55.565 multiply_cannon 111 7.7 0.137 0.236 49.126 51.956 multiply_cannon_loop 111 8.7 0.099 0.101 46.028 46.465 ls_scf_init_scf 1 4.0 0.000 0.001 34.077 34.078 ls_scf_init_matrix_S 1 5.0 0.000 0.000 32.413 32.445 matrix_sqrt_Newton_Schulz 2 6.5 0.001 0.002 29.720 29.727 mp_waitall_1 4527 11.1 22.589 25.825 22.589 25.825 make_m2s 222 7.7 0.005 0.005 22.609 23.797 make_images 222 8.7 3.566 3.873 22.501 23.688 multiply_cannon_multrec 444 9.7 17.756 18.451 22.385 23.026 hybrid_alltoall_any 227 10.6 1.651 3.605 12.840 15.674 make_images_data 222 9.7 0.003 0.004 13.059 15.235 multiply_cannon_metrocomm3 444 9.7 0.001 0.001 10.913 11.689 multiply_cannon_sync_h2d 444 9.7 8.787 8.835 8.787 8.835 arnoldi_extremal 4 6.8 0.000 0.000 7.863 7.871 arnoldi_normal_ev 4 7.8 0.003 0.009 7.863 7.871 build_subspace 16 8.4 0.025 0.036 7.300 7.308 dbcsr_matrix_vector_mult 304 9.0 0.017 0.033 5.761 5.790 ls_scf_dm_to_ks 2 5.0 0.000 0.000 5.323 5.415 dbcsr_matrix_vector_mult_local 304 10.0 5.085 5.402 5.088 5.404 apply_matrix_preconditioner 6 5.3 0.000 0.000 4.880 5.111 dbcsr_mm_accdrv_process 1814 10.4 0.242 0.354 4.445 4.572 dbcsr_mm_accdrv_process_sort 1814 11.4 4.111 4.243 4.111 4.243 mp_allgather_i34 111 8.7 1.215 3.901 1.215 3.901 ls_scf_post 1 4.0 0.000 0.000 3.803 3.811 ls_scf_store_result 1 5.0 0.000 0.000 3.518 3.562 make_images_sizes 222 9.7 0.000 0.000 1.430 3.513 mp_alltoall_i44 222 10.7 1.430 3.512 1.430 3.512 calculate_norms 792 9.8 3.232 3.287 3.232 3.287 dbcsr_finalize 304 7.8 0.082 0.089 3.064 3.159 dbcsr_merge_all 275 8.9 0.884 0.918 2.851 2.939 qs_energies_init_hamiltonians 1 3.0 0.006 0.010 2.931 2.931 dbcsr_complete_redistribute 5 7.6 1.434 1.479 2.772 2.884 qs_ks_update_qs_env 3 6.3 0.000 0.000 2.572 2.574 matrix_ls_to_qs 2 6.0 0.000 0.000 2.416 2.519 rebuild_ks_matrix 3 7.3 0.000 0.000 2.505 2.507 qs_ks_build_kohn_sham_matrix 3 8.3 0.001 0.001 2.505 2.507 dbcsr_sort_data 325 11.1 2.434 2.491 2.434 2.491 dbcsr_data_new 6591 9.6 1.870 2.452 1.870 2.452 dbcsr_new_transposed 4 7.5 0.242 0.251 2.286 2.300 dbcsr_frobenius_norm 74 6.6 2.059 2.133 2.185 2.210 dbcsr_add_d 103 6.2 0.000 0.000 2.125 2.202 dbcsr_add_anytype 103 7.2 0.859 0.890 2.125 2.201 dbcsr_data_release 12724 10.6 1.965 2.162 1.965 2.162 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="510", plot="h2o_32_nrep3_ls", label="(8n/1r/12t)", y=106.991000, yerr=0.000000 PlotPoint: name="511", plot="h2o_32_nrep3_ls_mem", label="(8n/1r/12t)", y=6996.000000, yerr=0.000000 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ~~~~~~~~~ RESULT ~~~~~~~~~ RESULT file: /scratch/snx3000/mkrack/rt/../rt/CRAY-XC50-gnu/d1376dc4969b25e1013bb42635e3ab0fe8739442_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 593.182720E+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 774572143. 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 0.049 0.123 391.494 391.496 qs_mol_dyn_low 1 2.0 0.004 0.024 390.069 390.090 qs_forces 5 3.8 0.005 0.016 389.946 389.949 qs_energies 5 4.8 0.002 0.020 385.788 385.833 scf_env_do_scf 5 5.8 0.000 0.001 369.096 369.099 scf_env_do_scf_inner_loop 105 6.6 0.003 0.011 323.633 323.641 qs_scf_new_mos 105 7.6 0.000 0.001 258.208 258.713 qs_scf_loop_do_ot 105 8.6 0.001 0.002 258.208 258.712 ot_scf_mini 105 9.6 0.003 0.008 247.246 247.623 dbcsr_multiply_generic 1445 12.2 0.125 0.133 145.830 148.127 qs_ot_get_p 112 10.4 0.001 0.001 146.646 146.962 velocity_verlet 4 3.0 0.002 0.009 139.585 139.587 qs_ot_p2m_diag 40 11.0 0.020 0.030 132.176 132.296 cp_dbcsr_syevd 40 12.0 0.002 0.002 128.116 128.120 multiply_cannon 1445 13.2 0.282 0.296 119.770 124.855 multiply_cannon_loop 1445 14.2 2.828 2.995 116.098 121.718 cp_fm_syevd 40 13.0 0.000 0.000 120.393 120.563 cp_fm_redistribute_end 40 14.0 56.815 113.414 56.830 113.425 cp_fm_syevd_base 40 14.0 56.559 113.159 56.559 113.159 ot_mini 105 10.6 0.001 0.001 68.171 68.335 mp_waitall_1 488190 16.1 48.238 62.330 48.238 62.330 multiply_cannon_metrocomm3 69360 15.2 0.208 0.224 29.681 47.490 qs_ot_get_derivative 55 11.6 0.001 0.001 45.954 46.173 init_scf_loop 7 6.6 0.000 0.006 45.356 45.361 multiply_cannon_multrec 69360 15.2 28.934 34.036 38.704 44.118 rebuild_ks_matrix 110 8.4 0.000 0.000 39.572 39.823 qs_ks_build_kohn_sham_matrix 110 9.4 0.012 0.024 39.571 39.823 prepare_preconditioner 7 7.6 0.000 0.001 38.877 38.911 make_preconditioner 7 8.6 0.000 0.002 38.877 38.911 qs_ks_update_qs_env 112 7.6 0.001 0.001 36.315 36.533 qs_rho_update_rho_low 110 7.6 0.001 0.001 34.231 34.762 calculate_rho_elec 110 8.6 0.030 0.033 34.230 34.761 rs_pw_transfer 690 11.5 0.011 0.013 32.883 34.063 multiply_cannon_sync_h2d 69360 15.2 27.962 32.599 27.962 32.599 make_full_inverse_cholesky 7 9.6 0.000 0.000 28.372 28.443 density_rs2pw 110 9.6 0.006 0.007 26.903 28.101 apply_preconditioner_dbcsr 62 12.6 0.000 0.000 23.403 23.732 apply_single 62 13.6 0.000 0.000 23.402 23.732 ot_new_cg_direction 55 11.6 0.001 0.001 21.407 21.409 cp_fm_cholesky_invert 7 10.6 18.621 18.645 18.621 18.645 sum_up_and_integrate 60 10.3 0.028 0.031 17.233 17.253 qs_ot_get_orbitals 105 10.6 0.001 0.001 16.722 17.232 integrate_v_rspace 60 11.3 0.002 0.003 17.205 17.226 multiply_cannon_metrocomm1 69360 15.2 0.097 0.107 6.305 16.634 qs_ot_get_derivative_taylor 37 12.8 0.001 0.001 16.307 16.458 mp_sum_l 4764 12.2 14.789 16.064 14.789 16.064 pw_transfer 1645 12.4 0.084 0.112 15.351 15.857 fft_wrap_pw1pw2 1425 13.5 0.013 0.016 15.208 15.720 mp_sendrecv_dv 168740 12.6 15.406 15.696 15.406 15.696 rs_pw_transfer_RS2PW_30 110 11.6 1.542 1.627 13.733 15.343 init_scf_run 5 5.8 0.000 0.001 13.559 13.561 scf_env_initial_rho_setup 5 6.8 0.000 0.008 13.559 13.561 check_diag 80 13.5 8.594 8.886 13.164 13.330 fft_wrap_pw1pw2_240 915 15.0 1.174 1.272 12.428 12.867 calculate_dm_sparse 110 9.5 0.000 0.001 12.607 12.843 potential_pw2rs 60 12.3 0.003 0.004 12.542 12.663 qs_ot_get_derivative_diag 18 12.0 0.000 0.001 12.489 12.615 qs_vxc_create 110 10.4 0.002 0.005 11.828 11.900 make_m2s 2890 13.2 0.076 0.083 10.956 11.766 make_images 2890 14.2 0.244 0.264 10.852 11.662 fft3d_pb 915 16.0 2.369 2.666 10.510 11.074 dbcsr_mm_accdrv_process 154766 15.8 6.125 6.363 9.644 10.511 make_full_single_inverse 7 9.6 0.001 0.006 9.944 9.978 cp_fm_cholesky_decompose 14 10.2 9.112 9.125 9.112 9.125 mp_alltoall_z22v 2340 17.7 7.784 8.747 7.784 8.747 xc_rho_set_and_dset_create 110 12.4 0.076 0.097 8.055 8.315 xc_vxc_pw_create 60 11.3 0.039 0.051 8.023 8.095 acc_transpose_blocks 69360 15.2 0.358 0.390 7.438 7.945 ------------------------------------------------------------------------------- ~ ~ ~ ~ DATA POINTS ~ ~ ~ ~ PlotPoint: name="601", plot="h2o_512_md", label="(64n/12r/1t)", y=391.496000, yerr=0.000000 PlotPoint: name="602", plot="h2o_512_md_mem", label="(64n/12r/1t)", y=562.600000, yerr=3.322650 ~ ~ ~ ~ END DATA POINTS ~ ~ ~ ~~~~~~ END RESULT ~~~~~~~~ ========= END RESULTS =========== CommitSHA: d1376dc4969b25e1013bb42635e3ab0fe8739442 Summary: empty Status: OK